Matt Lerner told the story at BoS Europe 2026. Elemental Software made a WYSIWYG website builder, drag and drop, connect to a database, right in the middle of the dot-com boom. The CEO hired a professional comedy writer and sent Lerner’s team to executive conferences to perform fake product demos in front of the room’s most important investors, on the theory that laughter would translate into offers.

The first attempt was Atonement.com, a fictitious confession website where visitors could enter their sins and receive an appropriate penance. They performed it straight after Steve Forbes, then running for the Republican presidential nomination, used his dinner speech to describe an abortion procedure in graphic detail. “Worst warm-up act ever,” Matt said. The room was tense enough that the comedy demo landed hard. It got laughs, business cards, no offers.

The second attempt raised the stakes: a fake site where you could start your own religion, complete with two original songs from a hired composer and a professional singer performing live on stage. Again, the room loved it. Again, nobody bought the company.

The third attempt is the one that worked. Matt’s co-founder, Peter, went through the attendee list ahead of time and built the demo around the people in the room, giving investors like Mary Meeker and Vinod Khosla actual speaking parts in a fake IPO pitch. After the show, the chairman of the board of Macromedia walked up to Peter and said, “We’re actually looking to buy a company in your space. Can we talk?” Negotiations followed, backed, by Matt’s account, by nothing more sophisticated than a yellow copy of Negotiating for Dummies on Peter’s desk. The deal closed at $20 million.

That would be a fine story on its own. What makes it worth telling is what happened next. Matt joined Macromedia as part of the acquisition and was handed a cubicle previously occupied by someone else, complete with a shelf of abandoned binders. Cleaning them out, he found a board memo: the proposal to purchase Elemental Software, authorizing Macromedia to pay up to $78 million. His company had sold for roughly a quarter of what the buyer was prepared to spend.

He told Peter that night at dinner. Peter’s reaction was one word, followed by stuffing the memo into his laptop bag and telling Matt never to mention it again. Matt kept that promise for 25 years, until he caught up with Peter over the most recent Christmas and got released from the arrangement, on the condition that he tell the whole story, memo included.

The lesson isn’t really about comedy demos, though those clearly worked as a way to get noticed in a room full of people who’d seen every standard pitch before. It’s about the gap between getting a company bought and getting it bought for what it’s worth. Differentiation, memorability, and nerve got Elemental Software an offer that a conventional sales process hadn’t produced in two prior attempts. None of that told the team anything about how much room the buyer actually had. Only a forgotten memo in an old cubicle answered that question, and by then the deal was already signed.

Watch Matt Lerner’s full talk to hear the whole thing, including the specific line that made Peter swear him to secrecy.

That was the starting point for Chika Emebo and Mark Stephens at Business of Software USA 2025, in a talk about running a team that spans four generations, Baby Boomers, Gen X, Millennials and Gen Z, often in the same room, sometimes reporting to a boss the same age as their own child.

Mark, who runs UK software company IDRsolutions, and Chika, a Gen Z member of his team, built the talk as a back-and-forth between their two generations rather than a lecture from one. The point wasn’t that generational differences are a problem to manage around. It’s that they’re the raw material for a better team, if you’re deliberate about it.

Your team should feel like the wrong crowd

Mark’s argument for hiring across generations rests on a simple test: if hiring is done right, walking into your team for the first time should feel like you’re the normal one among a group of slightly odd misfits. A team isn’t a friend group. It’s built from people with complementary skills you wouldn’t naturally choose to hang out with, which is exactly what makes it work.

He points to the Beatles. John and Paul argued constantly and wrote the songs everyone remembers. George Harrison wrote “Here Comes the Sun,” now the most streamed Beatles track, ahead of anything Lennon-McCartney produced. And Ringo Starr, by his own admission not the best drummer in the world, or even in the band, was the ingredient that made the group work. When he left, the others asked him back. Every strong team, Mark argues, needs its Ringo: the person who isn’t the obvious star but holds everyone else together.

Culture without the emotion

The most practical idea in the talk borrows from Bob Moesta’s jobs-to-be-done framework, applied to people instead of marketing. Every six months, Mark sits down with everyone on his team and runs through a simple pushes-and-pulls conversation: what’s pushing you away from this job, and what’s pulling you toward it. Doing it with everyone, on a schedule, strips out the fear that normally comes with “can we talk.” There are good reasons to leave a job, wanting to move from a small company to a bigger one, and bad ones, feeling disrespected or unable to use your skills. The framework exists to tell the two apart before they turn into a resignation.

