Why Your First Ten Customers Should Not Look Like Your Next Hundred

Early customer acquisition is the job of finding ten people who will tolerate an unfinished product in exchange for a problem being solved. That is a different job from acquiring a hundred customers, and the people who say yes to the first are, by definition, not typical of the second.

Founders hear this as permission to ignore fit, which it is not. It is the opposite. The first ten should be unrepresentative in a specific, chosen way, and knowing which way is most of the skill.

What the first ten are actually for

They are not for revenue. Ten customers at any price a new product can command is not a business, and treating early revenue as the goal will push you toward whoever will pay rather than whoever will teach you.

They are for three things.

Evidence that the problem is real enough that someone will change their behaviour over it. Not that they find it interesting. That they stopped doing something else.

A description of the problem in language you did not write. You will use their sentences on your homepage for the next two years, and you cannot invent them.

A working relationship close enough that they will tell you the product is bad. This is the one that decides whether the first ten were worth having.

The ways they should be unusual

Your first ten should be people with the problem more acutely than average. Not slightly more. Considerably more, to the point where they were already improvising a solution before you turned up. Someone who built a spreadsheet to cope is worth twenty people who agree the problem exists.

They should be reachable without a channel. You have no distribution yet, so your first ten come from people you can contact directly: your network, communities you are already in, individuals you can email one at a time. That is not scalable and it does not need to be.

They should be tolerant of rough edges, which usually means they are technical, or desperate, or personally invested in you. All three are fine and all three are unrepresentative.

They should be close enough to talk to weekly. Ten customers you speak to every week are worth more than fifty you never hear from.

What they should not be is the wrong buyer entirely. Someone with a different problem who will pay you anyway is the single most expensive customer you can acquire early, because they will pull the roadmap toward a market you have not chosen, and they will do it politely while paying.

Why the next hundred are different

The hundred come through a channel, not a conversation. That changes what has to be true.

They will not tolerate rough edges, because they did not meet you and have no investment in your success. They will read the homepage and decide in eleven seconds. They will not answer your emails asking what they think.

Most importantly, they have the problem at ordinary intensity rather than acute intensity, which means the product has to do more of the work of convincing them. The first ten forgave your onboarding. The next hundred will leave during it.

This is the transition that breaks most early-stage companies, and it usually gets misdiagnosed as a marketing problem. It is a translation problem. Everything that worked at ten worked because you were personally present, and none of that survives contact with a channel.

When to stop selling like this

The signal to stop is not a number of customers. It is repetition.

When three consecutive customers describe the problem in the same words, and buy for the same reason, and use the product in the same way, you have found something worth generalising. That can happen at eight customers or at twenty-five. Counting to ten and switching mode on schedule is how founders end up scaling a pattern that was never there.

The opposite signal matters too. If your first ten bought for ten different reasons, you do not have a market yet. You have ten favours. That is worth knowing early and it is much easier to see if you wrote down why each one bought at the time, rather than reconstructing it later from memory that has quietly tidied itself up.

The honest cost

This approach has a real downside, and it is worth naming rather than glossing.

Selling manually to ten acute-pain customers teaches you a great deal about ten people and almost nothing about a market. You will finish with deep conviction and a sample size that would embarrass a first-year statistics student. Founders regularly take that conviction into a channel and discover the wider market does not care, which is an expensive lesson delivered late.

The mitigation is not more early customers. It is doing market research alongside the conversations rather than instead of them, so that when you generalise you are generalising from two independent sources rather than one. That is essentially what a launch plan is for: FirstFlight researches your market live and builds the plan around what you are shipping, including positioning, a first-100 playbook and a dated 90-day tracker, for £79 one-time.

What to do with the first ten once you have them

Do not graduate them. The temptation once a channel starts working is to stop the weekly calls, because they are expensive and the customer is already won.

Keep three of them. Specifically the three who were most willing to tell you something was bad. They are your early warning system for every change you make afterwards, and their value goes up rather than down as you stop being able to talk to everyone.