How Startups Grow with a $0 Marketing Budget

A tiny marketing budget is not a death sentence. Stacy’s Brain digs through the latest studies and reports to find growth tactics that still work at $0, from founder-led content to partnerships. Free weekly read for founders who want to build on evidence.

Subscribe here

Sometimes the line between life and death is very thin. You only notice it afterward…

A heart operation lasting several hours is over. A small child is about to be moved to intensive care.

You want to think, “The hardest part is behind us.”

But a lot has to change hands between the operating room and the ICU. Equipment, connections, information about the patient, responsibility for their care…

The team at Great Ormond Street Hospital in London wanted to reduce the risks of that transition.

Where would you start?

More training? More detailed checklists? New equipment? A look at how other hospitals do it?

All reasonable possibilities.

But one of the team’s sources of inspiration was rather surprising.

The story goes that two doctors were resting after long operations when Formula 1 came on television. Watching a pit stop, they noticed a similarity to their own work.

A car stops. Several people get to work at once. A series of connected tasks is completed. Then the car is on its way again.

So they went to meet the team.

Ferrari; Case study

What Were the Doctors Looking for at Ferrari?

Speed is the first answer that comes to mind.

But think about what the hospital actually needed. They weren’t trying to complete an ICU handover in a few seconds.

They wanted people working simultaneously to finish their tasks without getting in each other’s way or missing something important.

From the outside, a pit stop looks like people moving incredibly fast.

It’s easy to miss something else inside all that speed: everyone knows where to stand, what to wait for, and when to move out of the way.

Making room for someone else to do their job matters as much as doing your own.

The hospital team recorded its handover process for review. A choreographer even became involved, helping people work out how to move without obstructing one another.

A children’s hospital, Ferrari, and choreography.

Not things you expect to find in the same sentence.

They also assigned someone to coordinate the whole transition. The anesthetist held that responsibility until the handover was complete, after which it passed to the intensive care doctor.

Another idea they took from Ferrari was to discuss possible problems in advance: what could go wrong, what would we do if it did, and how serious would it be?

I think there’s a valuable distinction here.

Everyone knowing their own job isn’t the same as everyone knowing how the whole job will move forward.

You can put excellent specialists in the same room. The transitions between them still need to be designed.

Understanding More by Talking Less

Under the new protocol, the equipment transfer and checks came first. Patient information followed. Questions and the care plan came afterward.

People weren’t all trying to explain things at once. Those who needed to listen weren’t occupied with another task at the same time.

Read that and it’s hard not to think about your own meetings.

Someone is explaining something important. Someone else is trying to share their screen. A third person is messaging about the previous topic.

Afterward, we say, “But we discussed this.”

We may have. Whether the other person actually received the information is a separate question.

The hospital’s new approach combined ideas from Formula 1 and aviation. A study published in 2007 observed 50 patient handovers before and after the change.

The average number of technical errors per handover fell from 5.42 to 3.15. Information omissions fell from 2.09 to 1.07. Handover time fell from 10.8 minutes to 9.4 minutes.

It was a small before-and-after study. We can’t draw conclusions about mortality from it. But it offers concrete evidence that the same transition could be performed with fewer errors and less lost information; Study

Think about that. You haven’t changed anyone’s profession. You’ve organized what happens between them more carefully.

How Does This Transition Work in Our Companies?

Someone speaks with a customer.

They understand the need. Take notes. Send them to the relevant team.

Then the customer waits.

Ask the person who took the call, and the job is done. The notes have been sent.

Ask the person who received them, and a piece of information is missing. They’re waiting for it.

The person who could supply it doesn’t know anyone is expecting something from them.

Everyone in the company is working. The customer’s request isn’t moving.

Sometimes a person fills this gap. They remember who is waiting for what. They message, call, and remind people.

When that person takes a few days off, we discover how important their job really was.

Now Add AI

AI summarizes the call. Extracts the customer’s request. Creates a task. Notifies the relevant person.

Preparation that used to take half an hour now takes a few minutes.

That’s a useful gain.

But if the next person still takes over the work in the same way, we may simply be producing tasks much faster.

More summaries. More notifications. More waiting work.

And everything looks beautifully organized on screen.

• “Call completed.”

• “Request analyzed.”

• “Forwarded to the relevant team.”

The customer’s question may still be unanswered.

I think this is one of the places to look in an AI project. When an operation finishes, what do we expect to happen next? How will we know it happened?

What Can We Borrow from Ferrari Tomorrow?

In today’s pit stops, completion signals from the mechanics are part of the decision to release the car. The person in charge also checks the space around the car and the pit lane. Finishing one part doesn’t, by itself, establish that the whole system is ready to go; Formula 1’s explanation

You can try a small experiment in your own company.

Choose a recurring piece of work that passes between two people. Preparing a quote, processing a refund, or booking an appointment, perhaps.

Look at the last ten examples. What happened between the moment the work was sent and the moment someone actually took it over?

Then clarify four things:

• What information does the next person need?

• Who will take over, and how will they acknowledge it?

• What outcome will tell us the work is complete?

• What happens if something is missing or an operation fails?

Suppose a customer wants Friday delivery, but stock availability can’t be confirmed.

You could write “Forwarded to sales” and move on.

Or you could identify who will check stock, when the check is due, and who will get back to the customer. You could also specify who gets notified if the information doesn’t arrive in time.

In the second case, a temporary system problem can be caught before it turns into a forgotten customer.

The jobs you might give AI become more concrete, too. Flag missing information. Route work to the right person. Follow up on a waiting step. Perform an authorized operation. Record the outcome.

This is what we look at when designing an AI workflow at Next Big App: how will the work move forward after the conversation or analysis? How will it connect to existing systems? Who will follow through on the result?

Some of the results that surprise even us with enterprise clients come from these details. When we combine AI with user experience engineering, work that has depended on someone remembering to chase it for years starts moving on its own. Customers have less to explain again. Employees spend less time gathering missing information from different places.

That’s also one of the things our clients appreciate most: the small daily frictions their teams no longer have to deal with. Individually, these changes can look minor. Together, especially in larger organizations, their effect can be much greater than initially expected.

Sometimes what’s needed is a new assistant. Sometimes it’s a few missing steps between tools you already use.

To find them, how busy people are may not be the first thing to examine.

Look for where the work loses its owner.

The question Ferrari helped the hospital ask is a useful starting point for us, too: while everyone is doing their own job, who’s managing the transition between them?

- “I’ve got a dinner proposition for you.”

- “Go on.”

- “I’ll send you the location of a restaurant. You go there on Wednesday evening. There’ll be five other people at your table.”

- “Who?”

- “You don’t know them.”

- “Do you?”

- “Well… sort of.”

You tighten your grip on your phone.

- “So what do you need from me?”

- “A small fee for my services.”

- “Oh, right. We’re paying for dinner upfront.”

- “No, you’ll pay for your food separately. At the restaurant.”

A pause.

- “So I pay you. Then I go to a restaurant and pay again. To have dinner with five people I’ve never met.”

- “Exactly.”

- “Do I at least get dessert?”

- “No.”

Most of us would say, “Good luck with that,” and keep walking.

But this exact proposition has a name, an app, and an operation behind it: Timeleft.

And once you think about it, you realize there’s one question missing from that strange conversation:

How were you going to find those five people yourself?

A Full City. An Empty Evening.

Imagine you’ve just moved to a new city.

You’ve unpacked. Figured out your commute. Found somewhere nearby that makes decent coffee.

It’s Friday, and you’d like to go out. Your phone can show you hundreds of restaurants. You can browse photos, read reviews, and book a table in a few taps.

Finding someone to go with is another matter.

You don’t even have to be new in town. Some friends may have moved away. Others may be in a different stage of life. Perhaps you never built much of a social circle in the first place. Your days run between work and home, you see the same faces, and opportunities to meet anyone new are getting harder to find.

You’d like to socialize. You’re willing to set aside an evening.

You’re just not sure where to start.

Of course, you could go to an event alone. But walking through the door brings a fresh set of questions. Who do you approach? Does everyone already know each other? How do you join a conversation halfway through?

Timeleft organizes that first step.

• You answer a personality quiz and choose an evening.

• The platform puts a group together, books the restaurant, and provides conversation starters.

• You pay for your own food and drinks at the venue.

There’s an easily overlooked detail in this arrangement: everyone at the table has come to meet new people.

That removes some of the hesitation you might feel about striking up a conversation with a stranger at the next table. You don’t need to find an excuse to introduce yourself.

Being there gives you something in common before anyone says a word.

The Infrastructure of Adult Friendship

At school, many friendships grew out of repeated encounters.

You were in the same class. Saw each other during breaks. Worked on an assignment together, then crossed paths again the next day.

You were becoming part of each other’s lives before you became friends.

Some of those settings disappear in adulthood. Making room for a new relationship takes a more deliberate effort: finding somewhere to go, making time, taking the first step, then arranging to meet again.

None of these tasks is impossible. Together, though, they become an invisible to-do list standing between you and the experience you want.

This is where Timeleft’s offer starts to make sense.

You hand part of that list over to the platform. In its standard matching process, you don’t choose each person at your table. The platform assembles the group and selects the venue.

There’s a tradeoff: you might not click with everyone.

There’s also a relief: you don’t have to work out every detail to make the evening happen.

When we build products, it’s easy to assume that more choice means a better experience. More filters. More profiles. More personalization.

But every option asks the user to make another decision.

Sometimes the value lies in being able to trust someone else with a few of those decisions.

Think of the relief you feel on a well-organized trip or at a thoughtfully planned gathering. Someone has taken care of the details. You get to take part.

When Our Job Ends, the Customer’s May Just Be Starting

For anyone building a product, this story raises a useful question:

When we say “done,” how much work does the customer still have left?

• Someone signs up for an online course. They may still have no idea when they’ll study, where to get help when they’re stuck, or how to put the first lesson into practice.

• An analytics tool delivers a report. The manager still has to figure out what decision to make with it.

• A new member joins a professional community. They look at a channel with hundreds of people in it, then leave without posting a word.

On the company’s side, each event can register as a success: a new enrollment, a generated report, an activated membership.

The outcome the customer wanted may still be some distance away.

We get used to leaving this work to the user. We expect them to make time, find the confidence, coordinate with other people, and put what they’ve learned into practice.

Then we wonder why they don’t come back.

Timeleft draws our attention to that gap.

Could we introduce a new community member to a small group instead of leaving them to find their own way in? Could we help a customer make their first decision after delivering the report? Could the first week of a course include a working session where participants actually produce something?

These changes can take more effort than adding another feature. They require us to understand people’s circumstances, think through the details, and follow through on the outcome.

In return, something the customer has been trying to do finally moves forward.

Your First Version Could Be a Restaurant Reservation

One of the most encouraging details in Timeleft’s story is how the dinner concept began.

According to an interview with cofounder Maxime Barbier, the early tests used Typeform, WhatsApp, and Stripe.

Those tools were enough to test the basic assumptions: would people pay for an evening like this? Would they show up? Would they enjoy it?

The first version of an idea can be as tangible as a well-organized evening.

Of course, getting people to the table doesn’t guarantee a good experience. Someone might not show up. The venue might be a poor fit. Conversation might stall. A pleasant dinner might never turn into an ongoing friendship.

So the quality of the product is determined well beyond the booking screen.

How many people actually came? How did the evening go? Did they want to return? Did anyone stay in touch?

Each question follows the promise a little further into the customer’s life.

Take this seriously, and the boundaries of the product begin to expand. The restaurant, the arrival experience, the first few minutes of conversation, and the way people reconnect afterward all become part of the design.

The product keeps working after the phone goes back into your pocket.

Follow an Ordinary Customer Through Their Day

To spot a similar opportunity in your own business, follow one customer this week.

Find out what happens after they sign up, make a payment, or download their report.

Where do they stop? Who are they waiting for? What do they put off? At what point do they think, “I’ll deal with that later”?

Then take responsibility for one task standing in their way.

Arrange the first meeting. Work through the first exercise together. Help them make the first decision.

See what changes in their life.

You might find a better service, a new revenue model, or the beginnings of another business. Some of the work keeping people from what they want sits in places nobody has claimed responsibility for yet.

Which brings us back to the stranger on the street.

We can now see a little more clearly what they would have to earn that fee for.

Sending a restaurant location takes seconds.

Finding the people, setting expectations, and creating an evening where they can meet takes work.

The meal still costs extra. But you might leave with someone new in your life.

How?

Platforms like Spotify pay small royalties when a song is streamed. Each play is worth a fraction of a cent. Musicians complain about this all the time. How can platforms pay so little? How did music become so “cheap”?

But billions of streams still add up to millions of dollars. And that was exactly what Smith was counting on.

He didn’t have a single hit. Or a real fan base. Yet millions kept landing in his account, year after year.

He hadn’t written the songs himself. As you’ve probably guessed, AI had.

The kind of songs you could make today by downloading an app like Suno. Except he didn’t make a handful. He made hundreds of thousands.

AI songs have become surprisingly enjoyable. You’ll happily listen to them, especially the ones you made yourself. But who has the time to sit through everyone else’s AI experiments?

So who listened to Smith’s hundreds of thousands of songs and made him all that money?

Nobody.

The “listeners” were automated accounts pretending to be people. Bots. Thousands of them, playing the songs billions of times.

AI wrote the songs. Software played them. The platforms treated those plays as genuine listening and paid out, a fraction of a cent at a time.

Quite a setup.

Eventually, it was uncovered, and charges followed. It sounds like a straightforward fraud case, but it has been described as the world’s first AI music fraud case.

That isn’t even the strangest part.

Fine, the bots play the songs and you collect the money. But why generate hundreds of thousands of tracks instead of five or ten?

Hoping at least one might become a hit?

Exactly the opposite. He wanted none of them to stand out.

A single song getting a billion plays would make platforms suspicious. Smith knew that. In a 2018 email cited by prosecutors, he explained the strategy: he needed lots of content, with relatively few streams per track, to stay under the radar.

Artists usually try to rise above the crowd, make the best music, attract the most listeners, and eventually get paid. Smith was producing music to disappear into the crowd.

The scheme ran for years before he was caught. As mentioned, it became the first U.S. criminal case involving AI-assisted music streaming fraud. Prosecutors sought at least 46 months in prison. The defense asked for no prison time. On October 6, the sentence came down: 18 months.

Yes, it was a first. Yes, the scheme was devilishly clever. Yes, the sentence seems laughably light. But the sentence isn’t why I’m telling this story.

For years, the counters ticked up, reports arrived, and payments went out. Everything looked like a music career.

What made that possible?

AI Makes a Real Difference. It Doesn’t Take Sides.

The answer is AI. But not in the way you hear about every day.

Someone is always promising you can “get rich with AI.” Most of it is nonsense. This story isn’t, because we can see exactly what AI changed: scale, rather than creativity. Without it, hundreds of thousands of songs would have taken far more time and money to produce.

Fake streams didn’t begin with AI. According to the indictment, Smith started working with an AI music producer around 2018. The fraud already existed. What was missing was enough music. AI filled that gap.

AI is powerful. It doesn’t care whose purposes it serves. Whatever you put in its hands, it scales up.

We keep asking the same question about AI: Can it write a good song?

Here, that was the wrong question. The songs didn’t have to be good. They didn’t even need anyone to listen. They only needed to look as though someone had.

We know the familiar story: production gets cheaper, everyone starts producing, competition intensifies, and attention becomes harder to earn. Smith turned that story upside down. While the rest of us complained about the crowd, he used it as cover.

So the useful question goes beyond “What can I do with AI?”

What has AI made cheaper, and whose metrics does that make less trustworthy?

The same AI can help a musician find the sound they’ve been hearing in their head. It can also produce the material needed to make an audience appear to exist.

The audio file alone won’t tell you which happened.

Who Is the Counter Counting?

Read this as a story about measurement. Something similar could be happening in your own reports.

A streaming record tells you something very narrow: a track was played.

We fill in the rest. Someone chose it, listened, perhaps liked it, perhaps played it again. Of course, every stream isn’t a declaration of love. Background music counts too. But the logic behind the payout assumes that somewhere, a person had an experience.

