The way companies get built has changed, permanently, and the founder is now the deciding variable. Not the market. Not the capital. Not the technology, which is the part everyone is focused on. The founder. You.

The rate of change over the past few years has been relentless and the velocity is increasing. Everyone and every industry is impacted, but the impact is most acute in startups and small to medium-sized businesses. Some of the changes are evolutionary and others revolutionary. I have almost three decades of experience, and as a repeat founder, my observation is this: being a founder today is massively different. While so much has stayed the same, everything about starting, building, and scaling businesses has fundamentally changed. We will examine this theory throughout the rest of the book, but more importantly, we will share how you can execute successfully in this environment with the right perspective, system, and support.

For roughly the last twenty-five years there was a path, and everyone knew its shape. You started small and scrappy. You did everything yourself because there was no one else to do it. As revenue grew you added people, because adding capability meant adding headcount. You hired specialists because the skills were scarce and expensive: someone who could build the product, someone who could read the data, someone who could run the systems. You matured in stages, and so did your company, and the two of you climbed together, slowly, toward the point where the business could finally run on an organization instead of on you. It was a ladder. It took years by design, and the years were the tuition.

That ladder is gone. Not weakened. Gone. Before I tell you what replaced it, I want to show you what the top of it used to look like, because I had the strange luck of standing next to it early in my career.

What Exceptional Used to Cost

Seventeen years ago I took my first job in a startup. I was thirty-two, fresh out of my MBA, and I had been in banking since I was eighteen, where I had worked my way up to a vice president title with authority over daily operations. My mentor at the time, the chief technology officer of one of the big consulting firms and an angel investor on the side, looked at all of that and told me the truth: you have done well here, and you do not know anything about anything yet. Go work for one of the companies I fund.

The company he sent me to was run by a man I will call Ken. It was Ken’s second company. His first had sold for nine figures. This one was a shipping aggregator: software that watched live rates across every major carrier, picked the least expensive option that met a retailer’s requirements, printed the label, and notified the carrier and the customer, all automatically, at a time when the retailers using it had been doing that comparison by hand, order by order, or not at all. I was hired to stand up the financial and internal operations, and I did, and along the way I got a front-row seat to something I did not have the vocabulary for at the time: a professional founder operating at full discipline.

Ken had spotted a gap in the market that others had walked past. He prototyped fast, tested what he built, and got it in front of customers while his competitors were still holding meetings. He took outside funding even though he could have written the checks himself, and when I asked him why, his answer stayed with me for two decades: he did not want to play with house money. He wanted the stakes to be real, because real stakes keep you honest.

But the thing that actually set Ken apart was not vision or speed. It was how he ran himself. Ken was religious about his priorities and ruthless about his time, and he expected the same of everyone around him. You did not bring him something you had the authority to handle yourself, and if what you brought him did not rise to the priority of that minute, it waited. It did not matter what it was. It waited. That sounds cold on the page. In the building it was the opposite: everyone was focused, everyone was empowered, and everyone kept moving, because the discipline at the top made the priorities legible all the way down.

I did my job well enough to eliminate it. Within a few months the operations I had been hired to run were standing up and largely automated, and Ken, reasonably, wanted to move me onto technical client work I was not equipped for and did not want. We parted ways quickly and without drama. He went on selling companies. As far as I know he is on his fifth, and every one of them has exited. In almost twenty years of watching founders since, I have never seen anyone run themselves and a business with that combination of consistency, focus, and follow-through.

Hold that picture. A founder like Ken, operating at that level, needed capital, a funded team of specialists, scarce technical skill bought at market rates, and a personal operating discipline so rare that I have effectively seen it once. That is what exceptional cost a founder seventeen years ago.

What Just Changed

Now look at what has happened to that list.

The research capability Ken paid analysts for is available to you tonight, for the price of a subscription, at a depth and speed no analyst team could match. The synthesis, the modeling, the drafting, the automation of routine work: the same. A working software prototype used to be a funded team’s quarter. It is now a founder’s focused week. The specialized skills that justified entire departments, the ability to query data, to build integrations, to produce professional-grade material in any discipline, are being dissolved into tools that anyone can operate. This is not a prediction about where AI is going. This is an inventory of what agentic tools already do, and every month the inventory grows.

