Where you are · 6 min read
Some users, and no growth
A flat line with real users is usually not an acquisition problem. Check retention first: if the people you have do not come back, more of them changes nothing. Then check repeatability: if you cannot name how your last ten users found you, you have nothing to do more of.
Forty users. Maybe a couple paying. New signups arrive at a trickle that neither grows nor stops, and it has been like this for two months.
This is a more interesting position than it feels like. Unlike a founder with zero users, you have signal. Real people chose to try this, some of them stayed, and both of those facts are readable. Almost nobody reads them, which is why the default response to a flat line is to go and acquire harder.
Do not start with acquisition
Acquisition is the expensive answer and usually the wrong one at this stage, because it multiplies whatever is already happening. Before adding people, find out what happens to the people you have.
More users into a leaking bucket is not growth. It is a faster way to run out of prospects.
Question 1: does anybody come back?
Take the people who signed up four weeks ago. How many used it in the last seven days? You do not need a cohort tool for this, a database query and a spreadsheet is fine.
- Most of them are gone. You have a retention problem, and it is a product or onboarding problem rather than a marketing one. Stop all acquisition work and go and find out what happened between signup and disappearance.
- A small group keeps using it. This is the best possible finding and it is the whole game. Those people are your product-market fit in miniature, and everything else on this page is about doing more of whatever produced them.
If a handful of people use it every week without being chased, you have something real. The job is no longer to invent demand, it is to find more people who look like those users. That is a much easier job and it needs completely different work.
Question 2: can you name how the last ten arrived?
Not the channel in the abstract. Ten names and ten routes. A specific thread, a specific person who told them, a specific search, a Slack group.
Most founders at this stage cannot answer, and it is the reason the line is flat. If you do not know what produced user thirty-five, you cannot do it again on purpose, so growth arrives as weather rather than as a result.
How to find out, this week
- Ask them. One question by email: how did you come across this? Ten sent, four or five answers, and you will be surprised by at least one.
- Add the question at signup. One optional free-text field. Crude, self-reported, and far more useful than referrer data once attribution has been stripped.
- Read your referrers before dismissing them. Direct traffic hides real routes: a private Slack, a newsletter, a podcast mention you never heard about.
Question 3: are the ones who stayed the same kind of person?
Look at the retained group and ask what they have in common. Not demographics. Situation.
Very often the answer is narrower than your marketing. You built a project management tool and the people who stayed are all agencies billing hourly, or you built a developer tool and everyone who stayed works alone. That pattern is worth more than any keyword research: it tells you which words belong on your homepage and which rooms to be in.
This is also the moment most founders resist, because narrowing feels like shrinking the opportunity. It is the opposite. The pre-fit sequence depends on taking your positioning from the people who actually stayed.
What the three findings each imply
- Nobody retains. One move: talk to five people who left. “What were you hoping this would do that it didn’t?” Do not defend, do not demo, collect the answer. Acquisition work is wasted until this changes.
- Some retain, route unknown. One move: find out how the last ten arrived, then repeat the most common route deliberately ten more times.
- Some retain, route known. One move: do that route again, at three times the volume, and measure. This is the only situation on this page where scaling is the right instinct, and it is the situation almost nobody checks for before scaling.
How to run the conversation with somebody who left
This is the highest-value half hour available to you and most founders get it wrong in the same way, by defending the product. You are not selling. You are collecting.
- Ask for the story, not the opinion. “What were you trying to do when you signed up?” beats “what did you think of it?” People are unreliable critics and excellent narrators.
- Then the one question. “What were you hoping this would do that it didn’t?” Ask it, then be quiet for longer than is comfortable.
- Ask what they do now instead. The answer is usually a spreadsheet, a group chat, or nothing, and it tells you what you are actually competing with.
- Do not demo, do not explain, do not offer a discount. Every one of those ends the flow of information and replaces it with politeness.
- Write it down verbatim. Their exact words are the asset. Paraphrasing turns their language back into yours, which is the problem you are trying to solve.
Five of these conversations will usually rewrite your homepage for you, in the customer’s own words, which is the only version that reliably works. That is a better return than any other half hour you could spend this month.
The awkward case: retention is fine, nobody new arrives
A small group of people who genuinely rely on it, and no new signups. This is the most frustrating position on the page and the most promising, because the hard part is already solved.
It is now purely a distribution problem, and the retained users are the instrument for solving it. Three moves, in order:
- Ask the retained users where people like them congregate. Not where they found you. Where they talk shop. They know rooms you have never heard of, and they will tell you if you ask.
- Ask them for the sentence. “How would you describe this to a colleague?” Put that on the homepage, replacing whatever you wrote.
- Ask each of them for one introduction. Not a referral programme. One person each, by name. This is the highest-converting traffic that exists and it is available to you today.
Flat is not failure, it is a reading
A flat line means your current inputs produce exactly this. Nothing is broken and nothing is compounding. It will stay flat until one input changes, which is genuinely good news: it means a single deliberate change is visible against it, and you have a clean baseline to measure against.
What ruins that is changing five things at once, which is what a marketing plan usually is. One change, one measure, one reading. Then the next.
If you would rather have this diagnosis run against your actual site and stage, the Product Check is free and needs no account. It is the same sequence: what is the constraint, what is the one move, and what would count as it having worked.