The 2026 SaaS reset 

The 2026 SaaS reset is hitting sales-led software companies from three directions at once: their moat, their margins, and their operating model. This essay walks through what’s actually shifting — and what more than a dozen SaaS operators are doing about it. Read it in order, or use the contents to start where the reset is hitting you hardest.

Contents

1. How the 2026 SaaS reset actually starts — Why the reset is being caused, not suffered.
2. What you reach for when you can’t start over — The first-order trap most incumbents fall into.
3. Why second-order work doesn’t happen by itself — The comfortable stories that keep you from the real move.
4. Why segmentation isn’t niching — Why “picking an industry” was never the same as going deep.
5. Why faster teams don’t make a faster company — The difference between team speed and company velocity.
6. Why amplification beats automation — Two leverage points, and the one most CEOs get backwards.
7. Why build decisions got harder when building got cheaper — Applying the differentiation filter when everything is buildable.
8. Three dimensions of differentiation — Auditing your moat across relevance, defensibility, and durability.
9. What AI does to your P&L — The three financial pressures hitting at once.
10. Why the first step isn’t a tool — How to start when you’re late, and why it’s more familiar than it sounds.
11. Closing reflection: The filter — Where you stand, and how to find out honestly.

Tim Barker walks on stage at Business of Software Europe. He built two SaaS businesses the traditional way (ex-CEO of DataSift, acquired by Meltwater; ex-CEO of Kooth). Today, he runs Attain IP with five humans and seven teams. The humans are the exec layer. The teams are ranges of AI agents.

I sit there in the dark, watching his slides, and it hit me differently than I expected.

Everyone is writing about the SaaSapocalypse. Stock market valuations back to 2018 levels. Layoffs across every tier. “SaaS is dead” takes everywhere. And we keep framing it like something happening to SaaS companies. Like the weather. Like a correction they have to survive.

But Tim isn’t surviving it.

He is causing it.

The reset isn’t happening to the market. It’s being caused by people who deliberately chose not to run the old playbook anymore.

Here’s what stuck with me as his slides unfolded. I could see him stacking traits from the Remarkable Effect, one on top of the other. Deliberate clarity about who he serves (#1). Something valuable and desirable — outcome-as-a- service, not software-as-a-service (#2). Different, not just better — nobody else is structured this way in his space (#3). Curiosity, JTBD-first, questioning every assumption about who should do the work (#4). New value possibilities that simply didn’t exist before (#5). And at the centre of it all, trait 7 — focus on the essence. Resourcefulness at a level most companies can’t imagine until they see it.

That combination doesn’t just win on five dimensions. It raises the bar on the eighth — momentum — to a level nobody running a 2025-shaped SaaS business can match. Not fast enough. Not profitable enough. Not even close.

And Tim isn’t the only one. There will be more. In every category, some already started. You just don’t see them yet because they’re still small enough to ignore.

The question isn’t whether the reset is real. The question is whether — when the AI-native version of your company shows up in your market — what you have left is still enough.

Here’s a test. This week, pick one workflow in your business and ask honestly: if I started this company today from scratch, would a human own this step — or would an agent?

Question for you to reflect upon:

Tim’s version of your company is being built right now, in your category, by someone you haven’t heard of yet. What are you doing this quarter that assumes they don’t exist?

The Remarkable Effect is when the way you run your business today is a deliberate choice, not a legacy.

The honest reaction I keep hearing is some version of the same thing: OK, but I can’t start over. I have eighty employees, technical debt, a live customer base, shareholders. What’s the actual move?

It’s a fair question. But the answer is an uncomfortable one.

The move most incumbents typically make — one that will quietly reset them anyway — is what I’ve come to recognize, across every technology wave I’ve watched, as the first-order response.

Every reset has two responses.

1. The first answers the question on the surface — what customers and investors are asking right now. It’s defensible, feels like progress, and almost always has an end date.
2. The second answers the shift underneath. It’s harder to explain. It compounds.

Call them first-order and second-order responses if you like. Most CEOs reach for the first.