This claim, unlike the folklore in this genre, comes with numbers. Stanford’s AI Index, the standard annual accounting of the field, measured what it costs to run an AI at the level that stunned the world when ChatGPT launched: twenty dollars per million tokens in November 2022, seven cents by October 2024. A drop of more than 280-fold in eighteen months, with prices across AI tasks falling anywhere from nine to nine hundred times per year, while organizational adoption jumped from 55 percent to 78 percent in a single year. The capability did not get slightly cheaper. It effectively stopped being a cost.

Watch what that does to team size and capital. In 2025, Y Combinator, the accelerator behind Airbnb, Stripe, and Dropbox, reported that for a quarter of its newest batch of startups, 95 percent of the code had been written by AI, and that the batch was the fastest growing and most profitable in the fund’s history, with companies reaching ten million dollars in revenue with teams of fewer than ten people. YC’s own chief executive drew the conclusion for founders: you no longer need fifty or a hundred engineers, and you no longer have to raise as much, because the capital goes much longer. The venture world’s most famous filter, an institution built to allocate scarce capability, is now publishing evidence that the capability is no longer scarce.

The academics have now run the experiment properly. A joint Harvard and INSEAD research program studying AI-native companies found that startups built around AI operate about 25 percent smaller than comparable firms, with meaningfully fewer entry-level workers and managers on the chart. When the same researchers put 515 high-growth startups through an accelerator and taught half of them how AI-native companies reorganize the work itself, the taught firms grew faster while their demand for outside capital fell by nearly 40 percent, and their demand for labor did not rise at all. Read that the way a founder should: faster growth, on dramatically less money, without adding people, produced not by talent or luck but by a teachable way of organizing the work. The compression is real, it is measured, and it is exactly what the word on the cover of this book is claiming.

Here is the finding that tells you where the pressure goes next. When researchers from Harvard, MIT, and Wharton ran a field experiment with 758 consultants at Boston Consulting Group, the consultants using AI produced roughly 40 percent higher quality work, 25 percent faster. But the detail that matters most for this book is who gained: the bottom half of performers improved by 43 percent, while the top performers improved by 17. The tools compress the technical middle of the skill curve. When routine competence is available to everyone, what it used to signal disappears, and the differentiation moves to the things the tools do not supply: judgment, honest evaluation, the discipline to finish.

By the time you read this, every number in the last three paragraphs will be out of date, and it will be out of date in the direction of my argument. That is the point. The trend is not a wave to ride. It is the new water level.

I did not learn any of this from the reports. At Seagen, years before the current wave, I watched the compression begin from the inside. When I arrived there was no meaningful AI at work anywhere we looked, so we went hunting for the constraint where intelligence would matter most. We found it in pathology: images from live cancer-drug research piling up faster than the pathologists reviewing them could move, a months-long backlog quietly throttling the pace of research itself. We dedicated an engineer, partnered with outside machine-learning teams, and trained a system to pre-sort the images so the pathologists spent their scarce attention on the most promising candidates first. The backlog fell from months to days and weeks, and the development engine behind it sped up accordingly. Later, when large language models arrived, we stood up one of the earliest private, internal GPT systems I know of, built on the company's own knowledge, so people could finally search across a siloed organization. We piloted AI support agents on frontline IT tickets as one of the earliest customers of that kind of platform, and watched the queue and the staffing it demanded shrink. None of that made us an AI company. It made us a company that aimed the new tools at real constraints. That distinction is the whole game, and it is teachable.

There is a word for what happens when scarce, expensive capabilities become cheap and universal: democratization. And democratization always does the same thing to a market. It moves the differentiation somewhere else.

Because the speed cuts both ways. The same forces that put enterprise-grade capability on your desk put it on everyone else’s. The environment is moving faster than the old ladder was built for, and the ladder’s great hidden assumption, that time was on your side, that you could mature into readiness at your own pace, no longer holds. Standing still now moves you backward. In periods of tumult, and this is one, companies do not merely fail to grow. They regress. They lose what they have earned. They get passed, undercut, and sometimes taken. The founders who treat this moment as weather, something to wait out, are the ones it will happen to.