Chika’s answer, when Mark asked how a business selling PDF conversion software (not curing disease, in his words) gives Gen Z the meaning they’re often accused of demanding, was more specific than the usual “purpose” answer. She wants a safe space to make an impact, to see her ideas taken seriously and have the results reach customers, plus the autonomy to be herself outside strict role boundaries. The skills she’s picked up in the office have paid off directly: one brand partnership from her side work as a content creator was worth over $2,000, funded by communication skills she says she wouldn’t have built without the job.

Small things signal culture

The talk’s smallest example carried its biggest point. Mark’s office got asked to bring in their own milk. One team member spent a Saturday designing “milk monitor” badges as a joke, and now everyone wears one. The badges aren’t the point, the fact that the whole office is in on the joke together is. He contrasts that with humour that splits an office into two camps, which he calls a straightforward sign of a toxic culture.

The same logic runs through his advice on self-expression: let desks look different, let laptops carry different stickers, ask new hires directly how they like to work and receive feedback rather than assuming. Demand conformity all day and then ask people to be creative in a meeting, and you’ll get nothing, because you’ve spent all day training the creativity out of them.

None of this replaces the harder structural questions a mixed-generation team raises, hybrid schedules, who works from where, how much autonomy is too much. A Q&A question from the audience about a Gen Z developer who wanted to work from Spain all summer got a blunt response from Mark: if someone doesn’t want to be in the office, the more useful question is whether you’ve actually sold them on why being there is worth it.

The full talk, including the audience Q&A, is available on the BoS site.

They spend their coffee meetings with the same core circle, people they already know and trust. It’s comfortable, but it’s also where the time gets wasted.

John Knox, in his lightning talk at Business of Software USA 2025, put it plainly: “The edges of your network are the most valuable part of your network. Yet most people spend all their time on the core.”

When John ran an event for med tech startups, he reached out cold to eight founders he’d never met. Seven said yes and ended up speaking at his event, despite having no prior relationship with him and no particular reason to trust his outreach. The leverage was never in his existing contacts. It was in the people he hadn’t talked to yet.

Networking Is Not Sales (or Small Talk)

Founders often conflate networking with sales, assuming it means pitching, small talk, and conference badges. John draws the line simply: “As soon as you talk about prices or establish yourself as a seller and someone as a customer, that’s sales, not networking.”

Networking is the work of building relationships around shared interests and mutual opportunity. “A network is something you do something with,” he says. “It’s interactive.” The moment you’re positioning yourself as the seller and someone else as the customer, you’ve left networking behind.

You don’t need to be in the room to do it, either. If you can only make it to one event a year, Zoom fills the rest of the gap. On outreach itself, John’s advice is specific: “Explain why you’re contacting them. Show respect, maybe flatter them. Prove you’ve done some research and you’re not a bot… Keep it short. Nobody wants to read a novel.”

Two Modes: Use and Expand

John describes two ways to work a network.

One is using it: you have a problem or an opportunity, and you lean on people you already know to help you move forward.

The other is expanding it: you have a direction you’re heading, and you go looking for the people who can help you get there. CEOs often come to John for strategic advice, but he points them somewhere else. The best intelligence comes from the market itself, from talking to the founders you’re trying to learn from and the customers you’re trying to serve. Your existing network will mostly confirm what you already believe. The edges are more likely to tell you what’s actually true.

For the Introvert: Speak Without Hedging

If you’re an introvert, and John counts himself as one, this can feel exhausting. It’s learnable anyway. His first piece of advice: speak confidently about your own work. “I’d hedge or underplay my achievements, and it shows a lack of confidence to people,” he admits. “It undermines whatever you’re trying to achieve.”

From there it’s a matter of showing up, one conversation at a time. Meet the person standing next to you. Help someone find their way around the conference. Make an introduction. Suggest a follow-up call. None of it feels like much on its own, but each small move fills in more of the map.

The Tree Model

John’s last framework: picture your business as a tree. The branches are your business model, product, and market. The roots are your network, and as the tree grows, the roots need to grow with it, into the same new directions.

That work looks nothing like coffee with people you already know. It’s reaching out to strangers and asking for their help, which is uncomfortable by design.

As John puts it: “If you invite the right people to participate in an ambitious project, many people will say yes. Just ask Obi Wan.”

The ChatGPT group could recall 83% less of what they had supposedly just written, nearly every student in that group, and they reported a much lower sense of ownership over their own work.

Georgiana Laudi, co-founder of Forget The Funnel, used that study in her talk at Business of Software USA 2025 to make a point about SaaS go-to-market strategy: the same thing is happening inside marketing and growth teams right now, and it’s quietly undermining their ability to make good decisions.