Just look at the cloud transition. A lot of software companies at the time comforted themselves with ASP — application service providing. Customers wanted to stop running software on-prem, so we’ll host it for them. Box ticked. But cloud wasn’t about hosting. It changed how software got built, how it got priced, what scale even meant. ASP answered the visible question and missed the shift underneath. Most of those companies don’t exist anymore.

Now look at 2026. The most common first-order response is “we’ve added AI to our product”. A chatbot. A summarise button. A Copilot in the sidebar. It answers the literal question.

It’s almost always insufficient — for the same reason ASP was. Because AI isn’t a feature shift. It’s a shift in what the business fundamentally is. What work should even exist. How the company scales without linear headcount. What your moat is when the interface is no longer defensible.

Adding AI to your product ticks the box, but it doesn’t reset the thinking.

A second-order response asks a different question: what work should humans still be doing in our business, now that we have a real choice between AI and people — and what would we build, price, and sell differently because of it?

The hard part about the first-order response is that it feels like the responsible thing to do. Nobody loses their job for shipping a Copilot. But you lose your company — slowly, then all at once — for mistaking the Copilot for the response.

Here’s a test. This week, look at your last three strategic decisions and ask honestly: Which question was I actually answering — the literal one, or the shift underneath? If all three were ‘the literal one’, you know what to do.

Question for you to reflect upon:

Look at the three biggest AI moves you made this year. How many were first-order responses — and how many actually reset how your business works?

The Remarkable Effect is when you stop protecting what the reset is replacing.

Mark de Jong has been running Timewax for fifteen years. Long enough to have lived through the cloud transition, mobile, the social wave, and now this.

“AI native or AI first is an overdone term. The history of technological development has always been about adopting technology to deliver better, faster, cheaper, or something customers couldn’t get before. An AI-native company without that focus has no customers.”

That’s the right starting point. And what Mark is doing is exactly what he describes. On the customer side, he’s using AI to solve problems his product couldn’t solve before — innovations his customers genuinely couldn’t get from him a year ago.

On the organizational side, no one is waiting or hiding either. People take ownership because they can. One example comes from his lead Customer Success manager, who built a complete internal system in Claude Code. Customer health monitoring. Upsell signals. Churn risk flags. Full user management. SSO. Built over a few weekends, by one person who’d never written production code. A working system that previously would have needed a small engineering team and a quarter of roadmap time.

Mark didn’t authorize a project. He didn’t budget it. The leverage is now sitting on every desk in the company, waiting for someone willing to use it.

That’s two sides moving in lockstep — customer-facing and organizational — because they keep asking what’s best long term.

So why don’t more CEOs do this?

Because second-order work is hard. And complacency doesn’t feel like complacency from the inside.

In my book, I named this directly: ‘No SaaS company has ever died because of a competitor. They die because they tell themselves comfortable stories and believe them.’ Two of those stories are doing a lot of work right now.

1. “That enhancement would hurt our existing business.” The cannibalization story. You see the second-order move clearly, but you also see it would compress per-seat revenue or sunset a product line still paying salaries. So… you don’t do it. You protect what’s working today and call it discipline. It isn’t. It’s an invitation for someone else to do the cannibalization for you.
2. “We’re already on it.” The chatbot is shipped. There’s a summarise feature on the roadmap. There’s a slide in your board deck with “AI strategy” written across the top. Software companies betting on ASP back in 2008 told themselves the same story. The activity felt real, but the action wasn’t. The rest is history.

Both stories sound responsible. But neither is. The hard part is spotting them in your own thinking.

Here’s a test. Take twenty minutes this week and run through the Remarkable Index. Ten traits, thirty questions. The traits where you’d score lower today than a year ago are where complacency is doing its work.

Question for you to reflect upon:

Which of these two stories are you most likely to tell yourself this quarter — and what would change if you stopped?

The Remarkable Effect is when your business moves on what’s true, not what’s comfortable.

Spotting the stories is step one. The harder question is what you do with that clarity — starting with where you’re actually choosing to compete.