So here is the claim, stated precisely. The revolution is not AI. The tools are the occasion, not the event. The revolution is that enterprise-grade execution capability just became founder-sized, and the moment it did, the deciding variable in company-building stopped being access to capability and became the human being directing it. What took Ken outside capital, a staffed team, and rare bought skill is now available to a founder of two. Everything, that is, except the one thing on Ken’s list that was never for sale: the discipline. That part you still have to bring. That part is what this book teaches.

One more thing about the word, because I chose it for both of its meanings. A revolution is an overthrow: the old path, deposed. But a revolution is also, in the older sense, one full turn of a wheel. Keep both meanings in your pocket. By the end of this book, the second one will matter as much as the first.

The Differentiator Inverts

When capability becomes table stakes, what remains scarce is the person running it. I will show you how literal that is, because I learned it in a single meeting.

In 2017 I was deep in the blockchain world, taken by the technology rather than the token mania around it, and a former student of mine, twenty-two years old and just out of my capstone course, had co-founded a company with his father. The father is worth describing: a brilliant technical architect, a Microsoft alum, a serial co-founder with an exit behind him, the kind measured in eight and nine figures. Between them they had funded a team, a real product in development, and every credential a company could want in the room. They brought me in two days a week as a part-time chief strategy officer because they were stuck, and one of the places they were stuck was fundraising. They had been meeting with investors for months. Meeting after meeting, all of them warm, none of them concluding.

I asked what was limiting. The answer, once we dug it out, was remarkable: they wanted the investors to offer. They did not want to make the ask.

Within my first week I sat in on one of these meetings. The investor was engaged, the update went well, and near the end he asked the question he had probably asked at the end of every meeting: so what can we do for you? Because I was new enough to have no habit of silence, I answered him. I said we could use an investment, that a capital injection was the thing that would make the biggest difference to everything else they had just heard. The investor said: sure. How much do you need? The founder blinked, recovered, and said one point six. One point six million. The investor said, more or less, we will write it up.

That was the whole obstruction. Months of meetings, a funded company, a nine-figure track record at the table, and the constraint on the entire enterprise was one sentence the founders could not produce. The money had been sitting there. What was missing was not capital, credentials, product, or market. It was a behavior.

I have been collecting versions of that moment ever since. A few years later I spent about a year as an advisor at a startup lab in downtown Seattle, a full floor of very early founders, and I met with them week after week, dozens of them, close to a hundred over my time there. I watched who eventually made it off that floor and who did not, and the discriminating line was not intelligence, pedigree, or even the quality of the idea. The founders who made it out could describe, in specific detail, the future they were building: what it was, who it served, why it mattered, and what it would look like when it existed. The ones who could not, I came to call the key-lock problem. I have this key, they would tell me, in one form or another. I am just looking for a lock. They could build. They did not know toward what though. In all my time there, I never watched a key find its lock.

The résumé does not confer the behavior. The father in that funding meeting had a résumé almost nobody on earth can match, and the behavior the moment required was not in it. That is the through-line of everything I have watched founders do for two decades, and it is the first premise of this book: behavior is the variable. Behaviors, unlike résumés, can be diagnosed, developed, and changed.

The Three Pillars

If the old path is gone and the human is the differentiator, then the founder’s work reorganizes around three things. I will give you all three now, because the rest of the book is built to serve them.

The first pillar is the founder. You are the first system your company runs on, and in this environment you are also the first system that has to be upgraded. The founders who win from here are the ones who can read the environment, read their team, read the tools, and adjust themselves accordingly: not once, at some annual retreat, but continuously, as a practiced behavior. Notice that I did not name a trait. I did not say the visionaries win, or the bold, or even the agile, because after thirty years of watching this up close I can tell you that every behavioral strength has a failure mode, including the ones this moment seems to reward. What wins is not a type. It is knowing your own profile, honestly, and running it on purpose. That is why this book begins, after my own story, with a behavioral diagnostic rather than a strategy exercise.