Her talk wasn’t really about AI. It was about three things she believes are the actual source of the disruption founders are feeling, and none of them show up on an AI roadmap.

#1: Your customer’s problem hasn’t stayed still

Understanding your customer is not a new idea, and most founders would say they’ve already done this work. But the research most teams did even a year ago is already out of date. Gia pointed to three examples from the last twelve to eighteen months: –

• Klarna rolling out an AI assistant and cutting headcount equivalent to 700 support agents
• Zoom and Google shrinking the market for tools like Otter with built-in AI meeting summaries
• Notion and Canva eroding demand for writing tools like Jasper with AI features baked into products people already use.

Budgets have shifted, roles have changed, and your champion buyer inside a customer account might not be who they were a year ago. If your last round of customer research predates this, it’s telling you about a customer who no longer exists.

#2: There’s a gap only your team can see

Gia’s team ran a study asking founders and their employees what was keeping them up at night. Founders talked about strategy: repositioning, messaging, product gaps. Not one mentioned team alignment. Nearly every employee did, describing the exact same frustration in different words: nobody seemed to know what the company was trying to achieve or who was deciding. The average company in the study had $2.5 million in ARR. These were not scrappy two-person startups without the resources to communicate. The gap wasn’t new, but it’s widening, and founders are usually the last to notice because they’re the only ones who can’t see it from inside their own team.

#3: Understanding is what gets you unstuck

Gia described this state as strategic quicksand: hard to recognize until you’re stuck, more activity making it worse rather than better. Her clearest example was Bitly.

Two hundred employees, $100 million in ARR, 6 million users, a company that had quietly expanded from a link shortener into a multi-product platform without anyone translating that shift into a shared plan. Every team solved for its own version of the opportunity, and customers ended up with a disjointed experience stitched together from marketing, product, and sales.

The fix wasn’t a new campaign or a rebrand. Bitly’s team identified three real customer segments and made a deliberate trade-off to build their entire go-to-market strategy, messaging, onboarding, and even parts of the product roadmap, around the two that mattered most. One of those segments turned out to have three times the retention of the other, a fact the product team only found once the whole company was working from the same understanding of who they served.

None of this requires a bigger budget or a different AI tool. It requires putting in the work AI can’t do for you: understanding your customer’s changed reality, closing the gap between what leadership sees and what the team is living, and building that understanding into something the whole company shares.

Gia’s full talk goes deeper into how to operationalize this, including two more customer examples and a practical framework for keeping a team aligned as things keep changing.

Next Events

By 2019, the company was ready to sell. Several offers came in, and Raj was deep into the checks a private equity buyer runs before signing.

The Monday Before Christmas

Super Salon’s largest customer announced it was building a competing product. The buyer pointed to the material adverse change clause, the part of the contract that lets a buyer back out if something big goes wrong before signing, and left.

You can do everything right for four years straight, and still lose the whole thing to one decision somebody else made without asking you.

A few months later, a whistleblower came forward. That same customer had reverse-engineered Super Salon’s database and taken its intellectual property. Raj offered to buy them out nut they told him they’d rather bleed him dry in legal fees, so he sued.

There’s unlucky, and then there’s this bad timing you can shrug off. Getting robbed by the same customer you’ve been serving for years is harder to make peace with.

April 2020

Haircuts were declared non-essential. Super Salon went from serving 10,000 locations to almost none. Inside days, 70% of the company’s revenue was gone. Around the same time, the competitor that had taken its IP raised $150 million from Tiger Global.

He lost 70% of his revenue in days. In the same stretch, the people who robbed him were handed $150 million. I don’t know how you carry both of those at once.

“It Wasn’t the Strategy That Held Us Together”

Raj got vulnerable with his leadership team, by his own account for the first time ever. He had no idea what they would say back. Nobody on the hundred-person team left.

“It wasn’t the strategy that held us together,” he said. “It was the culture. The openness. The honesty.”

He built that culture years earlier with none of this in mind. He was just trying to run a good company. That’s the thing that held when everything else stopped working.

The Settlement

Raj had a real shot at winning the lawsuit outright, but the courts were closed and the timeline was anyone’s guess. Fighting indefinitely meant burning cash and attention the business didn’t have to spare. So he split the company in two, settled one side and sold the other.

Everyone made money, and the team and customers who stuck it out came out protected. It was a smaller ending than the one he had spent four years building toward.

“I Love Creating Culture”

Raj closed his lightning talk with four things he took from the experience.

