Proprietary · Hellion internal
Gap Finder
We built our own market research engine, and it tells us what to build.
Every Monday, Gap Finder gives us a ranked list of real, evidence-backed problems that cost someone money or time. It reads the public web — reviews, forums, complaints, job listings — extracts the pain, groups identical problems across sources, and hands us a leaderboard split into two lanes: build a product, or sell a service to the businesses feeling it.
Attribase came out of that leaderboard. So did Locals Club. So will the next one.
Why we built it
We got tired of building things nobody asked for.
Most product studios pick what to build the same way: someone on the team has an idea, everyone convinces themselves it's good, and six months later they're trying to find the customer that was supposed to already exist.
We built Gap Finder to invert that. Instead of starting with the idea and hunting for the customer, we start with the customer's pain — the actual complaint, in their own words, on the public record — and only build if the evidence keeps stacking.
Two lanes
Every problem sorts into one of two lanes.
Product lane
Problems worth building a product around — where the pain is common enough, painful enough, and unsolved enough that a shipped product would find its audience. This is where Attribase and Locals Club came from.
Service lane
Problems worth selling a service into — where affected businesses would pay us to fix it for them one at a time. This is what our client services team follows: local SEO, AEO, custom software, automation.
The evidence gate
Every entry has to clear three thresholds.
A problem doesn't reach the leaderboard just because it made noise. It has to prove it's real, widespread, and current.
3+
distinct authors
A problem doesn't reach the leaderboard until at least three real people, in three separate posts, have complained about it.
2+
platforms
The signal has to show up in at least two different sources. A single subreddit isn't evidence — it's an echo chamber.
365d
recent activity
The problem has to be alive right now, not something that peaked in 2019 and disappeared.
What we don't do
The rules we set for ourselves.
Nothing gets on the list without evidence.
No hunches, no "I have a feeling about this space." Every entry has real posts, real timestamps, real authors we can point to.
We never sell the raw data.
Gap Finder is not a data product. We do not resell what we collect. We do not train models on it. The output is the leaderboard we use to decide what we build.
Only public web, only what people said in public.
No scraping behind logins. No private forums. No leaked data. If someone wrote it publicly, it counts. If they didn't, it doesn't.
Want a gap report for your industry?
We'll run Gap Finder on your space.
Tell us the industry or vertical you're in. We'll run Gap Finder on it and send back a sanitized report of the top 5 evidence-backed problems in your space — the kind of thing you could build a product or a service around. Free, no strings.