Smith’s system contained the record of that experience, without anyone having it.

Here’s the trap: you have two signals. Lots of music. Lots of listening. Taken separately, they might suggest a thriving market. But the same machine produced both. Neither independently validates the other.

It’s like a sitcom laugh track. You hear laughter because someone pressed a button, not because anyone found the joke funny.

That’s where this story reaches beyond music: as producing things gets easier, so does producing the signals that suggest those things found an audience.

Who Paid the Bill?

It might look like a closed loop with no victims. A computer makes a song, a bot plays it, and nobody gets hurt.

But the money wasn’t created inside that loop. In the basic model Spotify describes, royalties come from a shared pool and are distributed according to each recipient’s share of total streams. Fake streams can shrink the share attributable to genuine listening. The Justice Department says Smith’s earnings were diverted from other musicians and rights holders.

The defense’s main argument was that no artist suffered any perceptible harm.

That captures the nature of the harm quite well. Nobody lost a fan. The same people kept loving the same songs. But that doesn’t mean there was no damage. Thousands of people’s shares quietly eroded, a fraction of a cent at a time, without any one of them noticing it in their own pocket.

That’s what makes it so insidious.

Your competitor isn’t persuading your customers to switch. They’re taking a cut of the money allocated according to your customers’ behavior.

Which is why debating whether AI music is as good as human music misses the point here. The songs could have been much worse. The system would still have paid.

The Innocent Version of the Same Trap

This time, there’s no fraud. The intentions are good. But the mechanism looks familiar.

You have an idea for an app. You build a prototype with AI, then ask simulated personas representing your target audience to evaluate it:

• “The price is a little high, but I’d pay if there were a weekly plan.”

• “I love that feature, but there are too many notifications.”

• “I already do this in Excel. Why would I switch?”

Three different people. Three different opinions. A working prototype. It feels like customer feedback.

But there are no customers. Those voices aren’t independent people. They’re echoes of your own brief. Adding a hundred more personas makes the echo louder. It doesn’t strengthen the evidence.

Don’t get me wrong: this exercise can be useful. It can uncover a question you missed, a feature that conflicts with another, or a need worth investigating. It’s excellent for generating questions. The moment you use it to supply the answers, you’re listening to the laugh track.

Automation can play the same game. The report was generated. The message was sent. The record was created. All of those things really happened. But the benefit you expected may not have materialized.

And once you reward the system for those intermediate steps, it starts optimizing for them. There’s an old rule: when a measure becomes a target, it stops being a good measure. Smith’s story is one of its most expensive illustrations.

Automation itself isn’t the culprit. You might truly love an AI-generated song. An assistant might buy exactly the right product on your behalf. Who pressed the button matters less than what happened afterward.

Was a real need met outside the system that produced the activity?

The Real-Person Test

Before trusting a number, ask three questions:

1. Did someone outside the system measuring this signal generate it?

2. Did that person give something up for it: money, time, an existing habit?

3. If I shut the system down, would the signal still arrive?

If the answer to all three is no, what you have is the echo of your own voice, rather than evidence.

Smith’s file contained everything: music, streams, money.

Only one thing was missing: someone who heard a song and thought, “I want to play that again.”

Production got cheaper. Finding a real audience didn’t.

If you can’t measure the difference, you’ll call noise growth.

One is plain gray. The other has an American flag and an enormous eagle on it.

Which one sells more?

If you picked the eagle, I can see the logic; Stand out. Be different. The gray pair isn’t exactly begging for attention.

Except, according to Jacques Spitzer, gray is the bestseller.

Now guess what most of the ads selling all that gray underwear look like.

Close-ups of the fabric?

Something about softness and comfort?

The gray pair photographed from every possible angle?

No.

They show the eagle.

The eagle gets your attention. You buy the gray ones.

What makes us look and what makes us buy can be two different things. Fair enough. But follow that thought a little further and it raises an awkward question about how we build a customer journey.

Suppose we take our best-selling product, our most-clicked ad, and our highest-converting page. Put all three together. Have we built the best possible route to a sale?

I’m not convinced.

A lot can change between the scroll and the checkout.

• When you’re scrolling, you’re deciding whether something deserves a moment of your attention.

• Once you’re browsing the products, you’re working out which one suits you.

• By checkout, you may be wondering whether you trust the brand, whether the product will live up to expectations, and whether it’s worth the money.

All of that can happen in a few minutes.

One plausible reading of the underwear story is that the eagle earned a closer look. Then, with the products in front of them, people chose something they’d actually want to wear every day.

Each part of that journey has a job to do.

• Trying to explain everything in the first ad would be exhausting.

• Repeating everything at every step would be confusing.

• And by the time someone is ready to pay, an answer to the question they had five minutes ago may be no help at all.

We need to understand the steps people take and design for what they need at each one.

The oddity that sparks curiosity in an ad might offer very little reassurance at checkout. The simplicity that makes paying feel easy might make an ad easy to ignore. An impression that helps us early on can even become a liability later.

Your winning page has a history

There’s another complication: each stage inherits whatever happened before it.

A page converts beautifully.

Great.

Who’s landing on it?

People who already know the brand?

Someone sent by a friend?

Someone wondering what on earth they just saw in an ad?

Even one person could read that page differently depending on the ad that brought them there. Promise comfort, and they’ll look for evidence of comfort. Show them an unusual design, and that’s what they’ll want to see.

The ad shapes what they’re looking for before the page has had a chance to say anything.

This is why winning an ad test and winning a page test don’t necessarily add up to a winning combination.

Change the ad and you change who arrives, what they expect, and what they’re trying to find out. Remember the U.S. Open story 👇

Calling something “our best page” leaves out quite a lot. It was best for a particular audience, arriving with particular expectations. Connect it to another ad and you may be running a different experiment altogether.

With this many moving parts, it’s a wonder anyone buys anything.

Imagine a language-learning app. It has quick lessons for travelers and a comprehensive program for people who want to study regularly. Both are genuine parts of the product.

The ad team tests two approaches.

• One promotes steady learning.

• The other pitches the travel lessons: “Order dinner on vacation without guessing.”

The travel ad gets more clicks. We have a winner.

Meanwhile, the team working on the product page finds that highlighting the comprehensive program gets more of its current visitors to start a trial. Another winner.

Naturally, the teams put the two together.

Someone clicks because they’d like a few useful phrases for their trip. They land on a page full of weekly goals, proficiency levels, and long-term progress. The lessons they wanted really are in the app. They might leave without ever finding that out.

Come Monday’s meeting, clicks are up and the page test looks good. Both teams have made reasonable, evidence-based decisions. The experience they’ve assembled still leaves the customer stranded.

Putting travel photos everywhere won’t necessarily fix it. A page that takes people straight to the travel lessons might help. Or the less-clicked ad might bring in more people who stick with regular study. The better choice depends on what the business is trying to achieve.

When do we call it a win?

That’s the question I keep coming back to with the eagle boxers.

We need to judge the ad, the page, and the buying experience on the jobs they’re meant to do. We also need to test how they work together.

An ad’s value depends on what happens to the people it brings in. A page’s performance depends on where those people came from and what they expected. And a sale still has to deliver something worthwhile when the customer uses the product.

So after asking which ad won, ask what happened to the people who clicked it.

Did they buy the gray pair and enjoy wearing it? Leave without finding what they wanted? Start a trial and never come back?

You won’t necessarily find those answers in the first chart.

Of course companies need separate teams. Someone has to run the ads, someone has to build the pages, and someone has to make the product. But customers carry the promise of an ad onto the next screen, long after the ad team’s report has ended.

Customers don’t follow our org chart.

Sometimes the most useful change we can make is to the point at which we declare success. The company could already be losing a customer while a team is celebrating a win.

And this goes well beyond marketing. Think about the decisions you’ve seen at work or in your own life that seemed impossible to explain. “Why would anyone do that? It makes no sense.”

Before we call those decisions irrational, it’s worth asking how much of the journey we’ve actually seen. There may be more going into them than our explanation allows.

One scores 80%. Another gets 85%. The latest release beats them both. We glance at the colorful bars, say, “Technology really has come a long way,” and get on with our lives.

But have you ever wondered what sits behind those percentages?

What do these tests actually ask? What is the AI trying to figure out? And how would we do if we faced the same challenge?

Here’s a suggestion: pause this article for a few minutes and try one of the games in ARC-AGI-3.

You’ll find yourself in a visual environment. You have to work out what you’re supposed to do and how the rules work.

You make a move, observe the result, form a hypothesis, and change your mind when necessary. The benchmark is designed to measure how AI agents learn this way in unfamiliar environments.

Once you’ve played for a while, look at the models’ results. As of September 24, 2026, GPT-6 Astra, running at the High setting with a setup called Provider Adapter, is listed at 99.9%.

That score incorporates task completion and action efficiency relative to a human baseline. It doesn’t measure the full range of human intelligence, and your attempt at the public demo isn’t directly comparable to it. But it gives us a useful starting point for thinking about a machine’s ability to learn its way through something unfamiliar.

Now let’s leave those colored squares behind and walk into a meeting room.

Imagine someone exceptionally smart joins your team. Grasp problems quickly, make connections nobody else sees, and get up to speed on unfamiliar subjects. Look at something you’ve spent days wrestling with and ask, “Could we be trying to solve the wrong problem here?”

This arrival changes more than the quality of your meetings.

You start asking for their advice. You hand certain tasks over to them. A project you once considered too difficult comes back onto the agenda because they’re now on the team.

Your sense of what’s possible changes.

I think this is one of the least discussed aspects of AI.

We all carry a mental map of the paths we believe are open to us and the ones we consider closed.

• “I’m not a technical person.”

• “I’m no good with numbers.”

• “My language skills aren’t strong enough.”

• “We’d need a much bigger team to pull that off.”

Some of these statements reflect real obstacles. Others are the residue of a failure years ago, a grade we received at school, or a judgment someone once made about us.

Eventually, we become so familiar with this map that we stop checking whether the roads marked “closed” still are.

An idea occurs to us, and we dismiss it before looking into it. An opportunity appears, and we decide it isn’t for people like us. We want to learn something, but assume we’ve left it too late.

Progress in AI may require us to redraw that map.

Imagine someone who runs a small manufacturing business. They’ve spent years putting off the idea of selling overseas. They don’t know how to find customers, lack confidence in their English, and worry they won’t be able to answer a buyer’s questions.

Give that person an AI tool, and they could use it to polish the emails they already write in Turkish.

Or they could prepare for the first customer conversation they’ve been avoiding for years.

They could rehearse the questions a potential buyer might ask. Practice explaining their product in English. Find gaps in their proposal. Have an unfamiliar term explained through different examples until it makes sense.

None of this automatically wins them a customer. They still have to establish whether their product meets the buyer’s needs, check their costs, deliver on time, and keep their promises in the real world.

But they now have a way to attempt that conversation.

To me, this is one of the biggest changes: things we once ruled out before starting become things we can actually begin.

Seen this way, learning to use AI takes on a different meaning.

It involves more than typing instructions into a tool. It means learning to:

• Recognize what you don’t know.

• Ask for the information you’re missing.

• Try the proposed approach.

• Investigate why it didn’t work.

You’re learning how to work with that smart person who just walked into the room.

Over time, you may get better at doing some of those things yourself. For others, you may continue to rely on assistance. To understand where you’re making progress, you need to assess both the result and your own understanding.

I think this also presents a valuable opportunity for our children.

When a child encounters a difficult problem, we care about more than whether they get the answer. We also care about how they respond to the difficulty.

• Do they give up immediately?

• Do they try things at random?

• Do they have a theory?

• Can they handle discovering that their theory was wrong?

• After getting help, can they explain what they learned?

You can observe all of this by opening one of the ARC games together.

Ask them to explain what they think will happen. Let them make a move. If the result surprises them, ask what they would change. When they get stuck, ask the AI for a hint. Then have them explain, in their own words, why that hint helped.

There’s something a child can learn about themselves here: not being able to figure something out immediately doesn’t mean it’s beyond their reach.

I think adults need that experience too.

In the coming years, some of the limits we currently place on ourselves may no longer apply.

But we won’t necessarily notice when those limits change. Every so often, we’ll need to walk up to them and try again.

So this week, alongside one of your routine tasks, bring your AI tool something you’ve been putting off for a long time.

Before saying, “Do this for me,” ask: “What would I need to learn to do this? How could we set up a small first experiment?”

A benchmark table shows us which problems machines can solve.

Its impact on our lives may be to reopen questions we stopped allowing ourselves to ask years ago.

Machines can keep learning while our ideas about ourselves remain stuck in the past.

Perhaps one of the most valuable habits we can pass on to our children is to put a little distance between these two sentences:

“I don’t know how to do this yet.”

“I can’t do this.”

During Naomi Osaka’s first-round match, a group in a hospitality suite started moving around, posing for photos and creating enough distraction that the chair umpire intervened.

Elsewhere, spectators were filmed using ring lights and flash photography during play.

Yes.

Ring lights.

At the U.S. Open.

Organizers eventually reminded spectators not to use lighting that could interfere with matches. The point was simple: if you come to a sporting event, understand the etiquette of the sport.

One detail matters: the people involved in the most discussed Osaka incident were not part of the USTA’s official creator program.

That actually makes the story more interesting.

Because this isn’t really about influencers behaving badly.

It’s about what happens when one system imports the incentives of another.

The Obvious Take Is Too Easy

You can turn this into a generational argument quickly.

• Influencers have no manners.

• Old tennis fans are snobs.

• Social media ruined everything.

None of those explanations are very useful.

The better question is:

Because inside the creator economy, it makes sense.

You are somewhere interesting. Your audience expects content. Your business rewards visibility. The venue becomes part of the content. The event becomes raw material…

That doesn’t make creators irrational.

It makes them rational inside a different incentive system. And businesses underestimate this all the time.

You Don’t Add a Channel. You Add a Behavior.

The U.S. Open wants younger audiences and more social distribution. Of course it does.

Creators are good at both.

But a distribution channel never arrives alone.

It brings its habits, economics and definition of success.

Traditional media has one incentive structure. Fans have another. Sponsors have another. Creators have another.

• A creator is rewarded for capturing something.

• A tennis spectator, during a point, creates value partly by doing nothing.

That’s the real conflict.

Not old versus young. Two systems rewarding different behavior.

Channels Move Upstream

We’ve seen this pattern before.

• Google started as a traffic source; then ranking became the incentive, and SEO tactics became the standard way content was researched, structured and written.

• YouTube started as a video platform; then clicks and watch time became the incentive, and reaction videos, repeatable formats, low-cost entertainment and endless cat videos flourished.

• Facebook started as a social network; then attention and engagement became the incentive, and relationships slowly turned into content, performance and measurable social currency.

• Twitter started as a place to share updates; then replies, retweets and outrage became the incentive, and conflict consistently traveled further than nuance.

• Instagram started as photo sharing; then likes, followers and visual status became the incentive, and everyday life turned into a comparison machine for lifestyles, bodies, careers and success.

• TikTok started as a short-video platform; then instant retention became the incentive, and video adapted to shorter attention spans: faster hooks, quicker cuts, less patience and almost no room for a slow start.

The pattern is always the same: First, the channel distributes the product. Then its incentives start redesigning the product and eventually the people using it.

Usually nobody announces this.

The incentives do the work.

Growth Channels Have Opinions

This is where companies get caught.

They evaluate channels with questions like: How much traffic can this produce? What’s the CAC? How fast can it scale?

Those questions matter. But there’s another one I’d put next to them:

Imagine paid acquisition becomes 60% of your growth.

• Suddenly, customers who convert quickly look better.

• Brand investments look harder to justify.

• Longer product education looks inefficient.

• Retention competes with acquisition experiments that can move this quarter’s numbers.

Nobody says: “We’re redesigning the company around Meta Ads.”

They don’t need to.

The economics start making decisions for them.

The same thing happens with partners.

A partner channel works. Then it works really well. Soon their integration requests enter the roadmap. Their customers influence priorities.

The channel isn’t just distributing the product anymore.

It is helping specify it.

Distribution Has Gravity

A small channel adapts to your product. A large channel eventually asks your product to adapt to it.