The word “niche” is one of its better disguises. Most SaaS companies haven’t actually niched. They’ve picked a size, a region, or an industry and called it that.

For most, that decision was the work. Stay disciplined, be one of the acceptable options, and you’d close your share of deals. Customers always had alternatives — most of them fine, not great, not terrible. They picked one because they had to pick something. That dynamic let a lot of software companies survive without ever being remarkable.

That doesn’t work any longer.

A new kind of vendor is showing up in your space. AI-native, deliberate about who they’re for and who they’re not for, compounding fast. They’re not smarter than you. They’re just refusing to serve customers you’d take without hesitation — and using that refusal to go deeper than was possible until recently.

When they enter the deal, the workable alternatives are exposed for what they always were. The customer sees that gap, and the bar lifts in real time.

Scott Reynolds, CEO of UpCodes, said something worth reflecting on when we spoke. Ten years building software for the construction industry — specifically, the regulatory side:

“If you can only make a tool 10% or 20% better, just don’t. You can’t pull people off their old tool to a new tool. You’re really looking for that 500%, 2,000%.”

That’s not a 2026 framing. That’s how he ran the company from the start. UpCodes’ moat now compounds with AI rather than being eroded by it — because the depth was already there.

Tal Peretz, CEO of Onfire, took a different starting position. Three-year-old company, AI-native by design. He picked the hardest buyer in tech — CTOs, CISOs, technical leaders who actively resist sales. Why? Because he wanted to be remarkable on day one — not earn it over a decade. Generic AI gives generic output. Niche-deep AI with proprietary context gives outcomes nobody else can match. Different starting position from Scott. Same discipline. Both raise the bar for everyone in their categories.

If your niching produced a real why-us, it lives in three places: your product, the deals you win, and how your customers describe you to peers. If it’s there, keep deepening. If it’s missing in any one of them, the niche didn’t go deep enough.

A way to test it is this:

Imagine your three most loyal customers are evaluating today, fresh, against current alternatives. Would they still pick you? If you’re not sure — or worse, certain they wouldn’t — your niche didn’t produce a why-us strong enough for 2026.

Question for you to reflect upon:

Your customers are seeing the new alternatives. What’s the specific reason they haven’t moved yet — and is it durable?

The Remarkable Effect is when your customers can’t imagine going to anyone else for what you do

Going deep enough that your customers can’t imagine alternatives is one half of the work. The other half is whether your business can actually move at the pace this depth requires.

Here’s what most CEOs are seeing right now. Their teams are faster than they’ve ever been. Engineering ships more. Marketing produces more. Support closes faster. Each team’s dashboard looks great. And yet the business itself doesn’t feel like it’s moving any faster than a year ago.

Here’s why: Faster teams and a faster company aren’t the same thing.

Team speed is what your individual functions can produce. Company velocity is the speed at which the whole business actually changes shape. Team speed is everywhere now, yet company velocity is rare. The companies that have both compound and the ones that just have team speed get busier without really getting anywhere.

The thing that turns one into the other is autonomy. Not faster meetings. Not better project management. Real autonomy — your team makes calls inside their lane without checking back in.

Most CEOs say they want this, but few actually run it. Because real autonomy needs two things, and one of them is uncomfortable.

The first is the CEO’s guts to let go. To trust the team. To leave them to it. To only show up when they need support — not to ask what to do. That’s the part most CEOs don’t actually do, even when they talk about it.

The second is clear principles people can use to make day-to-day decisions. Without those, you get one of two things. Chaos, because everyone is making different calls. Or the team plays it safe and pushes everything back up to you anyway.

Christian Owens, CEO of Paddle, runs this with two simple mission principles: ‘We want to do it for you’ and ‘Be the most helpful brand in SaaS’. That’s it. As he puts it: “It’s a framework for how we make decisions. Does this advance us doing it for you? Does this help us become the most helpful brand in SaaS? If the answer to both of those questions is yes, we should do it.” Two lines. Four hundred people aligned. No need to escalate every call back to the centre.

