Find the rooms your customer is already in.

The problem, as somebody put it

Spent half a year building a product and launching it and tried to just post a white paper on HN and Reddit to get some feedback: instant ban for self promotion. From what I've seen, it is not even clear that humans are involved in this process. So how should I proceed? Start paying for ads?

Hacker News, July 2026

The replies told him to buy an aged account with karma, or to do what one of them called shadow marketing.

That is the market's default answer, and it is the reason this exists. If you don't want to buy a fake account to talk to your own customers, your remaining option is to do it properly by hand — which is slow, and which nobody has time for while also building the thing.

What you actually get

Five things, delivered by a person and not a dashboard. The first two are free, the third is the $500 plan, and all five are the two-week engagement.

1. A ranked map of the rooms. Which platforms and which specific communities your customer is actually in — forums, Q&A sites, Discourse instances, subreddits, groups. Ranked by how alive they are, not by how big they claim to be. Most of the good ones are not the obvious ones.
2. Each room's rules on self-promotion, extracted and read. Including the unwritten ones and the ratio each place actually tolerates, so you know your budget before you spend it rather than after you're banned.
3. The interview plan your own evidence implies. Every claim your deck makes about your customer, turned into questions about their past — who specifically to ask, where to find them, what to listen for, and the answer that would prove the claim wrong, written down before you ask it. A question with no answer that could embarrass you isn't a test.
4. Twenty real objections, quoted, linked and dated. What people in your market already say is wrong with things like yours — in their words, not paraphrased into marketing copy. This is the part that changes what you build, not just how you sell it.
5. We sit with you while you post. You write it, in your voice, from your account. We tell you what will land and what will get you removed. You keep the relationship, the reputation and the account.

The Mom Test, in two halves

The rule in The Mom Test is that you must never ask people whether they like your idea, because they will lie to be kind. You ask about their life instead: what they already do, what it already costs them, what they have already paid to make it stop.

Half of that needs no interview, and that half is free. The other half is the interviews, and that is the half you are paying a person for.

Those answers already exist. Your customers wrote them down — in forums, in threads, in replies to each other — unprompted, before anyone was selling them anything. Nobody was being polite to a founder. So the first pass of that research is not an interview at all; it is a measurement over what they already said. That is the part we automate, and it is why the deliverable is quotes with dates and links rather than a summary.

What it does not replace: talking to people. So the paid half is the part that gets you into those conversations properly — which specific people are worth an hour, where to find them, and the exact questions, each with the answer that would tell you you were wrong. The measurement is what makes those questions worth asking: your phrases come back with counts beside them, and a phrase your market has never once written in public is the most useful thing on the page. And once the rooms are known they don't stop producing — the same rooms can be watched afterwards, so what your market complains about keeps arriving while you're building rather than at the next planning cycle. That part is monthly, and it is only worth buying after the verdict says these rooms exist.

What it costs

Fixed scope, fixed price. Sometimes the verdict is no — not in public, not anywhere we can read. That answer is free, you get it in writing, and it is the one this whole thing exists to hand over early rather than late.

Why the free one first

Because the answer is not always yes, and you should not have to pay to find that out.

The same tooling ran against two audiences in the same week, on the same free public sources. For a niche trade — California balcony-inspection providers — it found zero verbatim matches for the phrases that trade uses, across the whole of Hacker News, and only 2 of 12 candidate forums were even measurable. For founders with no distribution, it found those phrases thousands of times, and 10 of 16 forums ranked cleanly.

Nobody typed either result in. Same week, same method, same constraints — and the two audiences differ by four orders of magnitude. Two of the first three audiences this ran on were the first kind, which is exactly why the measurement is free now: it used to be the $500 purchase, and charging $500 to tell somebody their market is unreadable is a refund conversation, not a business. If yours turns out to be the first kind, you will know before you have paid anything — and what the $500 then buys is the set of conversations to have instead, since for an audience nobody can read in public, asking people directly is the only method left. I'll walk you through all three runs, live, on a call.

Platform-agnostic, on purpose

Platforms close. Reddit priced out its API, Facebook retired Groups access, Substack forbids crawling outright. Anything built on one platform is one policy change from being worthless.

So the first question we answer isn't how do we win Reddit. It's which rooms are worth your time at all — and that answer is usually spread across places you haven't heard of. There are more than twenty thousand independent community forums that are open, active, and unclaimed by anyone selling to the people in them.

Who this is for

Being honest about the stage

This is new. There is no customer list to show you, and inventing one would be exactly the behaviour this whole thing is against. Nobody outside my own circle has paid for this yet. You would be first, and you should price that in.

What there is: a lot of primary research into where these communities are and what gets people banned in them, and three engagements run end to end. The first was Nuvea, our own product — aimed at California balcony-inspection providers, though its homepage still says US remodeling contractors, and that gap is itself one of the findings. The second was this: the method turned on its own audience, which is how the comparison above exists. The third was somebody else's product, from their deck, and not one question in it was written by me — it came back a clear no, which is the run that changed the price.

There were two free places. Both went to products I own or a friend's, which taught me how to deliver the work and precisely nothing about what it is worth. So the free places are closed. $500 is the ask, and the reason the measurement in front of it is free is not generosity — it is that a free "yes, this looks useful" tells neither of us anything, so the free part is deliberately the part that costs me almost nothing to give and the $500 is the first thing that means something when you pay it.