That’s distribution gravity.

And it’s why the biggest channel is not always the best channel.

Sometimes you pay for growth with margin. Sometimes with control. Sometimes with customer ownership. Sometimes with product complexity. Sometimes with the product itself.

AI Agents Will Do the Same Thing

I think we’re about to see this again with AI agents.

Today, most products are still designed for humans. Humans search. Humans compare. Humans read landing pages. Humans click.

But if agents begin choosing products, comparing vendors and completing transactions for us, companies will adapt.

Structured product data becomes more important. APIs become more important. Machine-readable reputation matters more. Pricing may change.

Eventually, companies may start building things partly because agents prefer them.

Same pattern. New channel. More distribution. Then the channel starts rewriting the product.

The Agent-Ready Business

AI agents won’t just change how customers discover products. If they become a meaningful distribution layer, they can also change how your company structures pricing, exposes data, handles transactions, and designs the product itself.

So we built The Agent-Ready Business a practical strategic report for founders who want to understand how prepared their business is for agent-mediated discovery, comparison, and purchasing.

Inside, you’ll get:

• The 100-Point Agent Readiness Index; a scoring framework across discoverability, offer legibility, comparability, actionability, transaction readiness, trust, and post-purchase operations.

• A 100-Signal Database; real examples from the emerging agent ecosystem, including MCP, A2A, UCP, ACP, AP2, agentic checkout, machine payments, and verified agent identity.

• Six Business-Model Playbooks; specific implications for ecommerce, B2B SaaS, marketplaces, consumer apps, travel/booking, and fintech or digital services.

• The Agent Preference Matrix; a framework for separating changes that improve the customer experience from changes that merely make your business easier for agents to process.

• A 100-Point Implementation Checklist; concrete technical and operational actions divided into what to do now, what to prepare, and what to delay.

• Three Adoption Scenarios; from AI remaining mostly advisory to agents operating with bounded transactional autonomy.

• A 12–24 Month Roadmap; what to build in the next 90 days, what to test later, and what is still too early to justify.

• A Board-Level Dashboard; a one-page template for tracking agent readiness, channel exposure, dependencies, and the next investments to make.

• Strategic Red Lines; a framework for deciding what you are unwilling to let a new distribution channel change about your product, pricing, customer ownership, or brand.

You can access the full report via the link below ↓

Read more

China: $12.4 billion.

That’s a 23x difference.

Now here’s the more interesting number.

By March 2026, Stanford’s AI Index found that the performance gap between the leading U.S. and Chinese AI models had narrowed to just 2.7%. The two countries had already traded places at the top several times since early 2025.

Read those two numbers together.

23x the private investment.

2.7% the model-performance gap.

That should make a lot of executives in developed markets slightly uncomfortable.

Because there is still a deeply embedded assumption in the U.S. and Europe that goes something like this:

Historically, that was mostly true.

With AI, I’m not sure it will stay true for very long.

And if you run a company in New York, London, Berlin or Paris, assuming it will is becoming dangerous.

The Lead is Real, The Moat is Not

Let me be clear.

The United States has an enormous structural advantage in AI.

Capital. Chips. Research universities. Talent. Cloud infrastructure. Frontier labs. Venture capital. Customers willing to pay.

I’m not arguing that these advantages disappeared overnight.

I’m arguing something different.

The technology produced by those advantages is becoming globally accessible at an astonishing speed.

A breakthrough used to take years to diffuse.

Now it can take months.

Sometimes weeks.

Look at DeepSeek.

You don’t need to move to China, negotiate a technology-transfer agreement or build your own data center to use a Chinese reasoning model.

AWS made DeepSeek available as a fully managed model on Amazon Bedrock. Microsoft offers DeepSeek models through Azure AI Foundry. Google added DeepSeek to Vertex AI Model Garden.

Think about how strange that would have sounded a few years ago.

Three of America’s largest technology companies are providing infrastructure through which their customers can deploy Chinese AI models.

That’s not a criticism.

It makes perfect business sense.

But it tells you something important about where competitive advantage is moving.

The model itself is becoming a component.

And components get commoditized.

Open Models Have Changed the Speed of Catch-up

One of the most interesting AI charts I’ve seen recently isn’t about funding.

It’s about usage.

OpenRouter analyzed more than 450 trillion tokens processed on its platform between January and mid-June 2026.

At the beginning of the year, DeepSeek represented roughly 9% of token volume.

By June, it was around 18%.

More interestingly, Chinese models collectively surpassed American models in token share on OpenRouter in early June.

Now, OpenRouter is one platform. I wouldn’t confuse its traffic with global AI market share.

But directionally, it tells us something important.

Capability travels.

Fast.

And buyers increasingly have the ability to choose between models based on performance, price, latency and suitability rather than where the model was created.

This changes the economics of catching up.

A company in Istanbul, Bangalore, São Paulo or Jakarta no longer needs to reproduce hundreds of billions of dollars of AI infrastructure before it can benefit from the output of that infrastructure.

• It can rent the intelligence.

• Use an open model.

• Fine-tune it.

• Build on top of it.

• Combine several models.

• And increasingly, swap one for another when the economics change.

That last part matters more than it sounds.

Because if your competitive advantage is:

“We have access to the best AI.”

You may not have much of a competitive advantage.

Your competitor may have access to something almost as good next quarter for a fraction of the cost.

Your Company Does Not Automatically Inherit Silicon Valley’s Lead

This is where I think developed-market companies risk becoming complacent.

The adoption statistics look fantastic.

Stanford’s 2026 AI Index says 88% of surveyed organizations were using AI in at least one business function in 2025.

Sounds like the transition is well underway.

It is.

But using AI and reorganizing a company around AI are two completely different things.

McKinsey found that nearly 90% of organizations were at least experimenting with AI, while only 7% said they had scaled it across the enterprise.

That gap is the whole game.

Giving everyone ChatGPT, Claude or Copilot is not AI transformation. It’s access.

Your employees using AI to rewrite emails faster is useful. It is not a moat.

The hard part starts when you ask:

• Which workflow changes?

• Which decision changes?

• Which cost changes?

• Which customer experience changes?

• Which revenue number changes?

• And what can we now do that was economically impossible two years ago?

That is where organizations slow down.

Committees appear. Governance meetings appear. Twenty-seven AI pilots appear.

Everyone has a strategy deck. Nobody can tell you what happened to gross margin. And the latest McKinsey research makes the difference quite visible.

Among its AI “high performers”, nearly three-quarters report fundamentally redesigning workflows because of AI.

Among everyone else?

About one-quarter.

That’s not a tooling difference. It’s an operating-system difference.

The winners aren’t simply using AI more. They’re changing the machine around the AI.

Your Dangerous Competitor May Not Look Dangerous Yet

For decades, a mid-sized company in a developing economy had disadvantages almost everywhere.

Less capital. Smaller talent pools. Worse infrastructure. Less access to technology. More expensive knowledge. Smaller domestic markets.

If you were sitting inside a large American or European company, you could reasonably assume many competitors elsewhere were years behind.

AI doesn’t eliminate those disadvantages. But it attacks several of them at the same time.

• A ten-person export company can now perform market research that once required consultants.

• A small sales team can personalize communication across hundreds or thousands of accounts.

• A customer-service operation can handle dramatically more conversations without scaling headcount linearly.

• A developer who has never worked with a particular framework can become productive in it much faster.

• A local company can translate, localize, analyze, code, research and create at a level that previously required much more specialized talent.

This is why I think we make a mistake when we talk about AI only as a productivity technology.

Productivity is the obvious part.

The bigger shift is that AI makes capabilities affordable that many companies previously couldn’t afford at all.

And that disproportionately benefits the company that was constrained in the first place.

Imagine you already have 200 analysts. Making every analyst 20% more productive is valuable.

Now imagine your competitor could never afford an analyst team in the first place.

Suddenly, three people with the right AI systems can do a meaningful part of what previously required twenty.

Those two outcomes are not economically equivalent. One is optimization.

The other is capability creation. And capability creation is how gaps close much faster than incumbents expect.

The Real Moat is Becoming Learning Speed

There’s another reason waiting is expensive.

Companies often talk about AI adoption as if it were something they could postpone and then catch up on later.

• “We’ll let the technology mature”

• “We’ll see which vendors win”

• “We’ll implement once the use cases are clearer”

It sounds rational.

The problem is that your competitors aren’t only accumulating software while you wait. They’re accumulating experience.

Suppose a competitor starts redesigning its sales operation around AI today.

Over the next two years, it learns:

• Which leads should be contacted first.

• Which tasks should stay human.

• Which responses can be automated.

• Where hallucinations create risk.

• What data the system needs.

• How customers react.

• How salespeople actually use it.

• Where automation hurts conversion.

• Where it improves conversion.

• And how the workflow should change again.

After two years, you can buy the same model.

You cannot buy those two years of organizational learning.

This is the part I think many boards underestimate.

The compounding asset is not the AI. It’s the feedback loop around the AI.

The model can be copied. The API can be replaced.

The workflow knowledge accumulated through thousands of real customer interactions is much harder to reproduce.

That’s where the moat moves.

Don’t Start with an AI Strategy

This is also why I’m skeptical when someone tells me their company is “working on an AI strategy”.

AI is too horizontal for that to be a useful starting point.

I would start somewhere much less exciting: Which business number do you want to move?

Revenue? Conversion? Churn? Customer response time? Cost per ticket? Time to create a proposal? Time from lead to meeting? Customers served per employee?

Pick one.

Then find the workflow constraining that number. Then redesign that workflow using what AI makes possible today. Measure before. Change the system. Measure after.

Did the number move?

If yes, keep going. If not, kill it and attack another bottleneck.

You don’t need an 80-page transformation plan to begin. You need one economically meaningful constraint.

AI is leverage.

Leverage is only useful when you apply it to the right point.

The Comfortable Era is Ending

The West is not about to stop producing the world’s most important AI technology.

That’s not my point.

Something subtler is happening.

Producing the technology and capturing the economic advantage from the technology are becoming two different things.

We’ve seen versions of this story before.

Invention happens in one place. Manufacturing advantage emerges somewhere else. Distribution gets captured somewhere else again.

Eventually, the company or country that created the breakthrough discovers that creation alone did not guarantee value capture.

AI may compress that cycle dramatically. The Stanford numbers are a useful warning.

The U.S. can outspend China more than 23 to 1 in private AI investment while the performance gap between its leading models shrinks to a few percentage points.

Meanwhile, open models and global cloud infrastructure can put advanced capabilities into the hands of companies almost anywhere.

So if you run a company in a developed market, I wouldn’t spend too much time congratulating yourself for being geographically close to the AI revolution.

Your company does not automatically inherit Silicon Valley’s lead.

You still have to earn it.

The question is no longer: Who invented the technology?

Increasingly, it is: Who reorganized around it fastest?

Because the technology advantage you think separates you from the rest of the world is becoming cheaper, more open and more portable every month.

And if someone with fewer resources learns to use it faster than you do, don’t be surprised when they stop looking like the company that was supposed to be behind.

Deadpool. Free Guy. The Proposal…

The sarcastic Canadian guy who has somehow turned playing slightly different versions of Ryan Reynolds into a very successful Hollywood career.

Fair enough.

But while most of us were watching his movies, Reynolds was quietly building a second career.

And it may be even more interesting than the first.

He became an entrepreneur.

Then a marketer.

Then the co-founder of a creative company.

He bought pieces of companies most people had barely heard of, plugged his own attention and creativity into them, and helped turn them into unusually valuable brands.

• Aviation Gin was later included in a transaction worth up to $610 million.

• Mint Mobile became part of a $1.35 billion acquisition by T-Mobile.

But Reynolds wasn’t simply the celebrity appearing at the end of somebody else’s advertising process.

He was an owner.

He shaped the creative.

He helped write the ads.

He understood the audience.

He had his own distribution.

And, crucially, he could move before a normal company had finished scheduling the meeting about whether it should move.

The interesting part is how he got there.

Because Ryan Reynolds did not learn marketing at business school.

He learned it because Hollywood kept telling him no.

Before the Gin, There Was Deadpool

Reynolds spent roughly a decade trying to get Deadpool made.

It wasn’t an easy sell.

Deadpool was violent, foul-mouthed, self-aware, R-rated and constantly broke the fourth wall.

In other words, not exactly the safest foundation for a giant studio superhero franchise.

Reynolds kept pushing anyway.

Eventually, he and the team were allowed to create a few minutes of test footage showing what their version of the character might actually look like.

Then the footage sat there.

The movie still wasn’t getting made.

Until July 2014, when something strange happened.

The test footage appeared on the internet.

Fans went crazy.

Suddenly the studio wasn’t evaluating an abstract pitch anymore.

There was visible evidence of demand.

For more than a decade, Reynolds joked about who might have leaked the footage.

Then, in 2025, he finally admitted it.

He did.

Reynolds said he had listened to his instinct that people would be interested, even though he knew leaking it was the wrong thing to do.

Whatever you think of the ethics of the leak, it reveals something fascinating about how his brain works.

He couldn’t win the argument inside the conference room.

So he changed the environment outside the conference room.

He didn’t convince the gatekeeper.

He created enough demand that the gatekeeper had to pay attention.

That may have been Reynolds’ first great marketing lesson.

And he hadn’t even started a marketing company yet.

The Movie That Accidentally Created a Marketer

Eventually, Deadpool got made.

But compared with the giant superhero productions around it, the team had relatively few resources.

Reynolds later described the experience as having to make every dollar feel like ten.

So they replaced spectacle with character.

Then they applied the same thinking to marketing.

Reynolds took the Deadpool suit from the set and started shooting inexpensive promotional material with George Dewey, who would later become his partner at Maximum Effort. Reynolds has described this period as the origin of his move into marketing.

Instead of making only conventional trailers, they treated advertising as entertainment.

Deadpool appeared on Tinder.

The violent superhero movie was advertised as if it were a Valentine’s Day romance.

There were ridiculous billboards.

Public-service announcements.

Cheap internet videos.

Reynolds moved almost seamlessly between actor, character and advertiser.

And people voluntarily shared the marketing.

That was the breakthrough.

Marketing didn’t have to interrupt the entertainment.

Marketing could be the entertainment.

The lack of resources had forced Reynolds and Dewey to discover a different way of creating attention.

And it worked.

Deadpool became, at the time, the highest-grossing R-rated movie ever.

But Reynolds walked away with something arguably more valuable than another successful movie.

He had learned how to manufacture attention.

Then He Realized Attention Was an Asset

This is where the story becomes much more interesting than “celebrity starts a side business”.

Most celebrities already possess enormous amounts of attention.

But historically, they have monetized it in a very simple way.

• A company pays the celebrity.

• The celebrity appears in an advertisement.

• Maybe they make an Instagram post.

• The celebrity gets a cheque.

• The campaign ends.

• The company owns the product, the customer relationship and most of the long-term upside.

Reynolds gradually arrived at a more interesting model.

Don’t just rent your attention to brands.

Own the asset your attention makes more valuable.

In 2018, he bought an ownership stake in Aviation Gin.

Even the origin story feels strangely Reynolds-like.

He discovered Aviation while drinking Negronis in Vancouver. He kept ordering the same drink, realized the thing he particularly liked was the gin, and eventually found his way into the company.

Once involved, Aviation said Reynolds would play an active role in the business and oversee creative direction; Reynolds also talked about wanting to participate in marketing, sales and distribution.

Think about what had changed.

If Reynolds made a brilliant commercial for somebody else’s gin, he could get paid for a commercial.

If Reynolds made millions of people care about his gin, he could increase the value of an asset he owned.

Marketing was no longer simply a service.

It became a mechanism for equity creation.

Around the same period, Reynolds and Dewey formalized what they had learned during Deadpool into Maximum Effort.

Reynolds later described the company as something that happened almost by accident: they had developed a way of marketing Deadpool, then needed to market Aviation, and suddenly found themselves running a marketing business.

Now the machine was coming together.

Reynolds had:

• Attention.

• Creative ability.

• Distribution.

• A marketing company.

• And equity in the thing being marketed.