• You fall to your habits, not your ambitions, when you’re tested.
• Opportunity cost shows up in small daily decisions more than big ones.
• Process is fragile and rarely smooth, so build it while things are good rather than when you’re already burned out.
• And know your why.

“When things were darkest, I got out of bed because I love building people,” he said. “I love creating culture. I loved the impact we had – ten thousand locations times eight stylists times a hundred customers a day. We touched billions of people.”

The habits point is the one I’d actually act on. Ambition is what got the company to a deal in the first place. Habits are what were still standing after it fell apart. Worth asking what you’re building on a normal Tuesday, not just what you’re aiming at.

Next Events

Joel Gascoigne’s talk at BoS USA 2025, “Navigating a Multi-Year Decline to New All-Time Highs,” makes an uncomfortable case: Buffer didn’t start tanking because they made stupid mistakes. They started tanking because Joel followed the exact advice every SaaS expert tells you to follow.

The Playbook Trap

When growth first slowed down, Joel did exactly what the startup textbooks say to do. He decided to focus on a narrower type of customer, choosing direct-to-consumer e-commerce brands because they relied so heavily on social media. To make more money from each existing customer, the team restricted the free plan, raised the Pro tier from $10 to $15 while cutting features, and eventually hid the free sign-up option altogether. They also started building more products so they could sell customers more than one thing at a time, and hired search firms to run four executive hunts at the same time.

On paper, the early data looked good. They were making more money per customer. But their paying customer count was collapsing underneath it. They told themselves the drop was just a natural side effect of chasing bigger, pricier customers. By the time the pandemic added its own pressure, the business was in a real spiral.

Two Different Ways to Grow

It took Joel three years to see this: there isn’t one universal way to scale a tech company. There are two different ways to run one, and mixing them up is exactly how you end up squeezing a company that was never built to be squeezed. One way relies on a small number of customers who each pay a lot, get personal attention, and see their prices go up steadily over time. The other relies on a large number of customers, a simple product, a strong flow of people trying it free and slowly deciding to pay, and small, steady growth in revenue over time.

Buffer was built from day one to be the second kind of company. Forcing it to act like the first, just to chase bigger customers and higher prices, put them on a path where they were squeezing more out of a shrinking base instead of delivering more to a wider one.

Turning Inward

Before any of that, Joel hit his own wall. After the layoffs and after his co-founder left, he thought he had things stabilized. Then he burned out completely. In his words, he couldn’t get out of bed. He took six weeks off, did nothing for a while, then slowly built himself back up. It’s a pattern that shows up a lot after burnout: once someone recovers, they usually come back wanting a fresh start, wanting to change things so they don’t end up in the same place again. That’s probably part of why the decisions that followed were so personal. He’d already had to lay off ten people once. He didn’t want to be in a position where that happened again.

I keep noticing this pattern, not just with Joel. The best decisions a company ever makes rarely come from a strategy deck. They come from the worst, most personal moments a founder goes through, the ones they never planned for and can’t fully explain to anyone else. Joel didn’t get to four day work weeks or a people first culture through a growth framework. He got there because he broke down first, and decided he never wanted to put anyone else through what broke him.

That same instinct, to look after people first, is where the four day work week came from. It started in May 2020, in the middle of the pandemic, when Joel’s team was dealing with kids at home and life turned upside down. For Joel, trying it felt like an easy call. For the team, it felt almost too good to be true, a full extra day off with no cut in pay or hours squeezed into the rest of the week. They ran a one month trial, then a six month pilot, and it just stuck. It’s still how Buffer operates today, five and a half years later.

Two years after that, in 2022, the same instinct showed up again in a bigger decision. Joel stopped the executive searches. He realized bringing in more managers during a decline just added distance between himself and the actual work. He took on the direct reports himself, told the product and marketing teams who’d been stuck waiting for a new leader that they were a complete team now, and it was time to build.

They shifted the company out of just managing the decline and back into actually building again. They ran team-wide weeks focused only on shipping, to pick up speed, and finally tackled the old, avoided technical problems everyone had been too afraid to touch.

Strategically, Joel reversed the push toward bigger customers with a framework he calls Down and Wide. Down meant targeting individual creators, early entrepreneurs, and small businesses, the opposite direction of competitors like Hootsuite. Wide meant keeping the product flexible enough to handle a wide range of social channels.

They brought the free tier back as a core part of the business instead of something hidden. They changed the main button on their homepage to send visitors straight to free sign-up instead of a pricing page. Monthly sign-ups jumped from 65,000 to 85,000 almost immediately.

