one flowfirst
How it worksThe full processWritten from your side, not mine.

Here's exactly whathappens.

Two weeks, two steps, and a bar set before anyone sees the build. No surprises for you.

OneStep oneIf something fatal shows up, you find it here.

Is this worth building?

Before a line of anything gets made, you get clarity on the questions that decide whether the idea survives contact with the real world.

  • Is this a real, painful problem, or a feature that sounds cool? One gets paid for. The other gets admired and forgotten.
  • Who exactly is the customer, and can you reach them? "Everyone" is not a customer. A reachable group of people with the same problem is.
  • Would people pay, or would they just say "I'd use it"? Those are not the same sentence, and only one of them is evidence.
  • Is the exciting part actually the risky part? Usually not. The risk is somewhere quieter, and that's the thing worth testing.
TwoThe decision ruleA bar set after seeing the data is a bar that moves.

Agreed before testing

You agree in advance what a result means. Not after you've seen the data, when hope starts quietly moving the bar. Before.

Go

People do the thing without being nudged, and they come back to it. You know what to build next.

Adjust

The problem is real but the shape is wrong. You know specifically what to change, not just "iterate."

Stop

People are polite but nothing pulls them. You keep your money and move on to a better idea.

ThreeStep twoWhat gets left out is a decision, not an oversight.

The part that matters, made real

One idea, one flow, the one that matters. You end up with something a real person can pick up and use on a phone, without you standing over their shoulder explaining it.

Real enough to test means the core action works end to end, it's fast, and it doesn't look like a prototype. People behave differently in front of something that feels real, and that difference is where the honest data comes from.

What gets left out on purpose: accounts and settings you don't need yet, edge cases nobody will hit in a test session, admin screens, polish on flows that aren't the question, and anything built for scale you haven't earned.

FourThe decision briefRead the call first, then the evidence.

What the brief contains

  • The riskiest assumption. Stated in one sentence, the way you'd say it out loud.
  • How to read what comes back. What to count as a signal, which flattering moments to ignore, and what people doing rather than saying looks like on paper.
  • The call. Go, adjust, or stop, measured against the rule you set in advance and written once the sessions are done.
  • What you'd do next. Either the shortest path forward or the honest reason there isn't one.

If the call and the evidence ever disagree, trust the evidence and come argue about it.

FiveThe testing playbookThe method, handed over.

So the next idea needs no one else

  • Where to find the right people. The ones with the problem, not the ones who like you.
  • How many you actually need. Fewer than you'd expect. The big patterns show up early, and the rest is confirmation.
  • How to run a session. What to ask, what never to ask, and how to keep politeness from quietly ruining your results.
  • What to measure. The handful of signals that mean something, and the flattering ones that don't.

Finding the right people is the part that takes real calendar time, and it depends on strangers being free. That's why the sessions sit outside the two weeks rather than inside a timeline nobody can honestly promise. Run them yourself with this, or have them run for you as an add-on.

SixWhat happens afterAll three outcomes beat not knowing.

What happens after

Three ways this ends.

Continue

The core assumption held. You know what to build next, and you can build it with anyone, including me.

Pivot

Part of it held and part of it didn't. The idea points somewhere new, and the same two weeks can run again on the changed version. New question, same method.

Stop

It didn't hold. You'll get a clear reason why it's a real stop and not "try harder", and you'll have kept the money you would have spent finding out the long way.

Some people come back after a pivot, or later with a different idea. None of that is decided up front. The two weeks stand on their own.

See it run on a real idea

NextStep oneTen minutes of questions, then a reply from a person.

Ready to find out?

Ten minutes of questions, in your own words, then a straight read on whether this fits. The answers are yours to keep either way.

Tell me about your idea

Here it is. It’s yours either way.

Nothing has been sent yet — this site has no server to send it to, and a form that says “sent” when it means “discarded” is worse than no form. Copy it or open it in your mail app, and it reaches me the moment you hit send. Keep a copy either way: the answers are more use to you than they are to me.

  1. A few questions about your idea.

    About ten minutes if you take your time. There are no right answers and none of this is a test — a rough answer is more use to me than a polished one.

    Only the idea and your email are needed. Skip anything you’d rather say out loud. It saves as you go, so you can close this and come back. Nothing leaves your browser until you choose to send it at the end.

  2. One or two sentences. Rough is fine — this is not a pitch.

    A sentence is enough, but it needs to be a sentence.

  3. Not everyone. The narrower you can make this, the more useful the two weeks are.
  4. Whatever you’d be replacing, even if it’s a spreadsheet, a WhatsApp thread, or nothing.

  5. This is the one the two weeks gets built around.

    The one thing that, if it turned out false, would make the rest pointless.

  6. Halfway. Now the useful part.

    It’s usually not the part that’s hardest to build.

  7. This one does the most work.

    A number, a behaviour, anything you could actually watch happen.

  8. An honest answer here is worth more than a polite one.

  9. A doc, a note, a conversation you keep having.
  10. A date, a competitor, money running out, a person waiting on you. Or none.

  11. Nearly there. Pick what is true today.

    What do you want out of the two weeks? Optional

  12. Last one.

    Where should I send my reply?

    I read every one of these myself. No scoring, no auto-reply, nothing in between.

    Please add your name so I know who I’m talking to.

    Please enter an email address I can reply to.

Your answers stay in this browser until you send them, and are used to reply to you about this call and nothing else. No list, no sequence.