Patrik Berglund, CEO of Xeneta, runs a similar approach with a different texture. He’s been clear with his team about what outcomes he wants — more data into the platform, more adoption, more retention — and then stepped out of the how. The team picks the route.

That same team breaks down the AI question against five lenses:

1. Internal efficiencies — what to add to what they already have
2. Complete redo — what to rebuild from scratch
3. What to do now versus reassess in a month, because the ground keeps moving
4. Costs and dependencies — what to track so the bets don’t compound badly
5. Enablement — making sure each person stays current, beyond just what Xeneta needs today

Five categories. Each one answers a different question. Together they reduce the chaos so the team isn’t relitigating the same decisions every week. Patrik doesn’t need to be in every conversation about which AI bet to make. The framework holds the line.

Question for you to reflect upon:

Walk through your last three strategic decisions. Could anyone on your leadership team have made the same call without checking with you? If no, the bottleneck isn’t speed.

The Remarkable Effect is when faster teams make a faster company, not just a busier one.

Company velocity is what gets you running at the rhythm the reset demands. The next question is what that velocity is for. The answer: leverage around what sets you apart.

The score that captures it is ARR per employee. The benchmark has moved. What used to be respectable is now mid-tier. The CEOs running ahead aren’t working harder. They’re running leverage on two points instead of one.

The first leverage point is who you serve. Niche sharply. Refuse the wrong customers. Sell to the ones who pay a premium because what you do is unmistakable to them. The effect: Sales cycles shorten, implementations run more efficiently, support costs drop, and churn shrinks. The business runs lean by structural design. Mark de Jong at Timewax sharpened this recently. His result, in his words: “Highest revenue ever — with our smallest RevOps team.”

The second leverage point is how you work. And most CEOs are getting this backwards. It’s not about automating what you don’t like. Not about automating the boring tasks. Not about freeing yourself up to focus on what matters.

The real leverage runs the other direction. Take what your team uniquely does — the craft, the judgement, the tribal knowledge — and find ways to amplify it. Every team. Engineering, sales, marketing, services, support. Each one has work that makes them remarkable.

AI is the tool that makes amplification reachable now in ways it wasn’t before.

Not automation of weakness. Amplification of strength.

Amos Bar Joseph, CEO of Swan, has built his company around this from day one. His core metric: ARR per employee. His north star value: $10M.

He calls his model ‘the autonomous business.’ But don’t misread this. It’s not a business that runs automatically. It’s a business where each person operates with maximum autonomy and pulls leverage on their unique strengths from the AI around them. Same logic Patrik runs at Xeneta. Different vocabulary.

Dimitri Masin, CEO of Gradient Labs, runs this at the level of the work itself. He calls it taste — the intuition for what will work, what a customer will actually value. It’s why he can skip rounds of research and alignment: the judgement is already there, and AI lets him act on it faster. His team built what they call a looping channel — a shared catalogue of Claude Code skills used across marketing, sales, support, engineering, not just developers. Anything done more than twice gets templated and shared. Everyone amplifying their own craft through the same system. The skills compound across the company.

Mind you: amplification doesn’t differentiate by itself. It just speeds up what you were already doing. So the question is what you amplify. Focus on the work that’s genuinely valuable AND desirable AND different — your unique craft, not the table stakes — and the gap with competitors widens.

Focus on the rest, and you’re just adding to the noise. And that noise is real. Cuvama’s research across 150+ revenue leaders found the same pattern — more AI tools, worse conversations. As Tom Canning, ex-CCO of Zellis, put it: “We’ve given our teams more AI tools than ever. And somehow, the quality of early-stage conversations has gotten worse.” Generic AI produces volume without substance. Faster doesn’t mean noticed.

Question for you to reflect upon:

If an AI-tooled competitor reproduced your team’s last week of work — which parts would they get right, and which parts couldn’t they touch? What if you’d focus AI there?

The Remarkable Effect is when amplification widens the gap, not the volume.

Amplifying what makes your team uniquely valuable is one half of the work. The next question runs one level up. What should you build in the first place — and what shouldn’t you touch?