Back to All-Time Highs

By matching the business to what Buffer actually was, instead of what the textbooks said it should be, the business recovered. In 2024, Buffer grew revenue by 9%. By 2025, they were back to all-time highs in paying customers, hit a healthy balance of growth and profit that most SaaS companies chase and rarely reach, and were projecting $2.5 million in net income, all while still running on a four day work week. They even started proactively canceling subscriptions for customers who were paying but inactive, on the idea that long-term trust matters more than inflating short-term numbers.

Joel nearly lost Buffer by making good calls that belonged to someone else’s business. If you’re running a company right now and something feels off even though you’re doing everything the way you’re “supposed” to, that might be your answer too. Before you copy the next piece of advice from a founder you admire, ask yourself if you’re actually building the same kind of business they are. And ask yourself, honestly, whether the people inside your company are doing okay. Buffer only found its way back once Joel stopped treating both questions as separate.

Jason Cohen’s “Escaping the Growth Ceiling” talk uses Buffer’s real churn numbers as its worked example, and Cohen’s own approved post already references Joel’s talk as a companion piece. Worth cross-linking the two once both are published.

What is Prospect Intelligence Scoping?

Our new skill takes your ICP and walks you through a structured process to establish: where do those folk hang out (LinkedIn, Slack communities, email lists, specific industries)? What’s the best way for you to reach each one individually? It tests each assumption against real data before you commit to outreach.

As always with BoS OS, you, the human, are in charge.

The output is a working prospect list. Not a spreadsheet of guesses. Not LinkedIn filters. A list of people who match your ICP, with a mapped route to reach each one individually (email address, LinkedIn profile, mutual connection, event they’re attending).

You can route and monitor follow-ups in the BoS OS and existing systems. Once you’ve run it, you have a repeatable process. When your ICP sharpens (and it will, because every customer conversation teaches you something), you run the skill again. You scale what works.

This doesn’t replace your instinct or your sales skill. It removes the manual research work so you can focus on the conversations that matter. You spend time on people worth reaching, contacted through the channel that works for them.

Install Prospect Intelligence Scoping from BoS OS Advancing Skills.

Let us know what you think, suggest another skill or submit your own.

I had no idea what BoS OS was when I joined. I’d used AI tools before, sure, but just on my own, for my own stuff, no different from how a lot of people are dipping their toes into AI right now, half curious and half unsure it was worth the time. Then I walked into Business of Software right in the middle of a much bigger shift, one that started after a talk from Tim Barker got everyone in the room thinking differently about what AI could actually do for a company like ours. That talk is part of why BoS OS exists in the first place. I just happened to show up while it was still being built.

No map, no handover

Nobody handed me a system on day one. Jann was out on maternity leave. Mark was buried in conference prep. The person who had my role before me had one week left before she was gone. Great timing all around. There was no drive with everything sorted. No map of how anything worked. I inherited a pile of systems nobody had time to walk me through, and I did not know where to start.

There was no proper system yet, so I built one for myself, temporarily, just to keep things moving while we figured out the real one. We’re still figuring it out, honestly. The first thing I actually wrote down was our passwords. Not exactly the heroic starting point you’d expect.

The real work started with a meeting.

No onboarding doc existed. So I made one live, during the call, asking every question I had while someone answered. Whatever came back, I typed straight into a Google Doc as we talked.

That document got me through those first weeks. I built it out of pure necessity, not because anyone told me to.

Making the system fit me

Even with that doc, BoS OS itself felt like a lot to take in, the whole way we worked had changed at once, and some days I couldn’t tell what my actual job was anymore.

Then Mark said something that stuck: BoS OS is supposed to make our lives easier, not add more work on top. I stopped trying to force myself to fit the system and started making the system fit me instead.

I have ADHD, and a wall of text will lose me every time. So I told the OS how I needed it to talk to me: plain language first, the simple version before the technical one, and show me, don’t just tell me, because staring at pure text for too long pulls me somewhere else entirely. I built myself a dashboard where every task lives, and I cross it off when it’s done. Watching that happen gives me a kind of satisfaction a text reply never did. I even taught it to call a hard stop when I’ve hit my limit for the day, because left on its own, AI will happily pull you down a rabbit hole for hours and act like that’s normal.

That is when BoS OS stopped being something I opened and used. It started working with how my brain actually runs, and started helping me decide instead of just answering.

What’s still unfinished

We are still fixing things. Naming conventions are half sorted. Sync issues are getting better but not solved. What is different is what the process itself surfaced: parts of BoS that had been broken for a long time and nobody had noticed, workarounds nobody questioned, extra steps that never needed to exist. Building this properly meant we couldn’t look away from any of that.

