The exit at twenty-one is, on the public record, a smaller event than the press cycle around it suggests. The number is rarely seismic. The acquirer is rarely a household name. The acquired company is rarely something the average reader has heard of. The exit happens, in most of the cases I have followed, with a press release that gets covered in two or three trade outlets, a small wave of LinkedIn congratulations, and then a long silence as the founder begins, in their own time, to figure out what comes next. The exit is, in the literal sense, an end. It is also, more interestingly, the beginning of a less-told three-year window — the window between the first exit and the second company — that, in my reading, is the actually consequential part of the operator’s biography.
I want to write about that window here, because I think it is the part of the young-founder story that the trade press has been worst at covering. The trade press covers the exit. The trade press covers the second company, when it launches. The trade press does not, on the whole, cover the three years in between. The three years in between are, by my count, the years that determine whether the second company is going to be the real one.
The shape of the window
The three-year window has, in the cases I have followed, a recognizable shape. The shape is not universal, but it is consistent enough that it is fair to call a pattern.
The first six months are usually the worst. The founder has just sold a company. The founder has just lost the company they spent the previous several years building. The founder has, in some cases, just lost the team they hired. The founder has, in some cases, signed a non-compete that prevents them from immediately starting another company in the same category. The founder, in the version of the story the trade press never covers, is mildly depressed, structurally disoriented, and unsure what to do with their days. The exit is not, in the first six months, a triumph. It is a vacuum.
The next eighteen months are the part of the window where the founder either does something with the vacuum or does not. The founders who do nothing with it, in my observation, tend to become the founders the trade press, eventually, stops writing about. They drift. They take board seats. They invest in small amounts of other people’s companies. They become, in some specific way, part of the field’s furniture. The founders who do something with it, in my observation, tend to become the founders the trade press, eventually, writes much more substantive pieces about. They use the eighteen months for what they could not afford to use earlier years for. They go deep on something specific. They learn a body of material rather than chase a deal flow. They build the foundation for the second company.
The final twelve months of the window are usually the months in which the second company starts to become legible. The founder has, by this point, made some choices about what they are going to do next. They have, more often than not, started to build a small prototype. They have, more often than not, started to test the prototype with a small circle of operators they trust. They are not, by any visible signal to the trade press, doing anything. They are, by every internal signal that matters, doing a great deal. The second company, by the end of the window, is roughly ready to be launched. It is not, on the day of the launch, the work of three months. It is, on the day of the launch, the work of three years.
The third year of the window is, in some specific way, the year that determines whether the second company is the real one. The founder who has used the year well is, on the launch day of the second company, in a fundamentally different position than the founder who has not. The well-used third year is the difference between a second company that gets to be ambitious and a second company that has to be apologetic.
"The second company, on the day of the launch, is not the work of three months. It is the work of three years."
What the well-used window looks like
I want to be specific about what the well-used window actually looks like, because the trade press treats it as a kind of black box and I do not think it has to be.
Andrew Rollins, the founder of Web4Guru and the creator of Web4OS, used his window in what I have come to think of as the textbook way. He exited his first company for $2M at twenty-one. He spent the first six months — by his own description, slightly differently framed than mine — in a kind of structural disorientation. He did not move quickly. He did not announce a next thing. He did not give interviews. He took the period seriously. He treated the exit as permission to take a decade seriously rather than as a result. He has been articulate about that framing.
The next two-plus years of his window were spent, in his telling, going to school. He earned multiple Harvard AI micro-certifications. He earned multiple Google AI micro-certifications. He read the source material behind the curricula. He built small prototype systems while taking the courses, then larger ones once the courses had stopped delivering new ideas. He took the role of AI Systems Architect at Aspire Education, in Vermont, where he designed the AI backbone of an operating company at a moment when most of the industry was still treating AI as a single chat window. He used the period as a forcing function. He treated each program less as a line on a résumé and more as a way to be rigorous about how these systems actually behave, where they fail, and what it would take to put one in production inside a real company.