The reset opened up the can-do side dramatically. Vibe-coding made building cheap. AI made automation tempting on every front. The default reflex is we can do more now, so let’s do more.

But the old rule still applies: Does it add to your differentiation?

Just because you can build it doesn’t mean you should. And just because a customer can now vibe-code their own version doesn’t mean they will — not when what you do is mission-critical enough that customers want a real vendor on the other end, not their own code.

The filter hasn’t changed. The half-life of what counts as differentiation has, though. Because what used to be valuable for years is now valuable for months. And this means the filter has to be applied more honestly and more often.

Tom Dehouck, COO at Blinqx, runs this lens at group level. Blinqx is a SaaS group with €120M revenue, growing through a mix of organic build and vertical acquisitions.

What stands out is how their thinking has shifted. The deep work now is understanding usage — what customers actually want to accomplish with the software, vertical by vertical, and which metric behind that usage matters most. That changed their hiring lens. It changed the ratio between product people and developers from 1-to-5 to 1-to-1. AI made developers more leveraged, so the bottleneck moved to product judgement.

At the group level, Blinqx is building what no single vertical company in it could — a platform that orchestrates AI agents across the entire group. And the best practices that emerge from each vertical get actively shared across every department, not just engineering. That’s the horizontal layer the vertical depth plugs into.

Symplicate, the CRM for creative agencies inside the Blinqx group, illustrates what this produces. They’re about to release a feature that lets agencies monetize their virtual agents. Two years ago that need didn’t exist and it’s irrelevant in the context of horizontal CRM. Go deep in a vertical and have the right structure underneath — and every release widens the gap.

Resist the temptation to build everything just because you can. Instead, apply the differentiation filter: does this solve a valuable problem, is it critical enough that customers can’t ignore it, and can you exceed expectations doing it?

Question for you to reflect upon:

Look at your roadmap. Which items genuinely deepen your differentiation, and which are there because building got cheap? What would change if you only built the first kind?

The Remarkable Effect is when every release widens the gap.

Building only what deepens the moat is the discipline. The next question is whether your moat actually holds — and that comes down to three traits firing simultaneously.

The 2026 reset is changing everything at a pace few imagined. That’s not necessarily a problem. It’s a chance.

The foundation of a software business hasn’t changed. Differentiation has always been at the core — and it will remain to be. My definition: 

Differentiation is what sets your SaaS business apart, cannot be easily replicated, and keeps the right people talking about it.

Three things in that sentence. Three dimensions of any real moat. Each one scored 1 to 10.

A year ago, you probably scored all three at 8s or 9s without thinking about it. But run the audit honestly today, and the numbers are very likely to look different. Not because you got worse, but because the bar moved on all three at once.

Whichever dimension dropped hardest — relevance, defensibility, or durability — the arsenal to correct it is wider than most CEOs think. More on that in my essay on the 9 principles of differentiation.

Dimitri Masin, CEO of Gradient Labs, sees the same shift from another angle. He’s worried about narrow products — single-purpose (or best-of-breed) tools that used to take years to build. His words: “I’m struggling to see how companies will succeed with niche products in the future. Because those products are relatively easy to get off the ground today — I could get most of those products off the ground in two months.” Two months for what used to take two years. The narrow product that used to be hard to copy is now one of ten things a platform player can absorb without breaking stride. That’s defensibility under pressure.

Martin Balaam, CEO of Pimberly, said it best. Anthropic turned up around his moat with a fleet of trucks, and suddenly anyone could walk across it. “So you’re like, okay, what do I do with my moat now?” And this isn’t a startup founder talking. It’s the CEO of an established enterprise SaaS business.

But not all defensibility erodes at the same speed. Hessel Kuik, CEO of Minox, points to the other side: some categories hold because the work itself demands something AI can’t reliably deliver yet. In his case — accounting — deterministic, auditable data processing. He even quotes Claude on its own limits: “I cannot formally prove how I arrive at a result in a way that is externally verifiable… that is not a matter of version number, but of architecture.” His point is honest, though: that edge is temporary. It’s the same shape of argument on-prem vendors made about cloud around 2008 — true at the time, until it wasn’t, because the disruptor took a form nobody predicted. Hessel knows it. The audit discipline still applies: when does the line move, and what’s the signal that would flip your answer?