My own role got clearer too. I know what is mine to own, what BoS OS can carry for me, and what still needs my hands on it directly. The manual grind that used to eat my whole day is mostly gone now. What is left is room to actually work on growing BoS instead of just keeping it standing.

If you’re new somewhere and you got handed a mess with no map, I get it. I wasn’t trying to fix anything. I just needed a way through mine.

If you’re staring down that same pile with no map, that’s exactly what the Introduction to BoS OS workshop is for. Two sessions, real example, no jargon required to start.

Jason Cohen’s BoS talk, “Escaping the Growth Ceiling,” takes that excuse away in the first few minutes.

His diagnosis is simpler and a lot less comfortable. You hit a mathematical ceiling. You were always heading toward it. You just never ran the numbers that would have told you it was coming.

Jason walks through the mechanics using real data from Buffer and ConvertKit, showing how the math works against you by default, and what the companies that break through actually do differently.

I’m still new to marketing, so a lot of what I’ve learned so far is the “just post more, try more channels” school of thought. This talk was the first time someone showed me why that advice runs out of road on its own.

Marketing grows linearly, cancellations grow exponentially.

How you get customers and how you lose them are two completely different shapes.

Marketing is linear. Even when you optimize your channels, you add a roughly steady number of new users each month. Draw it on a graph and it’s a straight line.

Cancellations are exponential. This is the part that sneaks up on you. Churn is a percentage of your total base, so the bigger you get, the bigger that number gets too.

As Jason puts it: “Marketing grows linearly. Cancellations grow exponentially. That’s what a percentage is.”

Say you have 1,000 customers and 5% monthly churn. That’s 50 people gone a month. If marketing brings in 100 new users, you’re still up 50. Feels fine. Feels like you’ve cracked it.

Now triple the business. 3,000 customers. Same marketing engine, still pulling in 100 new users a month. But 5% churn on 3,000 is 150 people. You’re now shrinking by 50 a month, and marketing didn’t change at all. Nobody did anything wrong. The math just caught up.

The Ceiling Has a Number: Max MRR

Jason’s fix for founders flying blind is a metric he calls Max MRR. It’s the revenue ceiling your business hits once cancellations catch up to acquisition.

Max MRR = New MRR per month ÷ Monthly Churn Rate

Bring in £150k in new MRR a month with 5% churn, and your ceiling is £3 million. You can throw every ad dollar you have at it and you’ll still hit that wall.

This isn’t theoretical. Buffer lived through a hard 5% churn rate for years. Joel Gascoigne told that whole recovery story at BoS too, and you can see the ceiling sitting right there in their graphs the entire time. ConvertKit went the other way: a decade of work to grind churn down to 3%, which is what let them keep climbing instead of stalling out.

Max MRR moves months before your actual revenue chart shows anything wrong. It’s an early warning, if you’re watching it. I didn’t even know this was a number you could calculate before this talk, and now I keep wondering how many teams find out their ceiling exists only after they’ve already hit it.

Your Cancellation Survey Is Lying to You

When revenue dips, the instinct is to dig into the cancellation survey. Don’t bother. Most of those dropdowns are noise. People click whatever’s first just to get past the popup.

Open ended, casual questions work far better. Groove tested this. They switched the automated email subject line to a plain “quick question,” and got a 10% usable response rate. Then they changed the actual question from “Why did you cancel?” to “What made you cancel?” Same idea, softer framing. Response rate doubled to 20%.

But even good feedback has a trap built in. When someone writes “too expensive,” we panic and reach for the pricing page. Jason’s counter is simple. They already agreed the price was fair the day they entered their card details.

“Too expensive” is usually a translation of something else. A value drop that happened weeks earlier. Confusing onboarding, a feature that didn’t work right, a workflow that felt like friction. It wasn’t price. Value stopped showing up. The fix was never in the pricing page. It was in something that happened quietly, weeks before anyone typed “cancel.”

The real fix is catching people on the happy path, the sequence a healthy, sticking around user actually follows. Someone who falls off that path hasn’t decided to leave yet. They’re just lost. Wait until they hit cancel and you’ve already lost the window. Fix the onboarding instead, and it compounds across every cohort after it, because the steepest drop off is always right at the start.

Three Ways Out

These are the three levers Jason points to for breaking the ceiling instead of hitting it.