The second pillar is people. Here is the paradox of democratized capability: the cheaper the technology gets, the more expensive good judgment becomes. When anyone can operate the tools, what you are actually hiring is what the tools cannot supply: critical thinking, clear communication, the ability to decide well with incomplete information. I will tell you something uncomfortable from the decade I spent teaching at the front of a university classroom, including a capstone course built around critical thinking: those inputs are getting scarcer, not more plentiful. I watched the supply thin in real time, year over year, and everything I have seen in industry since confirms it. I say that as observation, not complaint, because the business implication is the point: there is a run coming on genuinely capable people, and the founders who know how to find them, fit them to the actual maturity of the company, and get the most out of them will hold an advantage that no software subscription can erase. Getting the most out of a small number of excellent people is no longer a management nicety. It is critical for success. The consulting experiment you just saw is the proof: the tools lifted the bottom of the skill curve three times as much as the top, which is exactly why what they cannot supply is now worth more.

The third pillar is what I call the three tenses. A founder now has to execute in past, present, and future tenses at the same time. The past: what your data and your history are actually telling you, read honestly, in the context of when it happened. The present: running today cleanly, the priorities, the decisions, the work in front of the team this week. And the future: the specific state you are moving toward, watched constantly, because the environment is moving fast enough now that an unwatched future does not wait for you. It skips you. Most founders live in one tense, visit the second occasionally, and hope about the third. Executing in all three at once sounds impossible, and done as heroics, it is. Done as a system, with a cadence for each tense, it is Tuesday.

Pillar one is carried by the first of this book’s two instruments. Pillars two and three are carried by the second, and the second pillar, people, runs through both, because you will use the diagnostic on your team as well as yourself, and the execution system is built to get the most from the people inside it. I will name both instruments shortly. First, I need to tell you exactly what kind of work they were built for, because it is not for all work. It is for the work that matters the most.

How Not to Get Left Behind

This book is about enabling founders to operate in today’s environment and arrive at their preferred future state.

It is often a transformation centered on challenging work, the toughest change initiatives that are one-way tickets. It is not strategy, tips, tricks or hacks. Those have their place, but here we focus on how to successfully execute and arrive at your preferred future state. Every business has a pile of projects. Most can be paused, reversed, or quietly abandoned, and the company goes on more or less as it was. But some initiatives are different. The new system the entire operation will run on. The acquisition you are integrating. The move from a company that runs on you to a company that runs on a team. The AI adoption that changes how the work actually gets done. Those are changes, or groups of changes, so profound that you and your business will not be the same on the other side. A bell will have been rung, and it cannot be unrung. That is the test. Projects can be rolled back. Ordinary changes can be reversed. These can’t. Once you take off, there is no return.

That distinction matters because it sets the stakes. If this kind of change were reversible, the cost of doing it badly would be an inconvenience. Instead, the cost of doing it badly is years of your company’s energy, of capital, of careers, of the one resource I have learned to count more carefully than any other: time.

Notice what the revolution does to this class of work: it multiplies it. The AI adoption is a one-way ticket. The move off the founder’s back is a one-way ticket. The repositioning the environment is about to demand of you is a one-way ticket. The old path handed a founder one or two of these in a decade. The new one hands you several at once, moving faster, with less forgiveness for a bad crossing.

As you might expect, founders struggle to execute, and most of these initiatives are done badly.

The Evidence People Seldom Want to Look

At You may have heard the statistic that seventy percent of organizational change (transformation) efforts fail. It gets repeated in nearly every book in this genre, and the honest history of that number is instructive. It traces back to an admittedly unscientific estimate made in the 1990s that somewhere between fifty and seventy percent of reengineering efforts fell short. Which over time hardened, through sheer repetition, into a precise sounding “seventy percent” that has been cited as an established fact ever since. I’m not going to build my argument on a number whose origin is a guess.

What I can do is show you the evidence that is defensible and let you watch it accumulate.

Start with the patterns. In 1995, John Kotter published research that became Leading Change: Why Transformation Efforts Fail, documenting the same recurring failure modes across more than a hundred companies. I have spent my career watching these in action: insufficient urgency, vision without communication, victory declared before the work was done, change that was never anchored in the culture. Then consider the impact on raw delivery odds. Bent Flyvbjerg and his team at Oxford assembled a database of more than sixteen thousand large projects across every sector you can name; the fraction that delivered on budget, on time, and with the benefits promised was less than one percent. Not failing outright, most things eventually limp to some kind of completion. But delivered as designed? Vanishingly rare.