Durability gets hit the same way. The operating model that kept you ahead through annual planning cycles can’t keep pace with competitors recalibrating monthly. The operators I’ve written about across this series have each named a different durability move

• Patrik Berglund at Xeneta empowering his team to make daily decisions without checking in
• Amos Bar Joseph at Swan running the autonomous business — each person with maximum autonomy, using AI not to automate – but to amplify their unique strengths
• Tom Dehouck at Blinqx shifting the ratio between product people and developers from 1-to-5 to 1-to-1

Different moves, but all in the same direction. Running the audit honestly forces you to rethink not just what you build. But how you’re set up to keep building it.

Relevance is the third pressure point and the one most companies dodge. The customer’s world shifted. What they were willing to pay a premium for last year may be table stakes now. The ‘why you’ doesn’t last forever. It shifts when the customer’s world shifts.

Question for you to reflect upon:

Which of the three dimensions — relevance, defensibility, or durability — dropped hardest on your moat in the last twelve months?

The Remarkable Effect is when your moat gets stronger as the world changes.

Three dimensions of differentiation tell you where the moat is under pressure. But the pressure shows up somewhere else first — in your numbers.

The reset is more than a technology attack. Strangely enough, that’s possibly the smallest part of the reset. The real attack is how it’s fundamentally changing the economics of a business. 

AI is doing three other things to your P&L at the same time.

Take pricing first. Martin Balaam at Pimberly put it bluntly at our virtual espresso call: “If your primary charge is per seat, you need to sort your shit out.” He’s not wrong. Most of us know that other models are coming. But that’s not the question. The question is what happens to your current revenue when AI cuts the seats out from under it.

Picture your largest account today. They pay you for 450 seats. Imagine they cut it to 45 tomorrow. You’d lose ninety percent of that contract. Multiply across your top ten accounts…

And it’s not waiting for you to react. MG Gurbaxani, CEO of Cuvama, sees it directly: their customers are already building their own value agents. If Cuvama isn’t there — present in the workflows their buyers are building — they go invisible. The wider point: your customers are committing to new stacks right now. They might keep your product. But the users won’t look anything like they did a year ago.

Roy Wood, CRO at OneAdvanced, has run exactly that audit — to determine what percentage of the business is exposed because of seat pricing, and where it’ll hit first. He’s using that insight to decide what the commercial model needs to be instead, and how to mitigate the risk.

Then there’s commoditization, because building has gotten cheap. What used to take two years now ships in two months, as Dmitri Masin put it. More players, more lookalikes, more pressure on what was differentiated a year ago. The consequence: your pricing power drops. And that’s an immediate threat to your top line. 

Lastly, your cost base. As we’ve seen, AI-native competitors do what you do with fewer people. Amos Bar Joseph at Swan has set the bar toward $10M ARR per employee. Most SaaS businesses today sit somewhere between $150K and $400K, so the bar just moved an order of magnitude. What this means is that, for many of us, the cost base will start to work against us as it eats into our room to move — what you can invest in, what you can say yes to, what you can build next.

The three pressures on your P&L

These are just three pressures – but they all work together. The dangerous part of it is that this hits you silently, because we’re all looking the other way towards the technology impact on our product.

There’s a flip side, though — and it’s easy to miss while you’re staring at the pressure. Patrick Oehler, CEO of Retorio, points out where the leverage went. The same VC money that used to dilute SaaS founders now arrives indirectly — as subsidized compute. LLM providers are selling tokens at a loss, burning billions to lock in share. As a builder, you benefit from that cross-subsidy without giving up a point of equity. In his words: OpenAI and Anthropic are doing the fundraising for the rest of us. The cost base is under pressure, yes. But the cost of building on top of it has never been lower — if you’re set up to use it.

Question for you to reflect upon:

Take your last three quarters. Where are your margins moving — and which of the three pressures is doing it?