1. Give and get referrals. Skip the one way cash incentive. Dropbox’s “we both get extra space” reads like a favor between friends, not a commission pitch.
2. Net Revenue Retention above 100%. This is the real lever. Above 100% NRR, upgrades from your existing base outpace what churn takes away, and the business starts compounding on its own. Dropbox has been acquiring fewer new customers year over year and still growing revenue, because the existing base carries it.
3. Show your receipts. Remind customers what you’re doing for them before the invoice lands. A security dashboard that says “blocked 5,000 attacks for you this month” builds the case for you. So when a price increase shows up later, it feels earned instead of random.

The full talk is on the BoS website. Watch it alongside Joel Gascoigne’s companion talk on Buffer’s recovery to see these ideas actually play out.

Jason Cohen

Jason has  built four software startups, both bootstrapped and funded, both alone and with co-founders. All of them grew to more than $1m annual revenue.He sold two, and currently serves as CTO of the fourth, WPEngine, with 380 employees headquartered in Austin, Texas.  More recently, he has also been an angel investor and was a founding member of Capital Factory, an Austin incubator and co-working space. He writes about software and startups at ASmartBear.

More from Jason.

A ChatGPT or Claude subscription someone uses for drafting. Copilot in the codebase. Maybe a chatbot on the website that was set up and then quietly forgotten. A few team members who’ve found prompts they like and use them for specific tasks. A Slack thread somewhere with “useful AI prompts” that nobody’s looked at in three months.

That’s not an operating system. That’s a drawer full of tools.

And the difference between a drawer full of tools and an operating system is roughly the difference between a 1x business and a 5x business.

Tim Barker, former CMO at Salesforce and CEO of a 600-person healthcare company, now running a five-person AI-native startup, spent seven months building the latter from scratch. He came to Business of Software Europe 2026 to explain what it actually involves.

The answer is not more tools. It’s not better prompts. It’s a system. And most businesses don’t have one yet.

What an AI Operating System Actually Is

The analogy Tim reaches for is a new hire.

Imagine bringing on the most brilliant person you’ve ever hired. Extraordinary capability. Across every function. Available instantly, at any hour, at a cost that would have seemed impossible five years ago.

There’s one problem: they wake up every morning with no memory of what happened yesterday. No institutional knowledge. No recollection of the decision made in last Tuesday’s leadership meeting, or the pricing change from three weeks ago, or the brand positioning you’ve been refining for six months.

So how do you manage that person effectively?

You’d do what you do with every high-performing hire. You’d onboard them properly. You’d give them your strategy, your values, your brand standards, your products, the markets you serve and those you don’t. You’d give them a clear role with explicit boundaries. Not vague, not broad, but specific enough that they know exactly what authority they have and when a decision needs to come back to you. You’d give them work to do, review their output, and build a feedback loop that makes their next output better than their last.

“You do all of this today,” Tim told the BoS Europe audience. “The great thing about building an AI-first system is you just need to institute all of these things in a way that AI can use them.”

That’s an AI operating system. It’s not software you buy. It’s organisational infrastructure you build.

The Architecture: Five Layers

Tim’s operating system lives in a folder structure. Five layers, each serving a specific purpose, each feeding into the others.

Strategy is the foundation. Your brand, your values, your products, your priorities for the year, the markets you serve. Everything that probably exists somewhere on a shared drive, probably stale, that every employee should know but most don’t fully. This is your first task: pull it together, refresh what’s out of date, and format it so AI can read it.

“It’s a good opportunity to refresh these things before you start feeding them in,” Tim noted. Most organisations discover that articulating their strategy clearly enough for AI to use it reveals gaps that existed long before AI was in the picture.

Agents is where the job specs live. Each agent has a defined role, clear constraints, and an explicit understanding of what decisions it can make independently and what requires human sign-off. Tim has 14 agents operating today. Some do work: writing, analysis, monitoring. Others are guardians, whose job is to check the work of the agents that produce it.

“Build guardians that check the work of agents,” he said. “You don’t want an agent having growth at any cost. You’ve got to balance it against trust and economics.”

Missions are time-bounded projects: the things you’re actively working on right now, with defined outcomes and deliverables. The output of one mission can become the input to another. Keeping missions discrete keeps token costs manageable and keeps agents focused.

State is your current reality: where the business is today, what decisions have been made, what’s in progress, what’s blocked. This is the layer that most businesses building with AI completely skip. That’s why their AI experience feels like Groundhog Day.

“Context rot is real,” Tim said. “Token windows are only so long. You need to make sure you are always compounding everything back into assets that AI can read every morning.”

Artifacts is the library of approved outputs (content, analysis, frameworks) that feeds back into the system as inputs for future work.