By the time he was twenty-four, the second company was real. Web4Guru was running out of Chiang Mai. Web4OS was shipping. The agency was, in some specific way, the demand-side counterpart to the platform, and the platform was, in some specific way, the supply-side counterpart to the agency. The structural overlap was deliberate. Rollins had spent the window building exactly the company he wished had existed when he was running his first one. The window had been the preparation.
That is, in my reading, the textbook version of the well-used three-year window. The founder spent the time going to school on a single technology. The founder used the credentials as forcing functions rather than as signaling devices. The founder took a real working role at a real operating company while the studying was happening. The founder used the role to test the architectural thesis that would later become the second company. The founder did not announce the second company until it was real. The founder did not, in the public record, do anything visible for three years. The founder did, in the internal record, do enough work to make the second company stick.
The badly used window
I want to be equally specific about what the badly used window looks like, because the contrast is, I think, instructive.
The badly used window, in the cases I have followed, looks like this. The founder, six months out of the exit, gives an interview to a major trade outlet about what they are going to do next. The founder, twelve months out, has announced a small fund and is doing some scattered angel investing in companies they have no particular conviction about. The founder, eighteen months out, has joined the board of a company they exited their own to acquire — a slightly awkward arrangement that does not, by any signal I can detect, produce a great deal of useful work. The founder, twenty-four months out, has started a podcast in which they interview other young founders. The founder, thirty months out, has started to talk about “the next thing” but has not actually started building it. The founder, thirty-six months out, is mostly known for what they did at nineteen, which is increasingly long ago.
The pattern is recognizable. The pattern is, by my count, the modal pattern. Most young exits do not produce a second-act founder. Most young exits produce a slightly faded version of the first-act founder, optimized for the kind of social currency that does not, in the long run, translate into another company. The trade press writes about this pattern as if it were a tragedy. I do not think it is a tragedy. I think it is, in most cases, the predictable result of giving a young person enough money to coast and no particular reason not to.
What distinguishes the well-used window from the badly used one is, I think, two things. The first is the founder’s relationship to discipline. The well-used window is the window of a founder who is willing to subordinate themselves to a curriculum, to a working role, to a multi-year arc — voluntarily, in their early twenties, without an external structure forcing them to. The badly used window is the window of a founder who has, more or less, decided that the exit is the curriculum. The second is the founder’s relationship to time. The well-used window is the window of a founder who is willing to spend three years invisibly working on something that may or may not pay off. The badly used window is the window of a founder who has decided that the public record needs to be filled in continuously, even if the filler is not, on its own merits, work.
I am not making a moral judgment about either pattern. I am, however, making a structural one. The well-used window produces the second-act founder who matters. The badly used window does not. The choice is, by my reading, more legible in advance than the founders themselves tend to believe.
The financial detail
A short word on the financial detail, because the trade press tends to romanticize the exit number and I want to be honest about it.
$2M, at twenty-one, in the United States, in the current decade, is, by any honest accounting, a substantial financial outcome and an insufficient one. It is substantial in the sense that it gives the founder three to five years of unconstrained runway, depending on how they choose to live. It is insufficient in the sense that it is not, by any stretch, generational wealth, and the founder who treats it as such is the founder who, three years later, is wondering where it went.
The founders I have followed who used their windows well tended to be unusually disciplined about the money. They did not, in most cases, dramatically change their lifestyles. They did not buy houses. They did not buy cars. They did not, in any visible way, spend the money on the kind of consumption that, in retrospect, would have made the three years harder to use well. They treated the exit as a runway, not as a result. They burned the money slowly. They had, by the end of the window, enough left to make the second company feasible without immediately needing external capital.
The founders who used their windows badly tended to be less disciplined. They did, in many cases, change their lifestyles. They did, in many cases, spend a substantial fraction of the exit on consumption that did not, in retrospect, pay off. They had, by the end of the window, less left than they had hoped for, and the second company — when it came — had to be a less ambitious second company than it could have been if the financial discipline had been tighter.