That combination is much more powerful than celebrity endorsement.

And it reveals something Reynolds understood unusually early.

His Real Asset Wasn’t Fame. It Was Distribution.

There is an obvious objection to this entire story.

Of course Ryan Reynolds could do this. He’s Ryan Reynolds.

Correct.

Reynolds started with an unfair advantage almost no entrepreneur has.

Millions of people already knew his face.

Entertainment media followed what he did.

A single post from his account could reach more people than many companies could reach with enormous advertising budgets.

Pretending otherwise would make the whole story useless.

But Reynolds understood something about fame that many famous people don’t.

Fame is not merely attention. Fame is distribution.

And distribution is an asset.

Most celebrities rent that distribution out.

Reynolds connected his distribution to things he owned.

Now a viral joke didn’t merely earn likes.

It could introduce millions of people to Aviation Gin.

A clever video could generate awareness for Mint Mobile.

A cultural moment could become media without buying that media.

His personality became part of the companies’ distribution infrastructure.

That is a much smarter model.

And it doesn’t only apply to celebrities.

Because although you probably don’t have Ryan Reynolds’ distribution... you can build your own version of it.

You Don’t Need 50 Million Followers

Maybe 3,000 people read your newsletter. Maybe 12,000 people follow you on LinkedIn. Maybe 40,000 people watch your YouTube channel. Maybe 600 people genuinely trust what you recommend.

It isn’t Ryan Reynolds scale.

That misses the point.

The important question is:

If the answer is yes, you own some distribution.

And once you own distribution, something fundamental changes.

• When you launch a new product, you don’t begin at zero.

• When you have an idea, there is already someone listening.

• When you learn something useful, you already know where to publish it.

• When you build your second product, the audience from your first product can come with you.

You are carrying distribution with you.

This is one of the most valuable assets a modern entrepreneur can build.

Reynolds figured it out at celebrity scale.

A generation of entrepreneurs is beginning to figure it out at individual scale.

But Reynolds added another ingredient.

Speed.

The Ad That Had a 36-Hour Expiration Date

At the end of 2019, Peloton released a Christmas commercial called The Gift That Gives Back.

The premise was simple.

A husband gives his wife a Peloton bike.

She documents a year of workouts and later shows him how much the gift changed her life.

The internet had... other interpretations.

People found the commercial awkward, dystopian, sexist, controlling or some combination of the four.

Memes exploded.

Hot takes followed.

The actress at the center of the commercial, Monica Ruiz, suddenly became known as the “Peloton Wife”.

Most brands would have watched the internet drama from the sidelines.

Reynolds saw an opportunity.

His team began trying to find Ruiz within hours.

They eventually reached her.

Then they wrote an Aviation Gin commercial.

Shot it.

Edited it.

And released it.

According to Reynolds: The whole thing took roughly 36 hours.

In the ad, Ruiz sits at a bar between two friends looking like she’s just escaped something.

Three martinis sit on the table. One friend tells her she’s safe. They toast to “new beginnings”.

Then Reynolds publishes the commercial with four words:

Exercise bike not included.

The internet loved it.

But I think most people learned the wrong lesson from the campaign.

They saw the joke. Reynolds saw the window.

Reynolds Himself Has Now Formalized the Idea

There is a nice twist to this story.

Fastvertising has moved from an eccentric Ryan Reynolds marketing trick to something people are seriously studying.

In the January–February 2026 issue of Harvard Business Review, Reynolds coauthored “Marketing at the Speed of Culture” with Harvard Business School professor Ayelet Israeli, Leonard Schlesinger and Matt Higgins.

Their argument is not simply “be funny quickly”.

Successful fastvertising requires organizational conditions that allow people to respond quickly: empowered teams, streamlined governance, relevance over unnecessary production polish, and enough human judgment to know when a cultural moment should or absolutely should not become advertising.

They also make an important point about AI: generative tools can accelerate production, but human judgment remains critical.

Which gets us to the real lesson.

Humor Wasn’t Scarce. Speed Was.

Thousands of people made funny Peloton jokes that week.

Humor was abundant. Execution speed was scarce.

The Aviation ad worked because the joke arrived while the joke still mattered.

That’s a very different way to think about cultural marketing.

Attention has a half-life.

A cultural moment does not maintain the same value forever.

At hour one, people discover it.

At hour twelve, people share it.

At hour thirty-six, everybody has an opinion.

A week later, your beautifully produced brand response can already feel like someone explaining an old meme.

Modern marketers obsess over: CAC. CPM. CTR. ROAS. Conversion rate.

I think they should start measuring another number.

TTP: Time to Publish.

How much time passes between: “We should do something with this” and published.

That number rarely appears on a marketing dashboard. It should.

Because in a world where attention moves quickly: Latency is a cost.

Founder-Led Marketing’s Hidden Advantage

Founder-led marketing is usually explained with one word: Authenticity.

People trust people more than corporations.

A founder has a real voice. Customers like seeing the person behind the product. All true.

But I think authenticity is only half of the advantage.

The other half is much more mechanical. Decision compression.

Think about the traditional process.

• Someone notices an opportunity.

• Someone writes a brief.

• A manager reviews it.

• Brand reviews it.

• An agency gets involved.

• The agency produces concepts.

• Marketing chooses one.

• Legal reviews it.

• Management asks for a safer version.

• Someone schedules another meeting.

• Then eventually somebody publishes it.

Every handoff adds latency.

Now compare that with a founder who possesses taste, authority and proximity to the customer.

Idea. Decision. Publish.

Sometimes the entire loop occurs inside one person’s head. That is extraordinarily powerful.

Not because founders are automatically smarter. Many aren’t.

But when ownership + taste + authority sit close together, companies can move at a completely different speed.

Reynolds wasn’t simply the face of Aviation.

He was part of a low-latency decision system.

Then He Ran the Experiment Again

Aviation could be dismissed as a special case.

Celebrity. Alcohol. Funny ads.

Fine.

Then Reynolds bought an ownership stake in Mint Mobile in November 2019.

Discount wireless plans are considerably less sexy than gin.

Which makes Mint a much more interesting experiment.

The playbook remained remarkably similar.

Reynolds appeared directly in the advertising.

The production was often deliberately inexpensive.

The writing was self-aware.

The company made jokes about being a company trying to sell you something.

One Mint campaign proudly centered on a commercial that cost just $500.

Another had ChatGPT write the commercial. But perhaps the purest demonstration of the system happened in 2022.

Comedian Dave Foley tweeted Reynolds asking whether, if he switched to Mint Mobile, he could appear in one of the company’s ads.

Reynolds accepted.

George Dewey called Mint while heading to the set.

The concept was aligned in minutes.

The ad was written on the move.

Shot. Edited. Published.

From tweet to finished commercial: six hours.

Six hours.

There are companies that need six hours to decide who should attend the kickoff meeting.

The Acquisitions Matter. But Not for the Reason You Think

Eventually, the numbers became enormous.

In 2020, Diageo agreed to acquire Aviation Gin and Davos Brands in a transaction worth up to $610 million: $335 million upfront plus as much as $275 million tied to future performance.

Important detail: That number did not represent Aviation Gin alone.

The acquisition also included other brands in the Davos Brands portfolio. Reynolds retained an ongoing interest in Aviation.

Likewise, T-Mobile announced in 2023 that it would pay up to $1.35 billion for Ka’ena Corporation.

Again, important detail: That was not Mint Mobile alone.

Ka’ena also included Ultra Mobile and Plum.

The transaction closed on May 1, 2024.

So the lazy version of this story would be:

Ryan Reynolds made funny commercials and sold two companies for nearly $2 billion.

That’s a nice LinkedIn fairy tale.

It isn’t accurate. Product mattered. Pricing mattered. Distribution mattered. Teams mattered. Market timing mattered.

But the marketing capability clearly mattered too.

When T-Mobile completed the Ka’ena acquisition, it specifically pointed to Mint’s unusually strong direct-to-consumer marketing capabilities as one of the things that had fueled its success.

That’s the more interesting story.

Reynolds repeatedly connected: ownership + attention + creative control + distribution + speed.

And that combination created an unusually efficient growth machine.

Your Approval Process Is Part of Your Media Budget

Now imagine two brands notice exactly the same cultural opportunity.

Brand A can publish tonight.

Brand B needs nine days. Brand B may have: The better agency. Better designers. More research. A larger budget. A 140-page brand guideline.

It doesn’t matter.

Brand A owns the moment.

Which leads to an uncomfortable idea: Your approval process is effectively part of your customer acquisition cost.

We just don’t account for it that way. We measure campaigns that run. We measure campaigns that fail.

We do not measure brilliant campaigns that never existed because Legal could only join the meeting next Tuesday.

Nobody’s dashboard says:

Cultural moments lost because Steve was on vacation: 17.

But the cost is real.

Someone else captured the attention. You paid for safety with distribution.

There Is a Point Where Professionalism Becomes Latency

Companies add processes for sensible reasons; Consistency. Quality. Compliance. Legal protection. Risk management. Brand safety.

Initially, the processes help.

Then another stakeholder joins. Another approval. Another recurring meeting. Another document.

Eventually something strange happens.

Everyone contributes to the decision. Nobody can make the decision.

The work becomes more professional. The organization becomes less capable of acting. And the incentives push companies naturally in this direction.

A visible mistake has an owner. A missed opportunity usually doesn’t.

Nobody gets fired because an advertisement that never existed failed to go viral. So companies accumulate defensive processes.

Every individual layer feels rational. Together they create massive latency. And research suggests that slow does not necessarily mean better.

McKinsey found that only 48% of surveyed managers said their organizations consistently made decisions quickly, while just 37% said decisions were both high-quality and fast.

More surprisingly, organizations that made decisions quickly were around twice as likely to report high-quality decisions as slow decision-makers.

A much broader meta-analysis published online in 2024 examined 127 studies and 239 effect sizes and found that decision speed, implementation speed and competitive-response speed were all positively related to firm performance overall.

The lesson isn’t: Make every decision quickly.

The lesson is: Stop making reversible decisions travel through systems designed for irreversible ones.

A billion-dollar acquisition deserves friction.

A topical Instagram video probably doesn’t need six executive approvals.

AI Just Moved the Bottleneck

This is where the Reynolds story becomes much more important in 2026.

For decades, production itself was expensive.

Research took time. Writing took time. Design took time. Editing took time. Video took a lot of time.

Companies built enormous organizations around those constraints.

Then generative AI arrived.

In a study involving 758 Boston Consulting Group consultants, people using GPT-4 on tasks within the model’s capabilities completed work more than 25% faster, completed over 12% more tasks, and produced work rated more than 40% higher in quality.

Another study covering more than 5,000 customer-support agents found an average productivity improvement of roughly 14% after workers received generative-AI assistance.

And those are relatively early examples.

Today, a tiny team can research markets, write copy, produce images, generate voice, edit video, create landing pages and iterate on dozens of concepts at speeds that would have sounded absurd a few years ago.

Fantastic.

Now imagine this workflow:

• Research: 5 minutes

• Concepts: 10 minutes

• Copy: 30 seconds

• Images: 3 minutes

• Video: 20 minutes

• Management approval: 11 days

What exactly did we accelerate?

Nothing meaningful.

We optimized the part of the system that was no longer the bottleneck.

The bottleneck moved. From production to permission.

Companies are going to spend millions making employees five times faster with AI... and then feed their output into the same organizational architecture designed when producing a campaign took three months.

It’s like installing a Formula 1 engine and driving it through Istanbul traffic at 6 p.m.

Technically faster. Practically stationary.

Reynolds Now Has a Name for It: Fastvertising

There is a nice closing loop to this story.

What began with a frustrated actor trying to get Deadpool made eventually became an actual management concept.

In the January–February 2026 issue of Harvard Business Review, Ryan Reynolds co-authored an article with Harvard Business School professors Ayelet Israeli and Leonard Schlesinger and investor Matt Higgins.

The title: “Marketing at the Speed of Culture”

The concept they describe is fastvertising: rapid-response advertising designed to participate in cultural conversations before those conversations move on.

In other words, Reynolds’ accidental education became a system.

• Deadpool taught him constraints.

• Constraints taught him creativity.

• Creativity taught him how to create attention.

• Attention became distribution.

• Distribution became more valuable when attached to ownership.

• Ownership compressed decision-making.

• Compressed decisions enabled speed.

• And speed became an advantage.

Which brings us to the part of this story I think entrepreneurs should pay much more attention to.

The Ryan Reynolds Model Is Coming for the One-Person Company

Strip away Hollywood. Strip away the celebrity. Strip away 50 million followers.

What remains?

A surprisingly simple machine:

• Build attention.

• Own the relationship with the audience.

• Own the thing you sell.

• Create the marketing yourself.

• Distribute directly.

• Move quickly.

Twenty years ago, assembling that machine as one person was extremely difficult.

You needed a designer. A copywriter. A developer. A video editor. A media buyer. Maybe an agency.

And distribution usually had to be purchased.

Today?

One person can build an audience on LinkedIn. Publish a newsletter. Run a YouTube channel. Build software. Generate creative. Produce video. Write a landing page. Analyze customer feedback. Automate support. Sell globally.

And increasingly, AI can sit behind almost every one of those activities.

This is why Reynolds is such an unexpectedly useful model for solo founders.

Obviously you don’t have his audience. You don’t need his audience.

You need the smallest viable version of his machine.

Maybe your Maximum Effort is you plus five AI tools.

Maybe your Hollywood audience is 10,000 newsletter subscribers in an extremely specific industry.

Maybe your Aviation Gin is a $29 SaaS product.

The scale is different. The architecture isn’t.

And this may become one of the biggest entrepreneurial shifts created by AI.

For decades, a company had an enormous structural advantage over an individual.

A company could coordinate many specialized people. AI begins to compress those people into capabilities.

Which means something strange happens.

The individual starts acquiring some of the advantages of the organization... without necessarily acquiring all of its bureaucracy.

That is a very powerful combination.

The One-Person Company Has One Advantage Big Companies Can’t Easily Buy

People usually think small companies lose because they lack resources. Sometimes they do.

But smallness contains an advantage too. Distance. Or rather, the lack of it.

• Distance between customer and founder.

• Distance between insight and decision.

• Distance between decision and execution.

• Distance between execution and distribution.

A giant company can buy the best AI models. It can hire the best consultants. It can spend $20 million on transformation.

What it cannot instantly buy is an organizational architecture with zero friction.

A solo founder can hear something from a customer at 9:00 a.m., change the product at 10:00, record a video at 11:00 and tell the entire audience about it before lunch.

No steering committee. No alignment meeting. No deck explaining why the deck should exist.

That speed used to be limited by what one human could physically produce.

AI is removing that limitation.

Which makes smallness more interesting than it used to be.

Don’t Copy Ryan Reynolds’ Jokes

The wrong takeaway from all of this is: We need funnier marketing.

Maybe you do. But that’s the surface.

Don’t copy Reynolds’ jokes. Copy the architecture that allowed the jokes to exist.

Ask:

• How many people must approve a reversible marketing decision?

• Who actually owns the voice of the company?

• Can the person closest to an opportunity act on it?

• Do the people creating your marketing have direct access to customer reality?

• Which decisions genuinely carry major brand risk?

• Which decisions are slow simply because they inherited a process?

• How long does it take your company to go from: “That’s interesting.” to: Published.

And for founders, ask an even more important set of questions:

• Am I building an audience or merely buying traffic?

• Do I own the relationship with that audience?

• Do I own the products my distribution makes valuable?

• Can I move from insight to execution without waiting for somebody else?

Because those four questions describe a very different kind of business.

The Scarce Resource Is Moving

For years, companies competed on their ability to produce.

Who had the best agency? The biggest creative department? The largest media budget? The best production capabilities?

AI is making production abundant. Soon almost everybody will be able to generate: 100 ideas. 100 pieces of copy. 100 images. 100 videos. 100 landing pages.

So what becomes scarce?

• Taste. Knowing which of the 100 ideas matters.

• Judgment. Knowing which moment deserves a response.

• Trust. Having an audience that actually wants to hear from you.

• Distribution. Being able to reach that audience without renting access every time.