The Layer Most Businesses Are Missing: State

If there’s one thing Tim’s talk made clear, it’s that state management is the difference between AI that learns and AI that resets.

Every leadership meeting at Attain IP is recorded. A few hours after it ends, an agent processes the recording, extracts decisions, and presents Tim with a structured list to review. The ones he approves go into a decision log: a running timeline of every significant call the leadership team has made. That log gets fed back into the context that every agent reads when it starts a new session.

So when an agent wakes up tomorrow, it already knows about the pricing decision from last week, the hiring choice from three weeks ago, the new market focus from last month. It doesn’t need to be told. It reads the state, updates its context, and starts from an informed position rather than a blank slate.

Without this, you’re not building organisational intelligence. You’re re-explaining your business to a very capable but perpetually amnesiac contractor, every single day, from scratch.

Feedback Loops: Where the Compounding Happens

Tools produce outputs. Systems compound learning. The gap between them is feedback loops.

Tim gave a specific example. His agents write content in his voice and in the voices of his co-founders. Every time an agent produces a draft and a revised version gets approved, the system logs the gap between the two, capturing what changed and why. Over time, the gap narrows. The output gets closer to what Tim would have written himself, without Tim having to explain his preferences from scratch each time.

“There is a compounding effect if you just do this on all the assets you produce or draft over time.”

Another example: his LinkedIn monitoring agent delivers ten engagement recommendations every morning at 8am. His co-founder, a patent attorney of 30 years and not particularly active on LinkedIn, reads them, then replies to the agent’s daily email with feedback: which ones she’ll use, which ones missed the mark and why. The next morning, the agent has incorporated that feedback. The recommendations improve. The process costs $0.20 a day.

That’s not a magic prompt. That’s a system that gets better because it has a mechanism for getting better.

The Hard Constraints Your System Needs From Day One

Tim learned this one the expensive way.

Early in building Attain IP, he asked an AI agent with browser access to stub out some LinkedIn ad campaign skeletons, just placeholders he’d optimise and launch properly later. He didn’t specify that the agent wasn’t authorised to spend money. The agent set everything up, switched the campaigns on, and spent $2,000 on traffic to a staging website before Tim noticed three weeks later.

The AI did exactly what it was asked to do. There was no bug. The problem was the absence of a constraint.

“Make it super crisp,” he said of agent instructions. “That’s what’s going to make the whole system efficient and effective.”

Every AI operating system needs hard constraints embedded from the start: what agents can spend, what they can publish, what they can send, what they can change. Not as a brake on capability, but as the governance layer that makes expanding capability trustworthy over time.

The Difference Between a Copilot and an Autopilot

Tim draws a distinction that matters a great deal in practice.

A copilot AI works with you: it produces something, you review it, you approve it, it ships. An autopilot runs without you: it produces, decides, and acts. Both have their place. The question is whether you’ve built enough trust in the system, through enough real-world iteration, to remove yourself from a particular loop.

“Move at the speed of trust,” he said. “I’m not yet at a point where I want agents interacting with prospects and customers directly – until I’ve got their tone of voice to a point where I’m happy with it.”

The path from copilot to autopilot runs through the feedback loop. You review, you feed back, the system improves, you review less frequently, you eventually hand off the loop entirely. That progression is only possible if you built the feedback mechanism in the first place.

Where to Start

Tim’s advice for founders who want to build this is practical and unambiguous.

First: do it yourself. “All of you as leaders cannot delegate this.” Build something: a personal project, a single-function prototype, that forces you to encounter the real problems directly. Not to evaluate AI. To understand it from the inside.

Second: start with your strategy layer. Pull together the documents that define your organisation: brand, values, positioning, priorities. Ask honestly whether they’re current enough to be useful. This is foundational. Everything else builds on it.

Third: define one agent’s role precisely. Not broadly (“help with marketing”) but specifically (“monitor LinkedIn conversations in this domain, score them against these criteria, surface ten recommendations each morning, send to this email address”). Crisp role definition is the hardest part and the most important part.

Fourth: build the state layer before you need it. The decision log, the session summaries, the current priorities. Set up the infrastructure to capture this before there’s urgent content to capture. The habit is the thing.

Build Your AI Company Operating System

Tim Barker walks through the full architecture in the Business of Software workshop specifically designed for this: strategy layer, agents, guardians, missions, state management, and feedback loops.

→ Build Your AI Company Operating System

Tim Barker spoke at Business of Software Europe 2026. His talk drew on seven months of building Attain IP, an AI-native legal tech business, after leaving the CEO role at a 600-person digital healthcare company.