How we scope an MVP in one week
See how we turn an idea into a validated MVP plan in just five days - without unnecessary features or delays.
JULY 22, 2026 • TEAM NFN
By the end of the first week working with a founder, we aim to have four things: the user journey on a whiteboard, a ruthlessly cut feature list, a locked tech stack, and a timeline in writing. After 100+ shipped products, we are confident every one of those fits in five days. Here is how the week actually runs.
Days one and two: the user journey on a whiteboard. The whole product as one flow on one wall. Not epics, not tickets, the actual path a real person takes from I have a problem to this product solved it. The arguments this session produces are the point. Founders discover their cofounders disagree about who the user even is. Better to have that argument in front of a whiteboard than in month three of development, where it costs fifty times more.
Day three: ruthless feature cuts. Every feature on the wishlist has to justify itself against the journey on the wall. The question is never would this be nice. The question is does version one fail without it. Asked honestly, most wishlists lose more than half their weight in a single session, and nothing important is ever in the discard pile. The discard pile is where features wait for evidence. If users ask for them, they come back. Most are never heard from again.
Day four: the tech stack, locked. This should be boring and fast, and with us it is, because we default to stacks we have shipped dozens of products on. Unless your product genuinely demands something exotic, novelty in the stack is a tax on your timeline. Spend your innovation budget on the product, not the plumbing.
Day five: the timeline, in writing. Dates, owners, and a definition of done for each milestone. Writing it down converts optimism into commitment. It is also the honest moment: if the timeline does not fit the budget, we find out on day five, not in month four.
Documents help for sure, but do not de-risk a product. Shipping does. Your real risks are things like users will not pay for this or the workflow does not fit how people actually operate, and no document can retire those risks. Only a working product in front of real users can, and every week spent on paperwork is a week that test gets delayed while your assumptions quietly go stale.
One honest requirement in return: this only works if the founder shows up. Week one needs the person who can say yes and no in the room, all five days. Send a delegate and every decision becomes a meeting to schedule.
Show up with authority, and one week is enough. It has been enough for a very long time.