When the major firms study changes directly, the findings rhyme. McKinsey’s transformation practice puts the share that falls short of their goals at roughly seventy percent. Yes, that is the same figure I just told you to distrust in its anonymous form, which is exactly the point: cited carefully, by a named source studying a defined question, it stops being folklore and becomes one data point among several. In 2024, Bain & Company surveyed more than four hundred executives and found that only about twelve percent of what Bain calls “business transformations” achieve their original ambition. Eighty-eight percent fall short. When Bain looked at why, the answer was not strategy or funding, it was that without clearly defined roles and behavioral expectations, execution fragments, and cultural change stays in slogans rather than systems. Hold onto that idea: change stays in slogans rather than systems. It is, in seven words, the entire problem this book exists to solve. The institutions don’t agree on a specific figure, and I’d be suspicious of anyone who claimed they did. However, they agree on the shape of the issue: most big change efforts do not arrive where they set out to go.

Then there is the newest evidence. It should get your attention because it lands at the precise moment you are most likely deciding to change how your own business runs. As organizations raced to adopt artificial intelligence, MIT researchers found in 2025 that ninety-five percent of enterprise generative-AI pilots failed to deliver measurable business impact. Here is the part that matters for this book: the researchers concluded the primary cause was not technology. The models worked. What failed was the organizational willingness to confront operational friction and redesign how the work actually gets done. A new tool, a new mandate, the same ancient failure pattern. Without willing people, a sound-sounding plan, and a result that never materialized.

Now the institutions that spent decades selling the technology response have started saying the quiet part out loud. In 2026, McKinsey’s North America Chair Eric Kutcher described AI as the biggest and most complex change businesses have faced and then put a number on where the work actually lives: it is “80 percent business transformation and 20 percent tech transformation.” Read that again, because it is a remarkable admission from the firm whose name is on the seventy percent figure. The models, the platforms, the licenses, the part everyone budgets for, is the twenty. The eighty is the organization: how decisions get made, how the work flows, how your people actually behave when the new way of working meets their Tuesday. McKinsey’s own analysis of where AI value comes from makes the same point in different clothes. Roughly ten percent from the algorithms, twenty from the technology around them, and seventy percent from the organization the tools land in. Which means the evidence has finally converged on the thing this book is built on: the failure was never in the strategy or the software. It is behavioral, and it always was.

Almost thirty years of my own practice agrees with all of it and adds a diagnosis these studies are only beginning to name. The failures are almost never strategy failures, or even the technology. I have read hundreds of strategic plans. Most of them were fine. Some were brilliant. The failures happen in the gap between the strategy and the human beings asked to execute it.

That gap is what this book is about. The revolution is why the gap is now yours to close, personally: the tools have removed every excuse that used to sit between a founder and enterprise-grade execution. What remains between you and arrival is behavior. Yours, and your team’s.

Why the Existing Tools Don’t Close the Gap

There are two well-developed bodies of practice that should, in theory, prevent these failures. Neither does, and it is worth understanding why before I hand you something different.

The first is the assessment industry. You have almost certainly taken a personality instrument at some point. As I’ll discuss in Chapter 3, those tools measure something real. But what they measure is style in the ordinary sense. How you prefer to communicate, how you process information. They do not measure whether you are behaviorally prepared for what a real change initiative demands. Whether you can hold a long-range vision under pressure, own outcomes in chaos, recover from the plan breaking, and see the system instead of the task. Execution behavior is situational, developable, and almost entirely unmeasured.

The second is the project and change management canon. The methodologies, the certifications, the frameworks. Much of it is genuinely valuable, and you will see me borrow from it with attribution throughout this book. But it was built for large organizations with dedicated staff, and it largely assumes the human side is someone else’s problem (out of scope). It tells you how to structure work. It does not tell you who you are when the work gets hard, and it has remarkably little to say to the founder who is changing the company and changing their own job inside it at the same time. Which, as the next chapter will make clear, is where I learned the most expensive lessons of my own.

What was missing, and what I went looking for through two years of searching, building, and killing ventures of my own, and could not find, was a single, integrated system that starts with an honest behavioral diagnostic of the person and then carries that person, and that team, through a complete execution architecture. Diagnosis plus route. Mirror plus map.

So I built it.

What This Book Gives You

The system has two key instruments, and the book walks you through both.

