Why Helping One Person Properly Beats Helping a Thousand Badly

There is a particular kind of ambition that skips a step. You want to help people, so you start by asking how many. A thousand sounds better than one. A million sounds better than a thousand. You build for the million, and somewhere in the scaling you lose the thing that would have helped anybody at all.

The step that got skipped is the one where you help a single person and watch what happens.

Reach is a proxy, and proxies drift

Reach is easy to measure, which is exactly why it takes over. You can count subscribers, downloads and impressions before lunch. You cannot easily count whether anyone’s life got better, so the number that is easy to see becomes the number you optimise, and after a while it is the only number anyone talks about.

Meanwhile the actual effect on one person is slow, awkward and specific. It involves finding out that the thing you built solves a problem they do not have, in a way they cannot use, at a moment when they are busy with something else. That is unpleasant information. It is also the only information worth having.

If you cannot describe how your work changed one named person’s week, you do not yet know whether it works.

Depth is what generalises

The counterintuitive bit: helping one person properly is how you eventually reach the thousand. Not the reverse.

When you sit with one person’s problem long enough to actually solve it, you learn the shape of the problem. You find out which part was hard, which part they had already worked around, and which part you had completely misunderstood. That understanding is portable. The next hundred people have variations of the same problem, and you now know what you are looking at.

Start with the thousand and you learn none of it. You get averages, and averages describe nobody. You end up with something that is a mediocre fit for everyone and a good fit for no one, and no idea which part to fix.

What this looks like in practice

Pick one person. A real one, with a name, who has the problem you care about. Help them all the way to the end, past the point where it stops being interesting and starts being logistics. Notice what you had to do that you did not expect.

Then ask whether the awkward parts were specific to them or true of everyone. That question is the whole method. It is how a favour becomes a product, and how a piece of advice becomes something worth writing down.

You will also find out something less comfortable: whether you wanted to help, or wanted to have helped. Those two feel identical right up until the work gets boring. Wanting to have helped runs out at about the same time the novelty does.

The number still matters, later

None of this is an argument for staying small. Scale is how a good thing reaches the people who need it, and refusing to scale is its own kind of vanity.

It is an argument about order. Get the thing working on one person, learn why it worked, then take it wide. A thousand badly is not a step towards a thousand well. It is a different activity that happens to produce a bigger number.