• And perhaps most surprisingly: Permission.

The ability to actually do the thing while it still matters.

That is why the Ryan Reynolds story isn’t really about Ryan Reynolds.

It’s about a larger shift in how companies can be built.

Reynolds happened to start with the ultimate unfair advantage: fame.

But what made that fame economically interesting was that he figured out the rest of the system.

He turned attention into distribution. Distribution into ownership. Ownership into speed. And speed into an organizational advantage.

AI is now giving ordinary entrepreneurs tools that make a smaller version of that machine possible.

So don’t try to become Ryan Reynolds. Build your own version of his operating system.

Own an audience. Own what you sell. Own your creative process. And shorten the distance between an idea and the world.

Because your next competitive advantage may not be another AI tool.

It may not be another agency. It may not be another $500,000 campaign. It may simply be this: fewer people between the idea and Publish.

Sounds ridiculous today, doesn’t it? Imagine calling someone for help with a video game. Just Google it. Ask a friend. Start the game over…

But there’s something in that phone call that billions of dollars of AI investment still haven’t quite managed to recreate:

The moment the customer needs help, there’s someone on the other end who actually understands them.

The Nintendo employee doesn’t read from a script.

• “Which quest are you on?”

• “How many heart containers do you have?”

• “Where exactly are you stuck?”

They understand the context. Diagnose the problem. Give the right answer.

What’s remarkable is that the kind of one-to-one attention a gamer could get 40 years ago is still considered a luxury at many large companies today.

We’re Using Technology to Solve the Wrong Problem

Whenever companies encounter a new technology, they tend to jump to the same question:

“How much money can this save us?”

Fewer people in the call center. Fewer tickets. Shorter calls. Higher automation rates.

Then the technology gets placed between the company and the customer.

IVR. Forms. Chatbots…

And now, AI.

Too often, the goal isn’t to serve the customer better. It’s to delay the moment they reach a real person for as long as possible.

I think that’s a fundamental mistake in how we think about technology.

Technology shouldn’t be used to reduce human connection.

It should be used to multiply it.

The right question isn’t:

“How can we do this with fewer people?”

It’s:

The first is a cost-cutting strategy.

The second is a competitive strategy.

Great Service No Longer Has to Be a Scaling Problem

Until now, truly personalized service has been expensive.

Your most valuable customers get a dedicated relationship manager. A private banking customer calls and someone already knows who they are. A VIP customer doesn’t wait in line. They don’t have to explain their history again. They speak to someone who already understands what they need.

Giving that experience to millions of customers simply wasn’t economically realistic.

So companies rationed attention.

At the top: one-to-one service.

At the bottom: a queue.

One of the biggest things AI changes, in my opinion, is exactly this:

The economics of one-to-one service.

Built properly, an AI system doesn’t just answer questions.

• It knows who the customer is.

• It knows what they’ve done before.

• It knows what they bought.

• It knows where they are in the journey.

• It knows what problems they’ve had in the past.

And it builds the current conversation on top of all that context.

So instead of giving the customer a bot that starts from zero every single time, you give them an assistant that actually knows them.

And you can do that for thousands, potentially millions of people at the same time.

You no longer need to hire thousands of additional employees to make that possible.

But you do need to put AI in the right place.

Customer Service Is Actually Part of Sales

On an org chart, sales, customer service, retention, and cross-sell all live in different boxes.

Customers don’t think in boxes.

They’re just trying to do something.

• “Is this product right for me?”

• “Which one should I choose?”

• “How do I set this up?”

• “Why didn’t this work?”

• “What am I supposed to do now?”

• “Is there a better option for me?”

The company may label some of these questions “sales”.

Others “customer support”.

Others “customer success”.

From the customer’s perspective, they’re all the same question:

Are you going to help me?

And every moment in which you help them has a commercial consequence.

• Help them make a decision and conversion goes up.

• Help them get more value from the product and retention improves.

• Solve their problem quickly and churn goes down.

• Understand what they actually need and the right upsell opportunity appears naturally.

That’s why treating customer service purely as a cost center feels increasingly strange to me.

What you actually have is an always-on growth channel.

The Biggest Opportunity Isn’t in Big Problems. It’s in Small Moments.

We usually think about customer experience when something goes wrong.

A complaint comes in. A ticket gets opened. Someone calls the support line.

But a lot of customer loss doesn’t happen during dramatic moments.

It happens through tiny moments of friction.

They don’t understand something. They can’t make a choice. They don’t want to spend another two minutes figuring it out. They can’t find an answer. They open another tab…

And then they never come back.

In many of these moments, the company doesn’t even know the customer needed help.

This is where conversational AI that can speak naturally across the phone, website, or mobile app becomes especially powerful.

It doesn’t have to be a system that simply responds when something goes wrong.

It can become a layer that stays with the customer throughout their journey.

It can step in at exactly the moment help is needed.

And it can make that help deeply personal.

This Is the Biggest Difference We See in Enterprise Projects

At Next Big App, our work with large organizations in Türkiye keeps revealing the same divide.

When AI is treated as little more than a smarter FAQ bot, the outcome is limited.

Call volume drops a little. Costs come down a little. But the customer experience doesn’t fundamentally change.

Once you bring the customer’s history, transaction context, current stage, and intent into the conversation, something very different happens.

The system stops merely answering questions.

It starts helping. It makes the customer feel looked after.

Let me give you a real example.

At one large company we work with, customers who need support fill out a form on the website. It could be a complaint, a suggestion, a request, anything.

What happens next looks roughly like this:

• The form lands in an internal system and waits for an employee to categorize it. Manually.

• Then it gets assigned first to a department and then to an individual employee. Again, manually.

• That employee has to find time in the middle of an already busy day to get back to the customer. Usually by phone.

• They listen to the customer, understand the issue, and genuinely try to help.

• After the call, they take whatever internal actions are necessary: call another employee, update a system, fix whatever caused the problem, and so on.

• Then they call the customer again with an update. If the customer is happy, great. If not, the whole cycle begins again.

• Finally, the employee has to document everything. They sit down and write a long email summarizing the entire case from beginning to end.

You can see what’s happening here.

This is the kind of process only a company that genuinely cares about its customers would be willing to take on.

The amount of effort being spent to make one customer happy is genuinely worth applauding.

So what did we do for them?

We built a system that uses AI to understand the request, the context, and the relevant information from a much larger pool of company data.

It can automatically call the customer, listen to the entire issue, resolve it immediately when possible, or generate a report and route the case to the right teams automatically.

The goal isn’t just efficiency.

It’s to improve the customer’s experience while also making much better use of employees’ time.

At the beginning, the concern was predictable:

“What if the AI makes a mistake while talking to a customer on the phone?”

Over time, that changed into something closer to:

“The AI is actually more patient, more knowledgeable, and better at finding solutions than I am when I’m trying to squeeze these calls into the middle of a hectic day.”

Now both customers and employees are happier.

That, in a nutshell, is how we think about conversational AI.

Not as “replace the human with a bot.”

Almost the opposite.

We see it as a way to take the kind of high-quality attention you could previously afford to give only a small number of customers — and make it available to everyone.

So that every customer, at any moment, can feel like there’s someone beside them who already knows who they are.

To me, that’s the real promise of AI in customer experience.

Same Technology. Two Completely Different Outcomes.

You can have the same model. The same integrations. The same data…

But the objective you give the system changes everything.

If your goal is:

“Let’s cut call center costs by 20%.”

You’ll probably end up putting a slightly smarter barrier between the customer and your company.

If your goal is:

“Let’s help every customer as well as the best employee in our company would.”

You’ll design a completely different experience.

In one, technology gets between you and the customer.

In the other, it brings you closer.

Over the next few years, I don’t think competitive advantage will come from answering the question:

“Which companies are using AI?”

Everyone will be using it.

The real divide will be somewhere else:

Between companies that use AI to replace human connection, and companies that use AI to scale it.

That Nintendo player was only trying to get the White Sword.

But what they actually received was something more valuable:

The feeling that, at the exact moment they needed help, someone was there who genuinely wanted to help them.

Today, we have the technology to give that feeling to millions of people at once.

Using it only to cut costs would mean choosing the smallest part of the opportunity.

Then something strange happened.

People started watching television shows made of 45-second to two-minute episodes.

Not clips.

Not trailers.

Actual serialized dramas.

They’re called microdramas, and the numbers are becoming difficult to dismiss.

The U.S. audience more than doubled last year, from 26 million to 66 million monthly active users. The category is expected to generate around $1.5 billion in U.S. revenue this year and nearly $2 billion next year.

And here is the number I find much more interesting.

ReelShort, one of the largest microdrama apps, recently reached roughly 38 minutes of daily mobile viewing per user.

That’s more daily mobile viewing time than Netflix, Amazon Prime Video or Disney+ on mobile.

Think about the contradiction.

The episodes are two minutes long.

The viewing time, in aggregate, can be nearly 40.

Maybe our attention spans aren’t the problem. Maybe our calendars are.

We Keep Blaming Attention

“People have no attention span anymore” has become one of those explanations that feels true because we repeat it so often.

Nobody reads. Nobody watches long videos. Nobody finishes anything. Everything needs to be shorter.

But microdramas suggest something more interesting.

People apparently have enough attention to spend 38 minutes watching melodramatic stories about billionaires, forbidden romances and improbable family secrets.

What they may not have is a clean, predictable 38-minute block.

That is a very different problem.

A 45-minute Netflix episode asks you to make a commitment before you start.

A two-minute microdrama asks almost nothing.

You can watch one while waiting for coffee. Three in an Uber. Two before the meeting starts. Seven while your kid is finishing practice. Another four in bed.

By the end of the day, you may have consumed half an hour of story without ever deciding to “watch TV”.

That distinction matters.

The scarce resource may no longer be attention.

It may be uninterrupted attention.

Time is Becoming Granular

For most of the industrial era, products were designed around relatively large blocks of time.

• A newspaper expected a morning ritual.

• Television expected a 30- or 60-minute slot.

• A movie expected two hours.

• A bank expected you to visit a branch.

• Enterprise software often still assumes you will sit down, open the application, understand where you are and complete a workflow.

The internet removed some of those constraints.

Mobile removed even more.

But many products kept the old assumption hidden inside the new interface:

That assumption is getting expensive.

Microdramas reverse it.

They don’t ask the user to create a viewing session. They fit themselves into the cracks that already exist. That is why calling them “short-form Netflix” misses the point.

The important innovation isn’t the duration.

It’s the minimum viable unit of time required to receive value.

That unit has collapsed.

Gaming Figured This Out Years Ago

Microdramas are not the first product category built around fragmented time.

Hyper-casual mobile games made essentially the same bet years earlier.

The best of them require almost no setup, almost no learning curve and almost no commitment.

Open the app. Tap. Swipe. Avoid the obstacle. Try again…

AppsFlyer describes hyper-casual games as simple, mechanic-driven products designed for broad accessibility, with some puzzles intended to be completed in a minute or less. The category became a major force in mobile downloads precisely because the commitment required to start playing was so low.

The business model evolved. Hyper-casual increasingly blended into hybrid-casual as studios looked for stronger retention and monetization.

But the behavioral insight underneath it was already there. You didn’t need to convince someone: “I am now going to play a video game.”

You only needed a spare 45 seconds.

Standing in line. Waiting for someone. Sitting in the back of a taxi. Pretending to listen on a conference call. We’ve all been there.

The game could enter time that previously wasn’t considered usable gaming time.

That is a bigger innovation than making games shorter.

It created inventory out of fragments.

Microdramas are now doing something remarkably similar to television.

And I suspect many other categories will follow.

This Changes How I Think About Product Design

There is a metric I think more companies should care about: Time to useful completion.

Not time spent. Not session length. Not even time to value in the conventional onboarding sense.

I mean:

Five years ago, maybe your customer could reliably give you 20 minutes.

Today, perhaps they can still give you 20 minutes in total.

But you get it as: 4 minutes + 90 seconds + 6 minutes + 3 minutes + 5 minutes.

Mathematically similar. Operationally completely different.

A product designed for the first world can fail badly in the second.

This is especially important because companies frequently interpret the failure incorrectly.

• Users abandon a workflow, so we assume the interface is confusing.

• They don’t complete onboarding, so we simplify the copy.

• They stop using a dashboard, so we add notifications.

• They don’t watch the training, so we make the video more entertaining.

Sometimes those are the right fixes.

But sometimes the product is simply asking for a unit of time the customer no longer has.

You are optimizing the experience inside the session.

The real problem is the requirement to have the session at all.

The Microdrama People Understood This

A traditional television episode needs setup.

Characters enter. Context develops. Tension builds.

A microdrama doesn’t have that luxury.

You arrive late to the party and someone is already sleeping with the billionaire.

Reuters describes these series as fast-paced and built around immediate hooks and cliffhangers; some run for 50 to 75 tiny episodes.

It sounds ridiculous until you realize how carefully the product matches the environment.

• Every episode creates a small unit of completion.

• Every cliffhanger creates a restart point.

• Every restart requires almost zero context reconstruction.

The content is not simply shorter. The architecture is interruption-tolerant.

That is the useful idea. A lot of software isn’t.

Stop halfway through a complex enterprise workflow and come back tomorrow.

Where were you? What had you already checked? Why did you make that selection? Which tab had the information you needed?

Before you can continue the work, your brain has to reload the save file.

You reconstruct the state. Reopen the tabs. Remember the decisions. Rebuild the mental model.

And the more complex the workflow, the heavier that reload becomes.

Some products don’t lose users because the task is difficult. They lose them because resuming the task is difficult.

That distinction is easy to miss.

AI Makes This More Interesting, Not Less

The obvious AI product strategy right now is to make existing tasks faster.

Take a 20-minute workflow and make it 12. Then eight. Then five…

Useful.

But there is a larger opportunity.

What if the workflow no longer needed to be a session?

Instead of helping me operate the software faster, the product could hold the context while I disappear.

• I do one small thing.

• The system continues.

• It asks me for a decision only when necessary.

• I answer from my phone in 20 seconds.

• It continues again.

• I return later without reconstructing everything that happened.

That is not just automation. That is a different relationship with time.

The software absorbs fragmentation so the human doesn’t have to.

This is where AI becomes particularly powerful. Not because it can generate another paragraph or summarize another document. Because it can maintain context across gaps that humans find expensive.

The product remembers. The workflow persists. The user comes and goes.

This is Also Why Some “Engagement” Metrics are Becoming Dangerous

If you run a SaaS company, you may instinctively want customers spending more time inside your product.

More sessions. Longer sessions. More screens visited.

For some products, that makes sense.

For others, it may increasingly be evidence that you are consuming a scarce customer resource.

Imagine two products that create the same business outcome.

• Product A needs a focused 40-minute session every Tuesday.

• Product B needs four two-minute decisions spread across the week.

Which one is easier to keep using when the company gets busy?

Which one survives a chaotic month?

Which one does the executive actually adopt?

Product B might show dramatically less “engagement”. And have dramatically better retention.

That is the kind of metric inversion worth paying attention to.

Locally, less time in product looks worse. Systemically, requiring less uninterrupted time may make the product far harder to remove.

The Question I Would Ask

Look at one important workflow in your business.

Don’t ask:

How do we make this faster?

Ask:

How much uninterrupted attention does this workflow require from the customer?

Then make the question more uncomfortable:

What would we build if we could only ask for 90 seconds at a time?

You may end up with a shorter workflow.

But you may also end up with a completely different product.

Different notifications. Different state management. Different defaults. Different automation. Different pricing. Different definition of engagement.

That is what new constraints do.

They don’t merely change features. They change architecture.

Hyper-casual games looked trivial.

Microdramas look like disposable soap operas shot vertically for a phone.

It would be easy to dismiss both as evidence that people can no longer concentrate.

I think that gets the lesson almost exactly backwards.

People still have time. They still have attention.

They can apparently spend 38 minutes a day finding out whether the billionaire finally marries the woman he pretended not to love.

What they increasingly lack is the ability or willingness to hand one product a large, uninterrupted piece of their day.

The next generation of products won’t only save people time. They’ll stop requiring time to arrive in one piece.

