101 people told me who they are. 11 told me what they need.

I ran the same experiment twice and got a ten-to-one answer. Here is what I am changing because of it.

September 23, 2026

In August I sent this list one question with one tap as the answer: which one are you right now?

101 people answered.

Founder 45
Service provider 15
Investor 14
Operator 13
Supporter 10
Hiring right now 4

Four days later I sent a question that was, as far as I could make it, identical in cost. One tap, six options, no form, no login: what do you need most right now?

Eleven people answered.

Ten to one

I want to be careful here, because this is not a clean experiment. Different days, different subject lines, each question shown once. If you wanted to dismiss it you could, and I thought about doing exactly that.

But the gap is not close. It is roughly ten to one on the same list, in the same week, for two questions that took the same half a second to answer. Something real is in there.

Here is what I think it is. Saying who you are is free. Saying what you need is not.

Tapping "Founder" costs you nothing. It is a flattering thing to be, it is already on your LinkedIn, and it commits you to precisely nothing afterwards.

Tapping "Money" is a different act. It is an admission that you are stuck, sent to someone who might tell other people, at a moment when you are probably telling everyone else that things are going well. That is not a one-tap decision no matter how few taps I make it. I built the interface down to nothing, and it turned out the interface was never the expensive part.

Why this matters more than a poll result

This network only does one thing. It moves help between people who have it and people who need it. Introductions, capital, customers, a person who has already solved the thing you are stuck on.

Every version of that depends on somebody saying what they need. And the clearest signal I have says that most people, most of the time, will not.

So there are two honest options. Keep asking in a nicer voice and keep getting eleven. Or stop building around the assumption that people will tell me what is wrong, and build around what they will actually do.

I am doing the second one.

What changes

Three things, concretely.

I stop asking you what you need and start working it out. From what you clicked, which rooms you turned up to, what you said you were building six months ago and have not mentioned since. If that sounds presumptuous, it is. The alternative is a network that can only help the eleven people comfortable enough to raise a hand, and that is not a network, it is a favour I do for the confident.

I have to be right about small things before I am trusted with big ones. A wrong guess about what you need is worse than silence, because you only give someone a couple of tries before you stop opening the email. So this starts narrow and boring and earns its way up.

And I ask the other half of the question instead. Which is the one below.

The easier half

The same six options. Inverted.

Offering something costs nothing to admit. Nobody has ever felt exposed saying they know a good lawyer. And the offer side turns out to be the half I genuinely cannot infer — I might guess from your behaviour that you are quietly looking for customers, but I have no way to guess that you happen to know the one person in Lagos who could unblock someone else on this list.

One tap. Pick the one you could actually do this month, not the one that sounds best.

Whatever comes back, I will publish the number here. Including if it is worse than eleven.

One more thing

If you were one of the eleven who answered the hard question in September: I read all of them. Three of you said money, two said expertise, two said an introduction to a specific person.

The specific-person ones are the easiest requests I ever get, and almost nobody makes them. If you know exactly who you need to meet, say the name. That is a solvable problem, and I would rather spend this week on it than on almost anything else.

— Stephen


This is the first of a weekly series where I publish what the numbers inside Propel actually say, including the ones that make the strategy look wrong. If that is not what you signed up for, the unsubscribe link is where it always is, and there are no hard feelings.



Powered by beehiiv

Get the next one by email. Once a week: what the network needs, what closed, the founders, and the rooms coming up. Free, and you can unsubscribe from any issue. Subscribe and read every issue · File a free Propel Card