One question, answered with evidence you can check.

You name one market or product idea. We assemble what the public record actually says about it — the recurring complaints, who is affected, what they already pay for, and where the evidence runs out — then watch it for four weeks and talk it through with you.

What you get

One deep Evidence Brief — the eight questions, each answer tied to dated source excerpts with links. Where the record is silent, the brief says so instead of filling the gap.
Four weeks of monitoring — we tell you when something material changes, and stay quiet when nothing does.
A 45-minute call to walk through it, including what we could not establish and what you would have to test yourself.

Scope, frozen up front

One question. Not a market landscape, not a competitor sweep across a category.
One language version of the brief (English or Chinese — your pick).
A capped number of evidence items; the cap is set after the first engagement and published here before we take a second.
One 45-minute call and one round of small revisions.
Work is capped at 8 hours. If the question turns out to be bigger than that, we narrow the scope with you — we do not silently deliver less or quietly work longer.

Two checks before we start

Does it survive without Reddit? Reddit is our largest single source and the one with the most restrictive terms. If the core finding only holds because of Reddit, we say so before you pay — a conclusion resting on one platform is a conclusion with a single point of failure.
Source-by-source rights check. Every source runs through our rights matrix before anything is quoted. YouTube, for example, is metadata and links only — no quoted comment text, regardless of how useful it would be.

If either check fails, we tell you what we can and cannot deliver, and you decide before any money moves.

$249 per engagement · first 3 only

This is a founding price for the first 3 engagements while we learn what the deliverable should be. We run at most two at a time, so the queue is real.

Email us your question

Tell us the one question you want answered. We will reply with the scope and how to pay.

Questions people ask first

Is this a report generated by a model?

No. Language models extract candidate signals from public posts; the counts, thresholds and status are deterministic code; the brief is assembled and reviewed by a person. Every claim links to a dated source you can open.

What if the evidence is thin?

Then the brief says the evidence is thin, names what is missing, and proposes what you would have to find out yourself. A thin answer delivered honestly is the outcome we would want if we were paying.

Do you tell me whether to build it?

No. We can show that a problem is real, recurring and connected to spending. Market size, reachability, acquisition cost and whether it suits you are outside what public posts can establish.

Can I see the format first?

Yes — the public dossiers are the same evidence discipline at lower depth. The brief goes further on one question and adds the gaps and the next validation steps.