The first is the ATLAS-ESI. A behavioral diagnostic that establishes your actual starting coordinates: how you, and the people on your team, show up when the work gets hard, across the four dimensions that predict whether an initiative gets finished. This is the first pillar made practical, and because you will run it on your team as well as yourself, it is where the second pillar begins too. That is Chapter 3.

The second is the DELIVER Operating System. A seven-step execution architecture that takes you from wherever an initiative begins to your preferred future state: Discover, Evaluate, Lock, Instantiate, Verify, Execute, Realize, with an iterative cycle that runs from launch to completion. It is how the three tenses become a working cadence rather than a wish: the past read honestly, the present run cleanly, the future watched constantly. That is Chapters 4 through 10. Embedded inside the DELIVER OS are three frameworks that address the three most common operational failure points: MUSIC for prioritization, PAUSE-D for decisions under pressure, and EARN for the founder’s oversight. They arrive exactly where the work needs them.

The architecture is deliberately scale independent. The same system that carried a six-hundred-person global IT organization through a multiyear transformation that helped position a company for a forty-three-billion-dollar acquisition is the system I used to build the business and the book you are holding, with a team of two. That sentence, by the way, is the revolution in miniature: the same discipline, first at enterprise scale with enterprise resources, then founder-sized. Where the large-company and founder-led applications diverge, I will show you both.

Who This Book Is For

I wrote this book for founders and the teams they lead. Specifically, for companies that have outgrown what one person can carry: the founder is still the engine, the team is still learning to run without being pushed, and the next change the business needs is a big one.

If you are the founder, this book is written to you first. You have a business that works, a team that is busy, and a strategy you are confident in. What you do not have is a reliable way to turn that strategy into reality without everything routing back through you. You can feel the initiative wobbling, the calendar filling with everything except the work that matters, and you may not have the vocabulary to say why. This book gives you both: a language for what you are watching (you will recognize the failure modes by name within three chapters) and an execution system your team can run with you rather than around you.

If you are on that team, this book is written to you just as directly. You are the person who has been handed the initiative on top of a full-time job, or the leader of a function being asked to become something it currently isn’t. For you the book is a complete operating manual: the diagnostic that shows how you and the people around you actually show up under pressure, the execution architecture from destination to delivery, and the oversight model that tells you, explicitly, what you are entitled to ask of the person sponsoring the work. Because the sponsor who disappears after kickoff has sunk more initiatives than any market force, and in a founder-led company the sponsor is usually the founder.

If the change you are executing is personal, a career that needs to become a different career, a working life that needs a different engine, the book holds for you too. That was me, three years ago, standing in a parking lot I will take you to in the next chapter, and I have refused at every point in this book to treat the personal application as a lesser cousin of the business one. The system is the same system. The stakes, as I learned, can be higher.

What these readers share is the thing this book actually teaches: getting a real change done is not a project type. It is a discipline. Behavioral first and structural second. It can be learned. That is what makes this a revolution rather than a lament: the old path is gone, but the thing that replaces it is learnable, and you are holding the manual.

How to Read This Book

Read it straight through the first time. The steps are sequenced for a reason, and the later chapters assume the vocabulary of the earlier ones.

Then use it as a reference, which is what it was ultimately designed to be. Each framework chapter ends with a “Field Guide”. A compact, do-this-now protocol for running that step yourself. At the back of the book, you will find an appendix that compresses the entire system onto two pages, and a glossary of every named concept (because a reference book should never make you hunt). Throughout the book you will also find short sidebars labeled “The Science”. Each unpacking a single piece of research that supports and occasionally complicates what I learned in practice. I built these frameworks from thirty years on the ground, not from a literature review. But where decades of published evidence independently arrive at the same place, you deserve to know it, and where I have borrowed an idea, you deserve to know whose it was. The full sources are in the “Notes” at the back.

One more thing. This book will ask you, more than once, to do uncomfortable work. To write down what you actually want, to interrogate your own motives, to look at your behavioral defaults without flinching. I will not ask you to do anything I have not done on the page in front of you, with my own failures as the case study. That is the contract.

The window for doing the work that matters is specific and finite. Your window is now.

A contract deserves to know who is on the other side of it. Before I hand you the system, you should know who built it, and what it cost. That story starts on a winter morning, in a parking lot I had driven to almost every day for four years, with a cardboard box on the seat beside me.