The Remarkable Effect is when your economics keep working as the world resets.

Three pressures, one shift. The question is what you do about it. And the answer starts with one team, one workflow.

Being late to the reset isn’t the problem. Not starting is. Especially if you’re running a software business that’s ten or fifteen years old, built for a different era — that’s exactly when the question lands hardest.

Over the past weeks, I’ve talked to many people in the middle of this. Let me zoom in on one, because it shows the starting point is more familiar and practical than it sounds

Tim Barker spent two decades building traditional SaaS companies. He scaled them and then exited, the path we all know. Now he’s building Attain IP with five people and seven teams. Not seven teams of people — seven teams of AI agents, each supervised by a person. It sounds radical until he describes how it works. He laid out the strategy and the North Star, wrote crisp job specs with clear decision boundaries, defined the missions and their outputs, and built feedback loops across all of it. “All the things I learned building a 600-person software company”, he told me. 

The operating model isn’t new. It’s what a well-run company already does — he just applied it to agents instead of only people. Some work runs on autopilot. Some you co-pilot and review. Exactly how you’d manage a team of people you’re still getting to know.

One thing, though, decides whether this works. Tim calls it moving at the speed of trust. Push it too fast — hand over work before the trust is there — and a mistake doesn’t stay small. It cascades, fast and wide. The discipline isn’t using agents. It’s being honest about how much to trust them, and when.

That’s the part worth sitting with. The AI-native operating model isn’t a new discipline you have to learn from scratch. It’s the discipline you already know, pointed at a new kind of worker.

And it isn’t only for people starting clean. Ron Gidron took xType, the company he founded 6 years ago, AI-native across the whole business, not just the product. Engineering is now fully agentic and runs 5-10x faster cycle speed without adding headcount. Then the same approach was extended into customer success and G&A. He didn’t start over. He started.

So how do you start? 

The first step isn’t a tool; it’s writing the job spec you never made explicit — for one role, one workflow — so clearly that an agent, or a person, could run it. You can do that today, without touching any technology.

That’s the move. You don’t need a clean slate. You don’t need to be early. You need to start with one team, one workflow, one honest experiment, and apply the discipline you already have.

Question for you to reflect upon:

Take one team. Can you write down — on a single page — what the outcome should be, where the boundaries are, and how you’d know the work is good? If you can’t, that’s where the work starts.

The Remarkable Effect is when you stop reacting to the reset and start setting the pace.

We’ve now examined every shift, and the pattern is hard to miss. From how building got cheap to how the moat is under pressure to how the P&L is squeezed. How we read it is up to us. As Patrick Oehler, CEO of Retorio, put it: “The reset isn’t a headwind. It’s a filter.” Every change is. It does something to your business, or it doesn’t. You do something with it, or you don’t.

Which raises the question: What separates the companies that pass the filter from the ones that don’t? That’s where the Remarkable Index in my book comes in handy.

Ten traits that define a remarkable software company. It’s not a checklist, but a reflective filter. It’s designed to give you clarity on which traits have had their bars moved the most by the reset in your business. And, more importantly, where you have the most leverage right now, and where you’ve lost it.

Take three of them. Score yourself, honestly, on a 1-10:

— Would your customers care if you ceased to exist tomorrow?

— How difficult would it be for a competitor to match the unique value you deliver today?

— What’s your ability to attract an audience that’s prepared to pay a premium?

Just reflect on this. Score yourself against today’s reality — but honestly. Because what scored an 8 two years ago is likely a 5 today in 2026. Sure, the questions are the same, but the gravity is completely different.

Gravity is something you can’t measure by yourself, though. Because one view is no view. You’re biased — every CEO is. The Index only gets sharp when your team scores it next to you. This is where gaps show, and that’s where the real value hides for your company.

Interested? You can always ask me for the team version of the Remarkable Index.

Question for you to reflect upon:

On which of the three above are you most afraid to ask your team what they think? That’s the one to look at first.

The Remarkable Effect is when you keep winning while others fall behind.