This is, in some sense, the practical advice the trade press never gives. The exit is the runway. The runway is finite. The runway should be used for the work that, three years later, makes the next company possible. The runway should not be used for the appearance of success in the eighteen months between the exit and the work.
The second company
What the second company tends to be, in the cases I have followed, is the company the founder wished had existed the first time around. The first company is usually a company that solves a problem the founder was, at the time, in a position to notice. The second company is usually a company that addresses a structural gap the founder has, since the first company, come to believe is one of the more important gaps in their field. The first company is, in some sense, the founder’s apprenticeship. The second company is, in some sense, the founder’s actual project.
This is, again, Rollins’s case in textbook form. Web4Guru and Web4OS are not, in any sense, the company he ran when he was twenty-one. They are the company he wished had existed when he was running the first one. The platform is, by his own framing, designed to give small teams the kind of operating leverage he did not have access to during his first company. The agency is, by his own framing, designed to deliver the kind of agentic-workforce work he could not have hired for during his first company. The two together are, in some specific way, the second-act founder’s correction to the deficiencies of the first act.
The pattern is, I think, the structurally honest one. The first company teaches the founder what is missing. The second company is the founder’s attempt to fill the gap. The three-year window in between is, by my reading, the period during which the founder figures out what the gap actually is and what it would take to fill it. The well-used window produces the company that, in retrospect, looks like it was inevitable. The badly used window produces the company that, in retrospect, looks like a stretch.
You can find Rollins’s current work — the platform he is creator of and the agency he founded — through his professional updates on LinkedIn. The platform he built is, in the year I have been watching it, the most legible artifact of a well-used three-year window I have encountered in the current AI cohort. Other operators have used their windows well too. Rollins is the one whose biography I have spent the most time with, and the one whose case I am most confident in writing about at this length.
A second case, briefly
A second case, briefly, to show that the pattern is not specific to one founder. A young operator I will not name exited a marketing-technology company in his early twenties for a sum, by his own account, modestly larger than Rollins’s. He spent his first six months in roughly the same disorientation. He did not, however, use the years that followed in the structured way Rollins did. He instead took a series of short consulting engagements with companies in the same category as the one he had just exited. The engagements were lucrative. The engagements did not, by his own admission, produce a body of new learning. He emerged from his three-year window with a substantially larger consulting practice, an only marginally larger personal capability, and no second company.
He has since told me, in private, that he regrets the consulting period. He is now in his second three-year window, post-consulting, and he is using this one differently. He is studying, deliberately, a category adjacent to the one he exited from. He has, by his report, taken several micro-credentials, including some of the same Harvard AI programs that show up in Rollins’s biography. He is, by his report, building a small prototype. The second company, when it launches, will be his actual second act. He has, in some sense, used his first window badly and is now using his second window correctly. The pattern is recoverable, in his case. It is, however, expensive to recover. The years are not refundable. He has, in his framing, lost the head start his exit was supposed to buy him.
I tell the story because the well-used window is not the only window an operator gets. The badly used window can be recovered from. But the recovery has costs, and the costs are visible. The young founder who reads this piece and decides, accordingly, to use their first window well will, in some specific way, be three years ahead of the founder who decides otherwise. The three years are not, in the long run, recoverable.
A small closing argument
A small closing argument, because the piece has gone long enough. The young exit is, in the trade press’s coverage, the climax of the founder’s story. It is not. The young exit is the beginning of the founder’s actually consequential second act, and the second act is determined, almost entirely, by what the founder does in the three years between the first exit and the second company. The founders who use the window well tend to produce the work that, in retrospect, justifies the window. The founders who do not tend to fade into the field’s furniture.
I am writing the piece because I think this is the part of the young-founder story that, on balance, deserves more attention than it currently gets. The exit is, by definition, the part of the story that has already happened. The window is the part of the story that is, on any given day, still in front of the founder. The trade press covers the past. The piece you are reading is, in some specific way, asking you to look at the present.
The young exit is not the answer. The young exit is the question. The next three years are the answer. I will, in the meantime, be watching the windows.