They had been waiting more than two years for it.

The concert is scheduled to end in 2640.

Yes, 2640.

In Halberstadt, Germany, an organ has been performing John Cage’s ORGAN²/ASLSP — As Slow as Possible since September 5, 2001.

The full performance is designed to last 639 years. On August 5, 2026, organizers added one pipe to the organ, changing the sound from seven notes to eight.

After almost 25 years, the concert is roughly 4% complete.

Even better: the performance began with silence.

For the first 17 months, the piece produced no audible chord at all. That wasn’t a technical problem. That was the performance.

Imagine presenting those results at your next quarterly review.

“We’ve been working on this for 17 months.”

“And what have we produced?”

“Nothing you can hear.”

Good luck with the next job search…

The Problem isn’t Always the Metric

Business culture has become very good at measurement.

Weekly active users. CAC. Conversion. Revenue per employee. Churn. Sales velocity. Tickets closed. Features shipped.

Usually, when the numbers disappoint us, we question the work.

Sometimes we should question the clock.

Because every system has its own natural timescale.

• A paid acquisition campaign can tell you something within days.

• A pricing change might take weeks.

• A sales hire can take months before reaching full productivity.

• A new positioning strategy might look worse before it looks better because the old message was optimized for the wrong customers.

• A retention improvement can appear almost trivial in the first month and enormous several years later.

The metric can be perfectly accurate while the conclusion is completely wrong.

You’re just looking too early.

Management Has a Sampling-Rate Problem

I’ve seen this pattern many times.

A company makes a change designed to improve the system over six months.

Three weeks later, somebody opens a dashboard.

Nothing dramatic has happened.

So we change something else.

Then something else.

By month three, we’re no longer measuring the original decision. We’re measuring the accumulated noise created by our impatience.

We call this being data-driven.

Sometimes it is just highly instrumented anxiety.

The faster we can measure things, the stronger the temptation to intervene.

That’s an important distinction.

The frequency at which you can observe a system is not necessarily the frequency at which you should manage it.

Imagine checking the temperature of an oven every second and changing the dial every time it moves by half a degree.

More data wouldn’t make you a better cook.

It would prevent the oven from stabilizing.

Companies do the equivalent constantly.

AI is Going to Make This Worse

AI is dramatically shortening the time between idea and action.

You can rewrite the onboarding today. Generate 20 ad variants this afternoon. Change sales messaging tomorrow morning. Build a feature prototype before lunch. Analyze customer feedback in minutes…

This is mostly good.

But faster execution creates a new temptation: faster judgment.

Those are not the same thing.

If an experiment that once took six weeks to launch can now be launched in two days, it doesn’t necessarily mean the customer behavior you’re trying to understand suddenly reveals itself 21 times faster.

The production cycle got shorter.

The feedback loop may not have.

That mismatch matters.

When execution becomes cheap, organizations can easily become addicted to changing things.

And constant change can destroy the very signal they’re trying to observe.

Some Progress is Invisible at First

The John Cage concert is useful precisely because it takes our normal intuition about progress and breaks it.

For 17 months, silence was progress.

For years at a time, the same chord can be progress.

On August 5, adding one pipe was progress.

The next chord change is scheduled for October 5, 2027.

Viewed hourly, almost nothing happens.

Viewed across decades, the piece is moving exactly as intended.

The system hasn’t changed.

Only your observation window has.

Businesses have versions of this everywhere.

A customer-success process may create very little visible revenue this quarter while quietly changing next year’s renewal curve.

Hiring someone excellent can initially reduce team output because good people ask uncomfortable questions before improving the system.

Removing low-quality customers can make revenue growth look weaker while making the company healthier.

Refactoring infrastructure can produce weeks of apparently zero customer value while removing a constraint that would have become expensive at 10x scale.

This isn’t an argument for patience as a virtue.

“Just give it more time” is one of the oldest excuses for bad strategy.

It’s an argument for something more precise:

Decide the clock before you look at the result.

Match the Measurement Window to the Mechanism

Before starting an initiative, there are three different questions worth separating.

• What should move?

• How much should it move?

• And when should we reasonably expect to see it?

Companies are usually reasonably good at the first two.

The third one is strangely neglected.

We define a success metric without defining its natural feedback period.

Then management cadence fills the vacuum.

• If leadership meets weekly, everything starts feeling like a weekly problem.

• If investors care about quarters, every investment starts getting evaluated quarterly.

• If the dashboard refreshes every morning, yesterday suddenly feels meaningful.

The clock silently becomes part of the strategy.

And that’s dangerous.

Because the wrong clock can make a smart decision look stupid.

It can also make a bad decision look good.

Some problems take years to appear.

A brand can be weakened slowly while short-term conversion stays healthy.

Technical debt can accumulate while shipping velocity initially improves.

A team can burn out while quarterly targets continue being hit.

Just as measuring too early creates false negatives, measuring too narrowly can create false positives.

Time changes what you can see.

The Question I’d Ask

When something in your company appears to be underperforming, don’t immediately ask:

“Should we change it?”

Ask:

Not how often the dashboard refreshes. Not when the next board meeting happens. Not when the quarter ends.

How long does cause need to become effect?

Sometimes the answer will be three days.

Sometimes three months.

Occasionally three years.

And apparently, sometimes 639.

The people running the John Cage concert have one enormous advantage over the rest of us: Nobody can pretend the project should be judged by next quarter.

In business, the mismatch is harder to see.

Which is why we often kill good systems not because they failed, but because our clock ran out before their feedback loop did.

Others become part of internet culture.

Nearly a year after American Eagle’s Sydney Sweeney campaign first launched, people are still debating it. Not because the ad was technically brilliant. Not because everyone loved it. Quite the opposite.

It started with a simple wordplay.

“Sydney Sweeney has great jeans.”

The ad played on the words genes and jeans, with Sweeney saying, “My genes are blue.”

Within hours, that joke became something much bigger.

Some critics argued it echoed ideas about eugenics, racial superiority, and outdated beauty standards.

Others dismissed the backlash entirely, insisting it was just a clever denim commercial and another example of internet outrage.

Politicians joined the debate.

Marketing experts turned it into a case study.

Millions of people who had no intention of buying jeans suddenly had an opinion about American Eagle.

The company’s stock even climbed as the conversation spread.

The internet kept the campaign alive long after American Eagle stopped paying to promote it.

Whether people loved it or hated it became almost irrelevant.

The conversation itself became the product.

This summer, American Eagle brought Sydney Sweeney back for another campaign.

That decision is more interesting than the campaign itself.

Companies rarely repeat creative.

They repeat outcomes.

American Eagle wasn’t betting on a celebrity.

It was betting on something far more valuable:

A conversation that refused to die.

Most people are asking the wrong question.

“Was it a good campaign?”

The better question is:

That distinction changes almost everything.

The Internet Changed the Rules

For decades, marketing was about broadcasting messages.

• You bought attention.

• You delivered a story.

• You hoped people remembered it.

The internet no longer works that way.

Today, the most valuable campaigns aren’t consumed.

They’re extended.

Every quote tweet, reaction video, podcast discussion, LinkedIn post, and group chat becomes free distribution.

The audience doesn’t just receive the message.

They become part of the media plan.

The internet has quietly changed the unit of marketing.

It used to be impressions.

Now it’s participation.

You’ve probably noticed this yourself.

The most thoughtful thing you’ve ever written on X or LinkedIn often disappears without a trace.

You spend hours putting together original insights, practical frameworks, or hard-earned lessons.

A few likes.

A handful of comments.

Then it’s gone.

But post something that divides people, even slightly and suddenly the algorithm comes alive.

The comments don’t stop.

People argue.

Others jump in to defend one side or the other.

The post keeps resurfacing for days.

That’s because social platforms don’t optimize for agreement. They optimize for interaction.

And nothing generates interaction more reliably than a conversation people feel compelled to join.

Distribution is no longer something companies own. It’s something audiences decide.

AI Is Changing the Economics

This is where AI enters the picture.

Most people think AI’s biggest impact is making it cheaper to build products.

That’s true.

But it’s only half the story.

The bigger shift is what happens after creation becomes cheap.

• When everyone can build faster...

• When everyone can generate content...

• When everyone can launch products...

• Creation stops being the bottleneck.

• Distribution becomes the bottleneck.

AI isn’t just commoditizing software.

It’s commoditizing creation itself.

And whenever something becomes abundant, something else becomes scarce.

In this case, what’s becoming scarce is attention.

Not purchased attention. Earned attention.

The kind people voluntarily pass to someone else.

The New Competitive Advantage

The companies that win over the next decade won’t necessarily build the best products.

They’ll build the products that generate the strongest self-reinforcing conversations.

Because every conversation creates more awareness.

• More awareness creates more users.

• More users create more stories.

• More stories create more conversations.

That’s a flywheel.

And flywheels are much harder to compete with than features.

Look at the companies that have defined the last decade.

Figma didn’t become valuable because it made design files.

It became valuable because it changed how teams think together.

People shared workflows.

Templates.

Ideas.

Not because Figma asked them to.

Because using Figma naturally created conversations.

Notion followed the same pattern.

People didn’t recommend it because it stored documents better.

They recommended it because it became part of their identity.

Their dashboards.

Their second brains.

Their productivity systems.

The product became content.

The content became distribution.

Distribution became growth.

ChatGPT followed the same pattern.

Yes, it had an incredible model.

But dozens of companies have built incredible models.

Very few became household names.

People shared prompts.

Screenshots.

Unexpected answers.

Funny mistakes.

Creative use cases.

The product didn’t just solve problems.

It generated stories worth telling.

That’s a very different kind of advantage.

The Question Every Founder Should Ask

This is where many founders still get stuck.

They think distribution begins after the product is finished.

In reality, distribution begins while the product is being designed.

Not with:

“How do we explain this?”

But with:

“What will make someone tell another person about this?”

No advertising budget can manufacture that moment at scale.

Only products and ideas that naturally create conversations can.

This doesn’t mean manufacturing outrage. Or designing fake controversy.

Artificial polarization is obvious. And it rarely lasts.

The lesson isn’t to build controversial products.

The lesson is to build products that are inherently conversational.

So instead of asking:

“How do we explain our product better?”

Ask:

“What conversation does our product make inevitable?”

Because in a world where AI makes creation cheaper every month, creating will no longer be the hard part.

Being talked about will.

And that may become the most valuable competitive advantage of the AI era.

Sold out.

I checked the next day.

Sold out again.

Then the day after that.

Still sold out.

I honestly can’t remember the last time a film I decided to see on a whim was completely booked for three or four days straight.

Isn’t that a little strange at a time when so many people keep telling us cinema is dead?

We have enormous screens at home. We have Netflix. We have Disney+. We can pause a film, rewind it, watch it from bed, and glance at our phones whenever we lose interest.

Humanity has reached the technological point where we can watch Odysseus’s ten-year journey home without ever leaving the sofa.

And yet people are still fighting for seats to see a film in a cinema.

And not only in Istanbul.

Thousands of people have reportedly travelled from across Europe to Prague just to see the film in true 70mm IMAX.

One family drove for three days from Norway. By the end of July, the Prague cinema had welcomed more than 22,000 viewers, running four or five screenings a day, including midnight showings.

The reason this particular cinema could do that is almost absurd.

While everyone else was switching to digital, it never got around to replacing its old projector.

That “laziness” eventually made it one of only 41 cinemas in the world capable of showing the film in true 70mm IMAX.

The technology its competitors abandoned became its competitive advantage.

Of course, keeping an old projector in your basement is not enough to persuade people to drive for three days.

Someone has to make that projector meaningful again.

That person was Christopher Nolan.

Nolan’s Real Product Isn’t the Film

The Odyssey is the first feature film shot entirely with IMAX film cameras.

That is easy to say.

It is much harder to do.

IMAX cameras are not merely enormous. They are also extremely loud. In the past, filmmakers generally used them for large action sequences, then switched to smaller cameras for quiet dialogue scenes.

Nolan refused to accept that compromise.

To shoot the entire film in the same format, he worked with IMAX to develop a custom enclosure that reduced the camera’s noise. Together, the camera and its soundproof housing became a colossal piece of machinery weighing roughly 180 kilograms.

The film stock also had to be changed about every three minutes.

A normal producer would look at all this and ask:

“Is any of this really necessary?”

Nolan’s answer is almost a summary of his entire career:

“Yes.”

But this insistence is not simply a technical fetish.

Nolan loves scale.

He loves explosions, oceans, wars, outer space, atomic bombs, collapsing cities, and time bending in on itself.

But those things are not what make his films truly interesting.

Nolan searches for enormous emotional contradictions inside enormous stories.

He does not use IMAX only to enlarge the world.

He uses it to enlarge the fracture inside a human being.

The Small Knot Inside the Epic

Most of us already know at least something about the world of The Odyssey.

The Trojan Horse. The Cyclops. The Sirens. The gods. A great warrior trying to make his way home.

But the centre of the story is not merely a collection of mythical creatures and naval battles.

It is a man who cannot escape the consequences of his own intelligence.

Odysseus is known above all for his cunning. He wins through strategy rather than brute strength. The Trojan Horse is the most famous example.

He turns an object people believe to be a gift into a machine of death.

It is a brilliant tactic.

It is also a deception that pushes ethical boundaries to their limit.

His intelligence wins the war.

But it does not end the story.

On the contrary, it begins the real one.

To me, Odysseus’s journey home is not only a struggle against gods and monsters. It is a long reckoning with his own decisions, his pride, his desires, and the consequences of the intelligence he chose to use.

That structure is perfectly suited to Nolan.

There was a similar man at the centre of Oppenheimer.

A man intelligent enough to solve one of the most complex scientific problems in history, yet unprepared to live with what that solution would do to the world.

Building the atomic bomb was a technical achievement.

Living with that achievement was a human tragedy.

This is exactly the territory Nolan keeps returning to:

The distance between the triumph of intelligence and the defeat of the human being.

The difference between an idea that works and an idea that is right.

The way winning a war changes the person who won it.

The vast stories, images, and budgets exist for this reason.

Not to conceal the emotional conflict, but to magnify it until it becomes impossible to ignore.

Why Do People Still Go to the Cinema?

At this point, it becomes easier to understand why the old projector in Prague suddenly became so valuable.

The projector alone is not an advantage.

The IMAX camera alone is not an advantage.

Homer’s ancient story alone is not an advantage.

Even Nolan’s name is not enough on its own.

The success comes from the way all these elements reinforce one another.

Nolan took an old story that everyone thought they already knew and connected it to the inner contradictions human beings have never managed to escape.

To tell that story, he chose a technology so difficult that most directors would never dare to use it for an entire film.

Rather than accept the limitations of the technology, he had the camera redesigned.

The small number of true IMAX projectors left in the world then became part of the film’s distribution system.

Scarcity transformed a screening into an event.

People travelled not merely to consume the content, but to become part of the experience.

Remove any link from this chain and the result changes.

If the film had been shot in an ordinary format, the projector in Prague would matter less.

If the story were nothing more than a visual spectacle, the film would become a three-hour technical demonstration.

If the projectors had not been preserved, there would be nowhere to show Nolan’s images in the form he intended.

And if the emotional centre failed, people might praise the image quality but forget the film a few days later.

This is why it is not enough to say that preserving old technology creates a competitive advantage.

A more accurate statement would be: An abandoned capability becomes a competitive advantage when it is connected to a powerful new human need.

Netflix Didn’t Kill Cinema

Netflix became exceptionally good at performing one particular function of cinema:

Making content easy to access.

Going to the cinema is inconvenient.

You have to choose a specific time. Sit in traffic. Find somewhere to park.

The person next to you may chew popcorn like an industrial crushing machine.

Watching at home is easier than all of that.

So if cinemas try to compete with Netflix on convenience, they will lose.

But the purpose of cinema is not merely to move an image onto a screen.

It had started to feel that way because the industry spent years making the experience increasingly ordinary.

The same auditoriums. The same small screens. The same digital images. The same popcorn.

Cinemas had already started turning themselves into Netflix before Netflix arrived.

Netflix simply performed that job better at home.

Nolan is moving in the opposite direction.

