The first week in the Incubator, hour by hour

Most incubators treat the first week as a warm-up. Onboarding sessions. A vision workshop. A logo conversation. A calendar full of introductions and a Notion full of templates. By Friday the founder has met everyone and built nothing, and the first week, the one with the most energy and the least baggage, is gone.

We run ours the opposite way. By the end of week one a Moonlabs Incubator founder has something live on the internet that a stranger can use. Not a deck about the thing. The thing. Here is how the five days actually unfold, because the shape of week one tells you more about how we work than any description of the programme could.

Monday: kill the darlings, find the wedge

The founder arrives on Monday morning with an idea they have usually been carrying for months, sometimes years. It is almost always too big. That is not a criticism; big is why it is worth doing. But you cannot ship "the platform" in a week, so the first job, and the whole of Monday morning, is finding the wedge.

We sit down, the two of us and the founder, and we take the idea apart. What is the single sharpest thing this product does for one specific person? Not the vision, the wedge: the smallest version that is still worth paying for. This is uncomfortable, because it means setting aside nine tenths of what the founder walked in excited about. We are ruthless about it, and we are ruthless on purpose, because the founders who succeed are the ones who can hold the big vision in their heads while shipping the small wedge with their hands.

By Monday lunchtime there is a wedge on the whiteboard, defined tightly enough that you could build it. Monday afternoon is scoping: what is the thinnest slice of that wedge we can put in front of a user by Friday. We cut, and then we cut again. Almost everything a founder thinks is essential in week one is not. The cutting is the lesson.

Tuesday: build the spine

Tuesday is a build day, and it starts early. The founder is at the keyboard, and one of us is alongside, not doing it for them but building next to them, which is the entire model of the Incubator. This is where the operator-led part stops being a slogan.

We are not teaching abstractly. We are saying, on this specific repo, at this specific decision, here is the call we would make and here is why. The stack is whatever ships fastest for this product, usually the same battle-tested patterns we run across our own companies. By the end of Tuesday the spine exists: the core loop of the product works end to end on the founder's machine, ugly but real. It does the one thing. Nothing around it is finished. The one thing works.

Tuesday evening we do the first honest gut-check. Is the wedge actually as sharp as it looked on Monday now that we can see it running? Sometimes the answer is no, and we would rather learn that on Tuesday of week one than in month three. If it needs to move, it moves now, while moving is cheap.

Wednesday: make it real enough to touch

Wednesday is about turning the working spine into something a stranger could use without one of us standing next to them. That means the unglamorous work: the sign-up, the empty states, the one error message that stops a user getting stuck, deployment to a real URL. Founders often want to skip this and add features. We hold the line. A second feature nobody can reach is worth less than a first feature a real person can.

By Wednesday afternoon the product is deployed somewhere with a link. It is thin, and it is live. This is also when we start the conversation that most incubators leave until much later: who exactly is the first user, and how do we get the product in front of them this week. Not next month. This week. Distribution is not a phase that comes after building. In week one it runs alongside it.

Thursday: put it in front of someone real

Thursday is the day the product meets a human who is not us. We line up prospective users, the actual person the wedge was built for, and we watch them use it. Not a friendly demo where the founder narrates. We stay quiet and we watch where they hesitate, what they misunderstand, what they ignore, and whether the one thing actually lands.

This is the most valuable hour of the week and often the most bruising. The thing the founder was proudest of gets ignored. The detail they nearly cut turns out to be the whole point. A real user in week one resets a founder's instincts faster than a month of theorising, because you cannot argue with someone quietly failing to find your main button. Thursday afternoon we act on what we saw. Small changes, fast, straight back in front of another user if we can.

Friday: ship, and look back

By Friday the product is live, it has been touched by its first users, and it has been changed in response to what they did. We spend Friday morning getting it into a state we are not embarrassed by, and then we ship it properly and, more importantly, we stop.

Friday afternoon is the retrospective, and it is where week one earns its keep. We look back at Monday's idea and hold it against Friday's reality. What survived contact. What we killed and were right to. What the users taught us that no amount of planning would have. And we set the wedge for week two, sharper now because it has been tested against something other than our own conviction.

The founder ends the week tired in the specific way that means work happened, holding a live product and a set of instincts that are five days more honest than they were on Monday morning.

Why we run it this hard from day one

We could ease founders in. Spend week one on vision and relationships and strategy, and start building in week two once everyone is comfortable. We deliberately do not, for three reasons.

The first is that week one sets the metabolism of the whole twelve weeks. A founder who ships in week one believes, in their body rather than their head, that shipping weekly is normal. A founder who spends week one talking learns that talking is what an incubator is for, and that belief is expensive to unlearn. We would rather set the pace on Monday morning than fight to raise it in month two.

The second is that building immediately surfaces the truth about an idea faster than anything else. Ten weeks of runway is not long when a real round is meant to be closing at the end of it. We do not have a month to discover that the wedge is wrong. Shipping in week one buys us that discovery on day four, when it is nearly free to act on.

The third is simply that it is how real companies are built, and the Incubator exists to build real companies, not to simulate the experience of building one. The intensity is not theatre. It is the actual condition of shipping something people want under a clock, and week one is where a founder finds out whether they thrive in it. Most of the ones we choose do. Finding that out on day five, together, is exactly what the first week is for.

Louis O'Connell-Bristow and James Freestone are the co-founders of Moonlabs, the operator-led AI incubator and academy, and previously built the home.co.uk, Homemove and homedata.co.uk stack. The Incubator runs twelve weeks, building alongside the founder until the product ships and the round is closing. Site: moonlab.ventures.