What a real MVP is
Almost everyone who's asked me to lead something new has used the word MVP in our first conversation. Almost none of them meant it. What they described was version 1: 25 features, a polished UI, and everything they'll "eventually need." That's not an MVP. That's a wish list with a deadline.
A real MVP is 1-2 things that solve the core problem. That's it. It might look unfinished. It probably will. But I've seen a two-screen app give users an 'aha!' moment that a 20-feature product never could. If someone gets value in two clicks, you've won.
When a team can't pick those 1-2 things, it's not ready to build yet. That's not a product problem, it's a discovery problem. You don't know your user well enough. Go back. Talk to them. Watch them work. Stay there until the one thing they actually need becomes obvious. I've had projects where that conversation changed everything.
Scaleups do this too, with a new product line. The core product sells, so the new one gets planned like the core: full feature set, full launch, big date. It competes with the core for the same team, and loses, and a year later it's still "almost ready".
And whoever builds it has every reason to nod along. All 25 features, half-baked versions of each, no pushback, no launch date in sight. A good partner tells you no. A bad one tells you "sure, we can add that" until you run out of budget and still don't have a product.
Be razor-focused with every decision. With AI, anyone can spin up a large app that looks nice in a weekend. But looking nice doesn't mean useful. Useful doesn't mean desirable. And none of it means viable or maintainable. Build the one thing that matters, ship it, see what happens. Adapt, add, grow.
A new product line stuck at "almost ready"? That's what a special-projects lead is for.
More writing, or see how it plays out in previous work.