He has not made cinema easier.

He has made it even more difficult.

He does not tell the audience, “You can watch this film anywhere.”

He tells them, “If you want to truly see this film, you need to come to the right place.”

According to conventional business logic, that sounds like a terrible distribution network.

Fewer venues. More expensive infrastructure. More difficult operations. Limited capacity.

But as this example shows, reducing friction is not always the solution to a distribution problem.

Sometimes, when friction becomes part of the product, it strengthens the power of the distribution network rather than weakening it.

When you create an experience people will want to talk about once they reach it, the journey and the difficulty do not reduce its value.

They prove it.

This Is How We Should Look at “Dead” Businesses

In business, we love announcing that certain categories are dead.

Physical retail is dead.

Sales calls are dead.

Newsletters are dead.

Human customer service is dead.

Cinema is dead.

But perhaps the category itself is not dead.

Perhaps only the version that can be easily replaced has died.

An ordinary cinema experience can be replaced by a large television at home.

A screening that makes people travel to another country cannot be replaced so easily.

• A generic sales call can be replaced by an automated email. But understanding a customer’s business better than they understand it themselves is much harder to replace.

• A newsletter written like a press release can be produced by AI in seconds. But an original insight drawn from twenty years of experience, one that genuinely changes the reader’s decision is not so easily replaced.

• Standard customer service can be automated with a bot. But human judgement capable of understanding an angry, complicated customer and saving the relationship is much harder to replace.

I do not think the greatest opportunities of the future will emerge only in entirely new categories.

They will also emerge in the parts of supposedly “dead” categories that digitalisation abandoned, even though human beings still need them.

The goal is not to bring the old business back exactly as it was.

Imitating the 1980s is not a strategy for selling vinyl records.

Keeping an old projector operational is not a strategy by itself either.

The real question is:

Ritual?

Trust?

Belonging?

Craftsmanship?

Excitement?

The feeling of experiencing something together?

Once you find that value, an old capability that everyone else abandoned can become meaningful again.

Not Everything Old Is a Future Opportunity

There is an obvious danger here: nostalgia.

Something does not become valuable merely because it existed in the past.

Nor does difficulty make something valuable.

Doing a job manually does not automatically make it better than automating it.

Nolan’s success does not come from romanticising old technology.

It comes from understanding what that technology can still do uniquely well.

He then creates modern demand around that capability.

That distinction matters.

Successful revivals do not recreate the past exactly as it was.

They connect a unique capability from the past to an unresolved human need in the present.

The cinema in Prague preserved the projector.

Nolan gave it a purpose again.

Homer provided the story.

IMAX provided the scale.

Nolan placed human conflict at the centre.

And the audience contributed its time, its money, and sometimes a three-day journey.

Competitive advantage, as you can see, rarely comes from one brilliant decision.

It comes from a chain of decisions that reinforce one another.

So the next time I encounter a business that people say is “finished,” I no longer want to ask only:

“Would anyone still use this?”

There is a better question:

Because cinema did not die.

The kind of cinema people could more easily watch at home did.

That distinction will always matter.

Not because they weren’t looking hard enough.

Because they weren’t looking the right way.

That changed.

Using the Atacama Large Millimeter / submillimeter Array, researchers detected glycolaldehyde, a simple sugar and one of the molecular building blocks of life, in the gas surrounding a young star system.

The molecules had almost certainly been there long before we noticed them.

The universe didn’t suddenly become sweeter.

Our instruments and more importantly, our ability to interpret weak, noisy signals became better.

The Signal Was Always There

That’s how many of history’s biggest discoveries actually happen.

• Penicillin wasn’t invented. It emerged from a contaminated experiment that Alexander Fleming could easily have discarded.

• The cosmic microwave background wasn’t discovered because scientists were actively searching for evidence of the Big Bang. Arno Penzias and Robert Wilson were trying to eliminate what they thought was annoying interference from a radio antenna.

• John Snow didn’t need to collect an entirely new body of data to identify the source of London’s cholera outbreak. He plotted existing deaths on a map and revealed a pattern no one had seen before.

The data already existed.

The breakthrough came from seeing it differently.

Every breakthrough begins with the realization that the signal was always there.

We had simply mistaken it for noise.

Most Companies Don’t Have a Data Problem

This is also why I think most companies misunderstand AI.

They believe AI creates value by generating new content.

I think its biggest opportunity is somewhere else entirely.

AI is becoming humanity’s best tool for finding patterns that have been hiding in plain sight all along.

Most executives don’t have a data problem.

They have a perception problem.

Every company I meet tells me some version of the same story:

“We need more data.”

Then I ask a simple question:

“What data are you not collecting today that would completely change your business?”

Usually, the room goes quiet.

Because deep down, they already know the answer.

Almost none.

Companies are drowning in information.

• Customer support tickets.

• Sales call recordings.

• CRM notes.

• Emails.

• Slack conversations.

• Meeting transcripts.

• Product usage logs.

• Feedback forms.

• Invoices.

The average company has accumulated years of institutional knowledge.

It just exists as disconnected fragments.

The problem isn’t scarcity.

It’s interpretation.

Productivity Is Not the Real Prize

For years, AI has been marketed as a content machine.

Write emails. Generate proposals. Create presentations. Summarize meetings…

Those are useful capabilities. But they are not transformational. They are productivity gains.

The real shift happens when AI stops producing outputs and starts discovering relationships.

Imagine an AI that listens to every lost sales call from the past three years.

Not to summarize them.

But to uncover the objections that consistently appear six weeks before customers churn.

Or an AI that connects support tickets with feature usage and reveals that your highest-value customers aren’t leaving because of missing features.

They are leaving because onboarding created the wrong expectations months earlier.

No employee could realistically connect all those dots.

Not because people are incapable.

Because the volume exceeds human cognition.

That is where AI becomes more than a faster assistant.

It becomes an instrument of discovery.

The Best Companies Will Ask Better Questions

That’s why I believe the next competitive advantage won’t belong to companies with the biggest models.

Or even the most data.

It will belong to companies that ask better questions.

Not:

“What can AI generate for us?”

But:

“What has our business been trying to tell us for years that we’ve never been able to hear?”

That is a fundamentally different mindset.

One treats AI as a faster employee.

The other treats AI as a new scientific instrument.

History suggests the second group usually wins.

The telescope didn’t create new stars.

The microscope didn’t create new bacteria.

The map didn’t create the cholera pattern.

And AI won’t create the most valuable insights inside your business.

It will reveal the ones that have been quietly waiting there all along.

Your next breakthrough is probably already in the building. It’s buried inside yesterday’s noise.

At the World Cup, you have to give it back. The obvious reason is that the game needs to continue. The less obvious reason is that what you are holding is no longer just a football.

Inside Trionda, the official ball of the 2026 World Cup, is a motion sensor that records data 500 times per second. It tracks the ball’s movement, helps officials identify precise touches and sends information to the VAR system in real time. It even needs to be charged before a match.

Catching one in the stands is less like catching a souvenir and more like briefly taking possession of a very round computer.

During England’s quarter-final against Norway, that computer became the most controversial witness in football.

Norwegian goalkeeper Ørjan Nyland launched the ball high up the pitch. It climbed into the air, seemed to hang there for a moment, then dropped so abruptly that Norway’s players and staff became convinced it had struck one of the cables holding the camera system above the field.

Seconds later, Jude Bellingham scored England’s equaliser.

The rule was straightforward. If the ball had touched the cable, play should have stopped and everything that followed would have been irrelevant. Nyland protested immediately. Norway’s staff pointed toward the wire, the replays looked unusual, and manager Ståle Solbakken remained convinced that the ball had changed direction in mid-air.

But the most important witness was not the referee, the goalkeeper or even the Spidercam footage.

It was the ball itself.

FIFA examined the data from the sensor inside it, what it calls the ball’s “heartbeat”. A cable impact should have created an additional spike while the ball was airborne. According to FIFA, no such spike appeared, so there was no evidence of contact.

The goal stood.

The decision was settled. The argument was not.

Many England supporters and neutral observers accepted the sensor data as more reliable than human perception. Others watched the ball’s sudden movement and reached the opposite conclusion: they trusted what they believed they had seen with their own eyes.

The sensor had data. Norway had eyes.

Neither side thought the other had presented enough evidence.

That is what makes this story interesting. Not whether the cable was definitely touched, or whether millions of football supporters need an emergency lesson in the Magnus effect. The more useful lesson is that a system can produce an answer without producing trust.

Accuracy and trust are different products.

A machine may reach the correct conclusion and still fail to persuade people. That is because we do not trust technology simply because it gives us an answer. We trust it when we understand how the answer was reached, what evidence was used and whether the conclusion can be challenged.

Good technology should hide its operational complexity. It should not hide its reasoning.

AI is About to Learn the Same Lesson

Most companies currently have the opposite problem. Their AI systems are highly visible while the work is being done and strangely opaque when something goes wrong.

An employee chooses a tool, opens another tab, explains the context again and uploads the same files. They refine the prompt, inspect the result, move the output into another system and perhaps ask a second AI to improve what the first one produced.

Then they check the final work themselves, because nobody wants to discover a hallucinated number halfway through a board meeting.

We bought AI to reduce people’s workload. Instead, we created a new operating role: manager of artificial employees who require constant supervision.

One tool writes. Another researches. Another analyses spreadsheets. Another creates presentations. A fifth promises to automate the first four. Each product looks useful in isolation, but together they form a small digital organisation with no chief of staff.

The employee still has to break down the objective, select the right system, transfer the context, monitor the outputs and assemble the final result.

We automated production. We did not automate coordination.

This coordination cost rarely appears in a software budget. It appears in interruptions, repeated explanations, fragmented attention and the small decisions employees make every time they wonder which tool they are supposed to use.

One AI application may save ten minutes. Fifteen disconnected applications can create an entirely new management layer.

The obvious solution is to make the complexity disappear: one interface, one AI coworker and a collection of specialists working behind it. But simply hiding everything inside a black box creates a different problem.

The first time the system makes a questionable decision, the user will want to know what it did, which information it used and why it chose that particular path. They will also need some way to inspect the work, correct the result and remain responsible for the final decision.

A black box feels magical until the first controversial call.

The right standard for an AI coworker is therefore not that it never makes a mistake. Human coworkers fail that test before lunch.

A more useful standard is whether it can reduce the amount of coordination you perform without reducing your ability to understand what happened.

The work should become less visible.

The accountability should not.

THIS ISSUE IS BROUGHT TO YOU BY APPY.AI

One AI Coworker, Fifteen Specialists Behind It

This is what makes Appy.AI’s approach interesting.

Instead of giving employees another collection of separate AI tools to coordinate, Appy gives them one AI business partner called Violet. Behind Violet is a team of 15 specialists that can help with work such as marketing strategy, project management, financial modelling and presentation creation.

You describe the outcome you need in plain English. Violet determines which specialists are relevant, coordinates the work and returns with the result.

More importantly, this happens inside Slack or Microsoft Teams, the places where teams already communicate. There is no new dashboard employees have to remember to check, no code to write and no complicated workflow builder to maintain.

That does not mean people should blindly accept whatever an AI coworker produces. It means they can delegate the coordination without surrendering their judgment.

The aim is not invisible magic.

It is less management, fewer handoffs and a clearer path from request to result.

Hire Appy.AI and get $100 in free credits. No credit card required.

Hire Appy.AI

The AI industry remains obsessed with capability. Which model reasons better? Which agent can browse the web? Which tool can create the most impressive presentation?

Those questions matter, but they are not the questions most employees feel during the working day.

Employees feel the management burden. They notice how many systems they need to open, how often they must explain the same task and how much of the final output they still have to assemble themselves. They also notice when an automated system expects trust without offering enough evidence to deserve it.

The World Cup ball offers a useful model because the footballer does not have to operate the technology. The player simply kicks the ball. The sensor carries the context, sends the information to the relevant system and produces evidence when the moment requires it.

That is good product design.

But the controversy also showed that producing data is not the same as producing confidence. When the decision feels wrong, people want to inspect the evidence rather than being told that the machine has spoken.

That is trust design.

The winning AI products will not necessarily be the ones with the longest feature lists or the most specialists. They will be the products that remove the most coordination from the working day while making their actions easier to understand.

The World Cup ball did not ask the player to become a sensor operator.

AI should not ask every employee to become an AI manager.

Good technology should disappear from the workflow, not from accountability.

They see large companies hiring Chief AI Officers, building internal AI teams, running innovation workshops, buying enterprise platforms, and creating 73-slide transformation decks with words like “enablement”, “alignment”, “governance”, and “future-ready”.

Which is usually corporate language for:

Small businesses look at this and think AI is probably not for them yet.

• Too complex.

• Too expensive.

• Too technical.

• Too early.

I think that’s completely wrong.

The AI opportunity for small businesses is not building an AI department.

It is building one workflow that never sleeps.

And this distinction matters more than it sounds.

Because most businesses don’t lose money because they lack strategy decks.

They lose money in much uglier, more boring places.

• A lead comes in and nobody calls fast enough.

• A customer asks a question and waits too long.

• A cart is abandoned and nobody follows up.

• A sales call contains a buying signal and nobody notices.

• A customer says, “Your competitor is cheaper” and that sentence dies inside a call recording forever.

This is where money leaks.

Not in the cinematic business moments.

Not in the strategy offsite.

Not in the “we need to become an AI-first company” slide.

In the broken loops.

And this is where AI becomes useful.

Not as magic.

Not as a toy.

Not as another subscription your team tries for two weeks and then silently abandons.

As a worker inside a workflow.

That distinction is everything.

AI Looks Expensive When You Copy How Large Companies Adopt It

Small businesses should not copy large companies.

Small businesses have a different advantage: speed.

• No committees.

• No internal politics.

• No 18-month transformation project.

No one needs to schedule a meeting to decide whether another meeting is needed.

A small business can do something much better:

• Find one painful workflow.

• Put AI inside it.

• Measure the result.

• Improve the loop.

• Then move to the next one.

That is not an AI department.

That is leverage.

Where Small Businesses Actually Lose Money

Most businesses do not lose money because they lack ambition.

They lose money because important things happen only sometimes.

• Sometimes new leads get called quickly.

• Sometimes missed calls get returned.

• Sometimes abandoned carts get followed up.

• Sometimes customer feedback gets reviewed.

• Sometimes the CRM gets updated.

• Sometimes old customers get reactivated.

• Sometimes sales calls are analyzed.

• Sometimes the team notices a pattern before it becomes a problem.

“Sometimes” is expensive.

And the smaller the business, the more dangerous “sometimes” becomes.

Because small businesses usually do not have extra people sitting around waiting to catch what falls through the cracks.

• The founder is busy.

• The sales team is busy.

• The support person is busy.

• The agency is busy.

Everyone is busy.

And when everyone is busy, the boring but valuable work gets delayed.

• Follow-up.

• Reminders.

• Call summaries.

• Lead qualification.

• Customer check-ins.

• Objection tagging.

• Review analysis.

• Reactivation.

None of this sounds exciting.

That is exactly why it matters.

Boring work compounds when it is done consistently.

Boring work leaks money when it depends on human memory.

The First AI Employee Will Be Boring

I don’t think the first AI employee for most small businesses will be a strategist.

• It will not be a genius analyst.

• It will not be a brand visionary.

• It will not sit in a digital corner wearing a tiny Patagonia vest and telling you to “rethink the customer journey”.

The first AI employee will probably be much more boring.

• The person who answers the phone.

• The person who calls back leads.

• The person who follows up after a form submission.

• The person who reminds customers.

• The person who asks for feedback.

• The person who summarizes calls.

• The person who updates the CRM.

• The person who notices when ten customers complain about the same thing.

Very glamorous, I know.

But this is how leverage usually looks in real life.

Less Iron Man suit.

More “we stopped losing leads because someone now replies in 30 seconds”.

Your first AI employee should not be impressive.

It should be useful.

The Tool Is Not the Strategy

A lot of companies will waste a lot of money on AI over the next few years.

Not because AI is weak.

Because their thinking is weak.

• They will subscribe to one tool for writing.

• One tool for meetings.

• One tool for research.

