Ideal Customer Profile: Why "Developers" Is Not an Answer

You launched three months ago. The product works. You have 47 GitHub stars, a handful of signups who logged in once, and nobody paying you.

Someone tells you to figure out your ideal customer. You think: fine, developers. My product is for developers. And you move on, because that felt like an answer.

It was not an answer. Your ideal customer profile is the single decision that determines what every cold email, landing page and community post says — and "developers" determines nothing at all. This post shows you how to narrow it to something you can act on by Friday.

Who has this problem

You are a developer-founder. You can build almost anything, and you have never sold anything.

You are comfortable with hard technical decisions and deeply uncomfortable with the idea of picking a subset of people and ignoring everyone else. It feels arbitrary. It feels like guessing. And it feels like throwing away most of your market on a hunch.

If you have ever opened a blank email to a potential customer, stared at it, and closed it without writing a word, this post is for you. That blank page is not writer's block. It is your audience definition failing in real time.

Why "developers" feels safe and fails anyway

There are roughly 30 million developers in the world. They work at two-person startups and at banks. Some install whatever tools they like. Most cannot install anything without approval. Some hold a budget. Most have to go and request one.

Nothing meaningful is true of all of them. So anything you write is either vague enough to apply to everyone — landing on nobody — or specific enough to be wrong for most of them.

There is a quick diagnostic for this. Open a blank email and write the first sentence you would send to a stranger to get them to try your product. If your audience is "developers", you cannot write it. There is no person on the other end, only a category.

The deeper reason founders resist narrowing is that it looks like shrinking the market. It is not. Narrowing does not reduce your eventual market. It reduces the market you are talking to this month.

And that is the constraint that actually binds you. You are one person with no ad budget. Your limit is not how many people could theoretically buy — it is how many you can find and speak to convincingly. A narrow group lives in one subreddit, one Slack community, one conference attendee list. "Developers" cannot be found anywhere, because they are everywhere.

Y Combinator's own post-mortem data puts "no market need" ahead of running out of money as a cause of startup death. A meaningful share of those companies had a market. They just never identified it precisely enough to reach it.

How to narrow your ideal customer profile

Step 1: Run three filters, in order

Most candidate audiences fail at the second one.

Do they have the problem acutely? Not "would they agree it is annoying" — everyone agrees things are annoying. Do they currently spend time, money or attention working around it?

Do they know they have it? A buyer who has not noticed the problem must be taught before they can be sold to, which roughly triples the length of the sale. Early on you do not have the runway. Pick the people who already know.

Can they say yes on their own? If your user must ask someone else for budget — someone who does not personally feel the pain — that user is not your buyer. They are an advocate. Useful, and much slower.

Step 2: Separate the buyer from the user

This is the expensive one in developer tools.

Take CodeReviewBot, an AI pull-request reviewer. The user is a senior engineer who wants their afternoon back. The buyer is that engineer's manager, who wants delivery to stop being unpredictable. Same product, two completely different sentences.

Write to the buyer's problem in your outreach. Solve the user's problem in the product. Get this backwards and you build something people love that never reaches anyone with a card.

Step 3: Write the profile as a sentence you could hand to a stranger

Here is the test that matters: could someone else take your profile and go find you ten of those people by tomorrow afternoon?

Compare the before and after for CodeReviewBot:

Before: Developers.

After: Engineering managers at 20–200 person product companies whose team ships daily and whose senior engineers spend over an hour a day on code review.

The second one writes the email for you. It also tells you roughly how many such people exist and where they gather — which is the entire point.

Step 4: Fill in the formula

Use this and resist the urge to soften it:

I sell to [role] at [kind of company], who currently handle this by [workaround], which costs them [cost] every [frequency].

If you cannot fill in the cost or the frequency, you have identified a preference rather than a pain. That is a research problem, not a writing problem — go and run a few customer interviews before you go further.

Step 5: Write it down before you validate it

Your first profile is a guess. It is supposed to be. Write it down anyway.

A guess you have written down is testable. A guess in your head quietly changes shape every time reality disagrees with it, which means you never learn anything. The purpose of writing it is to be provably wrong about something specific.

Why this framework holds up

None of this is clever. It is the same narrowing every company you admire did before it widened.

The Mark Platform builds this into step 4 of an eight-step journey, and it is deliberately placed after customer research rather than before. You cannot write a real profile without having talked to people, and you cannot run a useful interview without a rough idea of who to talk to. So the platform teaches the sketch, sends you to validate it, then has you write it properly — and it scores an assumed pain point lower than a validated one, because an assumption is worth less.

Where this will not work: if you have not spoken to a single potential customer, no framework will produce a real profile. You will produce a well-formatted description of yourself. That is why interviews come first, and why validating product-market fit is a separate discipline from writing a profile.

Key takeaway

Your ideal customer profile is right when a stranger could use it to find you ten real prospects by tomorrow afternoon. Anything less specific is a category, and categories do not reply to emails.

Frequently asked questions

Is it not risky to narrow this much so early? The opposite. Narrow is reversible in an afternoon; twelve months spent broadcasting to a category is not. You can always widen once one narrow group is working.

What if I genuinely have several types of customer? You probably do, eventually. Pick the one you can reach most cheaply this month and ignore the rest until it works. Running two profiles at once means you cannot tell which one produced a result.

How do I know when my profile is wrong? You get replies but no meetings, or meetings but nobody buys. That pattern usually points at the message or the offer rather than the market — the positioning statement is the usual culprit before the profile is.

Do I need interviews before writing one? No, but you need them before trusting one. Write the guess, then go and test it.

---

You now have the framework. The Mark Platform walks you through the persona builder field by field, with fill-in-the-blank formulas and real examples from three startups, then connects that profile to the offers, channels and outreach that come after it. See how the eight-step journey fits together, or start your marketing journey →