• One tool for images.

• One tool for automation.

• One tool for customer support.

• One tool someone saw on LinkedIn and bought before lunch.

Then they will wonder why nothing changed.

A long AI tool list is not a strategy.

It is usually a very expensive junk drawer.

The tool is not the strategy.

The workflow is the strategy.

AI becomes valuable when it makes something important happen every time.

Not when it sounds smart.

Not when it generates a clever paragraph.

Not when it creates a cute image of a dog wearing sunglasses on a skateboard.

Although, to be fair, that has its place in civilization.

AI becomes valuable when it changes the operating rhythm of the business.

The Never-Sleeping Workflow

Here is the concept I think every small business should understand:

A Never-Sleeping Workflow is a business process where AI makes sure an important action happens every time, not just when your team remembers to do it.

That is the unlock.

Not more tools.

Not more dashboards.

Not more “AI experiments”.

One workflow that never sleeps.

For example:

• Every new lead gets called within 60 seconds.

• Every missed call gets an automatic callback.

• Every abandoned cart gets a voice follow-up.

• Every old customer gets reactivated after 60 days.

• Every sales call gets summarized and scored.

• Every customer objection gets classified.

• Every negative review triggers a recovery flow.

• Every support conversation becomes product insight.

• Every qualified buyer gets routed to the right person.

• Every repeated complaint becomes visible before it becomes expensive.

This is not science fiction.

This is not a 2032 thing.

This is the practical frontier of AI right now.

Small.

Specific.

Boring enough to be real.

Valuable enough to matter.

The E-Commerce Example

Take a simple e-commerce business.

They spend money on ads. People visit the site. Some add products to cart. Many don’t finish checkout.

Classic problem.

The usual response is predictable:

Spend more on ads. Send more emails. Add another pop-up. Change the discount. Redesign the checkout. Blame the agency.

Sometimes those things help.

But often the business is missing something simpler.

• What if the customer just needed a nudge?

• What if they had a question?

• What if they were comparing options?

• What if they were ready to buy but got distracted?

Now imagine an AI voice agent that calls selected abandoned cart users within minutes.

Not tomorrow.

Not when someone has time.

Not after a spreadsheet is exported, forgotten, reopened, assigned, and half-followed.

Immediately.

• It calls.

• It understands the customer.

• It answers questions.

• It offers help.

• It sends a payment link.

• It books a callback if needed.

• It updates the CRM.

• It summarizes what happened.

Then it runs again tomorrow.

That is not “using an AI tool”.

That is installing a workflow that never sleeps.

Big difference.

The Second-Order Effect

At first, you think AI saves time.

Then you realize it changes what your business can afford to pay attention to.

That is the real shift.

Before AI, many businesses ignored most customer conversations because listening to all of them was impossible.

Before AI, many old leads were never followed up because the economics did not work.

Before AI, many sales calls were never analyzed because managers could only listen to a tiny sample.

Before AI, many small tasks stayed manual because automation was too expensive, too technical, or too annoying.

Now the math changes.

And when the math changes, strategy changes.

AI does not just make existing work cheaper.

It makes previously unreasonable work reasonable.

• Calling every abandoned cart customer used to sound ridiculous. Now it can be a workflow.

• Analyzing every customer conversation used to be impossible. Now it can be a dashboard.

• Creating personalized follow-up for every lead used to require a team. Now it can be a system.

This is where small businesses can build an unfair advantage.

Not by pretending to be large companies.

But by using AI to do the consistent operational work that large companies are often too slow, political, or overcomplicated to implement.

What AI Exposes

There is one uncomfortable part.

When you put AI inside a workflow, it does not only automate the work.

It exposes the business.

• If your sales process is vague, AI will expose it.

• If your lead qualification is messy, AI will expose it.

• If your CRM is chaos with a login screen, AI will expose it.

• If your follow-up depends on someone remembering something, AI will expose it.

• If your offer is unclear, AI will expose it.

• If your team does not agree on what a good customer looks like, AI will expose it.

This is uncomfortable.

But useful.

Because most businesses do not only need automation.

They need clarity. AI forces clarity. It forces you to define the script.

• The trigger.

• The handoff.

• The success metric.

• The fallback.

• The next action.

• The owner.

• The outcome.

This is why “just add AI” rarely works.

If the workflow is broken, AI will not magically fix it.

It will show you exactly where it breaks.

That is not a bug.

That is the work.

The 5-Step AI Workflow Test

If you want to use AI in a way that actually matters, don’t start with tools.

Start with this test.

1. Find the leak

Where does money, attention, speed, or customer intent disappear?

Examples:

• New leads are not contacted fast enough.

• Missed calls are not returned.

• Abandoned carts are ignored.

• Customer complaints are not analyzed.

• Sales calls are not reviewed.

• Old customers are not reactivated.

The leak should be specific.

“Improve sales” is not specific.

“Call every qualified lead within 60 seconds” is specific.

2. Define the repeated action

What should happen every time, but currently happens only sometimes?

This is the most important question.

Because “sometimes” is where AI can create leverage.

If the answer depends on someone remembering, checking, copying, pasting, listening, tagging, or following up manually, you probably found something.

3. Remove human memory from the loop

Human judgment is valuable.

Human memory is overrated.

Do not build critical business processes around someone remembering to check a spreadsheet.

Or follow up tomorrow.

Or listen to a call later.

Or update the CRM after lunch.

That is not a process.

That is a hope with a calendar invite.

4. Put AI inside the workflow

Now decide what the AI worker should actually do.

Should it call? Answer? Summarize? Classify? Route? Remind? Score? Draft? Trigger? Update?

The goal is not to make AI look smart.

The goal is to make the workflow run.

5. Measure one business outcome

Do not measure vibes.

Measure one outcome.

Revenue recovered. Calls answered. Leads converted. Response time reduced. Manual hours saved. Churn reduced. Customer complaints detected. Sales objections classified.

If you cannot connect the workflow to a business outcome, you may still have something interesting.

But you probably do not have leverage yet.

The Practical Starting Point

So where should a small business start?

Not with:

“How do we use AI?”

That question is too broad.

It creates tool shopping, internal confusion, and long conversations that somehow end with someone suggesting a Notion template.

Start with this instead:

That answer is your first AI employee.

Maybe it is a follow-up agent. Maybe it is a missed-call recovery agent. Maybe it is an abandoned-cart voice agent. Maybe it is a customer feedback analyst. Maybe it is a review mining workflow. Maybe it is a sales call scoring system.

It does not need to be impressive.

It needs to work.

At Next Big App, this is how we think about building AI products.

Not more tools for the sake of tools.

Practical AI workflows that help businesses recover revenue, reduce manual work, and understand customers faster.

Because AI is not the advantage by itself.

The advantage is what AI makes happen consistently.

• Even when your team is busy.

• Even when it is after hours.

• Even when the lead came in at the wrong moment.

• Even when nobody feels like updating the spreadsheet.

• Even when the customer signal is buried inside the 847th call recording of the month.

That is the real promise.

Not an AI department.

Not another dashboard.

Not a bigger tool stack.

One workflow that never sleeps.

Start there.

He has four championships. Four MVPs. Every scoring record that matters. A son in the league. A body of work so complete that nobody would blink if he spent his last contract collecting a paycheck and a farewell tour.

He just told the Lakers he’s not coming back.

Not because he’s chasing a bigger check.

Not because LA benched him or disrespected him.

Because, by most reports, he wants to know the organization around him is still built to win, and he’s not convinced anymore.

So at an age where literally every incentive points toward comfort, he’s choosing the harder path. Free agency. A new system. Possibly a reunion with Steph Curry and the Warriors, walking into someone else’s locker room at 41 to chase one more real shot.

I want to talk about why that decision is more instructive than anything he’s ever done on a basketball court.

The Default Setting Is Coast

Here’s something I’ve watched play out over and over, in sports and in business: the moment someone has “made it”, their default setting flips from compete to protect.

Founders do it after the first real exit.

Executives do it after the corner office.

Operators do it the moment the business finally stops being an emergency every week.

It’s not laziness. It’s not even really comfort-seeking on purpose.

It’s quieter than that. You stop asking “is this still the best environment for me to keep getting better in” and start asking “is this still fine”.

Fine becomes the bar. And fine is a very low bar to clear once you’ve already won.

The tradeoff nobody wants to admit is this: the years right after you’ve proven yourself are usually the years you’re most likely to start declining, because that’s exactly when you stop demanding the environment prove itself back to you.

The companies that avoid this trap are the ones everyone remembers, because they’re rare.

• Reed Hastings didn’t wait for DVD rentals to die before he killed them himself, he redirected Netflix into streaming while the DVD business was still printing cash and still the undisputed market leader.

• Steve Jobs did the same thing with the iPod, a product generating obscene margins with zero real competition, by building the iPhone knowing full well it would cannibalize his own cash cow.

Neither man waited for the market to force the decision. They ran the audit early, while everyone else in the room was still enjoying the view.

LeBron inverted that.

What Inversion Actually Looks Like

Inversion thinking isn’t a productivity hack. It’s a discipline: instead of asking “what does winning look like”, you ask “what does losing look like, and am I already doing it”.

Losing, for someone in LeBron’s position, doesn’t look like a bad season. It looks like staying somewhere out of familiarity.

It looks like optimizing for the easy version of the story, the retirement tour, the jersey retirement, the soft landing, instead of the accurate version of the story, which is: am I still in the best possible system to compete for what I actually came here to do.

Reports say he spent last year monitoring whether the Lakers were building a genuine contender around him, even while he was still under contract.

That’s not entitlement. That’s someone refusing to let inertia make the decision for him.

Most people, at 41, with his resume, never run that audit. They let the calendar and the comfort of “this is just where I am now” make the call.

He ran the audit anyway. And the answer he got was no.

There’s an old story from Intel that captures this exact move.

By the mid-80s, Intel was still known as the memory chip company, even as Japanese competitors were quietly eating that business alive.

Andy Grove turned to his co-founder Gordon Moore and asked him a question that later became legendary: if the board fired us both and brought in a new CEO, what would that person do?

Moore didn’t hesitate. He said the new guy would get Intel out of memory chips.

So Grove and Moore walked out of the room, and walked back in as if they were that new CEO, and made the call themselves.

That single mental trick, firing yourself in your head and asking what a fresh set of eyes would decide, is precisely what LeBron ran on his own career.

He didn’t wait for a front office to tell him the Lakers weren’t built to win.

He fired himself as the loyal incumbent and asked what a player with no sentimental attachment to the jersey would do.

Why This Matters If You Run Anything

If you’re a founder or an operator, here’s the uncomfortable mirror.

You built something. Maybe it’s working. Maybe it’s more than working, it’s genuinely good, and you’ve earned the right to relax a little.

That’s exactly the moment the coast-default kicks in, and it’s exactly the moment you should be running LeBron’s audit on your own situation instead.

Is the team around you still built to win the next version of the game, or just the version you already won?

Is the system still sharpening you, or are you just occupying a familiar seat in it?

Are you optimizing for the comfortable story you get to tell later, or the accurate one about whether you’re still getting better?

Nobody forces this audit on you once you’ve had success. That’s the trap.

Early on, the market forces honesty on you constantly, you either grow or you die.

Later, you can coast for years on reputation before anyone notices the decline, including you.

I’ve watched a version of this play out with founders after their first real exit. The wire hits, the title softens into “advisor” or “board member”, and suddenly the same person who used to obsess over every metric is coasting on a comfortable retainer.

Nothing is technically wrong. It’s just quiet. It’s fine. And fine, again, is the trap.

The founders I respect most are the ones who take the exit money and immediately go put themselves back in a room with real risk, a new startup, an angel check with actual downside, anything that puts them back in a system that has to prove itself to them again.

Jeff Bezos used to tell Amazon the same thing in different words.

• Day one is intensity.

• Day two is stasis, followed by irrelevance, followed by a painful decline, followed by death.

His whole point was that day two doesn’t announce itself. It arrives disguised as comfort. Which is exactly why the mental move matters more than the calendar age.

LeBron isn’t behaving like a 41-year-old future Hall of Famer collecting a farewell tour. He’s behaving like it’s day one.

LeBron just showed 41-year-olds, and honestly everyone else, what it looks like to refuse that trap on purpose, in public, with everything to lose reputationally if it doesn’t work out.

The Sharp Takeaway

Winning doesn’t earn you the right to stop competing for your own environment. It just raises the price of staying somewhere that no longer deserves you.

The people who stay great longest aren’t the ones who coast on what they’ve already built.

They’re the ones who keep asking, at every level of success, whether the system around them is still worth staying in and they leave the moment the honest answer is no.

It may simply be creating a mystery.

More specifically, sparking a debate around whether your product, your content, or even the people behind it are real humans or “AI”.

Many of you have probably seen the recent viral wave around Mebrure’s song “Değilim”.

She hid her face, used a digital avatar, and suddenly everyone in the comments started arguing about whether the whole thing was completely AI-generated.

And what happened?

With zero ad budget, she generated millions of organic impressions in a very short time.

In her official statements, the artist clearly said that she is not AI. She said the lyrics, composition, and voice all belong entirely to her.

But that was almost beside the point.

Today, when every feed is full of content that looks and sounds the same, radical transparency does not always create attention. Sometimes, leaving a little uncertainty does.

When you introduce just enough ambiguity, people start debating in the comments. And that is exactly the kind of engagement signal algorithms love.

So these days, whether AI is actually involved or not can become a reason for virality.

You can probably think of many global examples where brands, artists, or products did not explain everything perfectly on day one. Instead, they left some room for speculation, curiosity, and public interpretation.

How something spreads is often bigger than the thing itself.

Maybe it is time we started treating curiosity not as an accident, but as a deliberate part of our growth strategy.

To boost subscriber retention, Netflix consistently experiments with new entertainment formats.

We saw this early on with Bandersnatch, their choose-your-own-adventure interactive episode, though it didn’t trigger a massive wave of follow-ups.

While the platform currently hosts a library of casual mobile games, Unhinged marks an ambitious shift toward a genuinely immersive horror experience.

Video:

Here’s how it works: you scan a QR code on your TV screen, and your smartphone instantly transforms into the character’s flashlight.

Wherever you point your hand in real life, the beam of light on the TV moves in perfect synchronization.

When your character receives an in-game phone call, your actual phone starts vibrating; the caller’s voice comes directly from your handset, while the eerie ambient noise continues to blast from the TV.

Instead of forcing players to buy a $1,000 VR headset to feel true suspense, Netflix achieved brilliant horror atmosphere through clever sensory synchronization.

They are seamlessly blurring the line between the digital world and physical reality.

As product builders, this is a great reminder that breakthrough innovation doesn’t always require inventing entirely new technology from scratch. Instead, it’s about how creatively we can bridge the tools and habits users already have. We should all take a page out of this design playbook.

The rig detects the incoming pigeon, takes aim, and —bzzzt— blasts it with water.

Video:

Honestly, this might be the clearest lesson in software and business development I’ve seen in a long time.

Think about it: no major brand is ever going to mass-produce a device for that hyper-specific, niche problem on a single balcony. But the friction this guy faced was so real and frustrating that he took matters into his own hands and built his own solution.

I believe this is exactly the crisis companies are facing in the software world today.

We hit a bottleneck in our workflow that is entirely unique to us. To fix it, we go out and buy a massive, clunky, off-the-shelf software license. We end up using barely 5% of its features while forcing our team to mold themselves into their rigid frameworks.

But there’s a flip side to the coin. Recognizing this need and dumping the job on your internal team with a casual, “Hey, let’s just build a quick automation here” is a massive trap.

You inevitably end up with amateur, fragile systems that collapse at the slightest breeze.

Hyper-specific problems require professional engineering. Just like you wouldn’t ask your accountant to pick up a welding torch to build a balcony rig, you shouldn’t expect your in-house team to become AI systems engineers overnight.

The real way forward is to stop forcing your business processes into cookie-cutter software. Instead, partner with “digital mechanics” who can custom-design and scale these digital rigs specifically for you.

We need to leverage the power of AI and technology the right way. Keep your team focused on their core competencies, and leave the bespoke AI systems to the experts.