By Dairon Canel · Founder, Proven · Updated June 2026
Process
How Proven works
Three steps. About four minutes. A verdict with sources you can verify.
You describe the idea
Paste anything from a single sentence to three paragraphs. The more specific you are (who has the pain, what triggers it, what they've tried), the more targeted the evidence will be. Proven doesn't need a business plan. It needs enough to know what market signal to look for.
The description you write is not scored or evaluated as writing. Proven uses it as a search brief: the raw input from which it derives the pain signals, keywords, and adjacent topics to go after. A description that focuses on the specific frustration (“agencies spend 2 hours a week chasing client approvals by email”) will return sharper evidence than one that describes the solution (“a tool for client communication”). Describe the pain, not the product.
If you're unsure how to frame it, the most useful format is: the person who has the pain, the moment it hits, and what they typically do about it today. That's enough for Proven to go find whether the market is expressing it.
We go find the evidence
Proven fetches live posts from Reddit, Hacker News, and App Store reviews. Not cached. Live at the moment you submit. The pipeline identifies the relevant communities and searches them for posts that describe the pain you outlined.
Each post is scored for pain intensity, specificity, and market signal. A post where someone describes losing a client because of a specific operational failure scores higher than one that says “this process is annoying.” Vague frustration gets filtered out. The specific, emotionally charged descriptions survive. Proven analyzes 150+ posts per brief.
While that runs, Proven maps competitors: the tools, workarounds, and alternatives people mention when they describe the pain. It builds the persona from demographic signals in the posts, pulls out the feature requests buried in the complaints, and maps the risk patterns in the evidence. All of it runs in parallel, which is why the brief is ready in about 4 minutes.
You read the verdict
A 10-section Founder's Brief lands in your dashboard. Pain score, evidence, competitors, persona, features, GTM plan, pricing signals, risks. And a top-level verdict: Build it, Explore more, or Kill it. Every finding links to the original post so you can judge the evidence yourself.
The verdict is not a recommendation from an algorithm optimized to make you feel good. It's a conclusion drawn from what the evidence actually shows, which is how startup idea validation is supposed to work. A weak Pain Score, thin signal volume, and a clear competitor with strong reviews will produce a Kill or Explore verdict, even if your description sounded compelling. A high Pain Score, evidence of competitor gaps, and clear willingness-to-pay signals will produce a Build verdict, with the specific communities and angles to pursue first.
The brief does not replace your judgment. It gives your judgment something real to work with instead of the pattern-matched optimism you get from asking a language model whether your idea is good.
Why Reddit, HN, and App Store reviews?
These three sources have the highest signal-to-noise ratio for what validation actually requires. Reddit is where people describe operational pain in detail. Not to network or promote. Because they want help, or want to vent. The specificity in Reddit threads is unusually high compared to most public platforms.
Hacker News attracts a different audience: developers, technical founders, and early adopters who describe problems at a more systemic level. HN threads about a given pain often surface the “why does nobody solve this properly” perspective, which is exactly the signal a founder needs before choosing a wedge.
App Store reviews are the most underused source for B2B validation. When someone takes the time to write a review of a competitor's product, they're describing what they needed that they didn't get: specific features, missing integrations, pricing frustrations, support failures. That's competitive intelligence and customer insight in one place, and most founders never read it systematically.
Frequently asked questions
What if there are no Reddit posts about my idea?
Your brief will reflect it honestly: a low Pain Score, low signal volume, and a verdict of 'Explore more' or 'Kill it.' This is not an error; it is a real finding. If the market is not visibly expressing this pain in public forums, that is information worth having before you build. Reframe the pain description and run another brief to see if a different angle returns stronger signal.
Which subreddits does Proven search?
Proven does not search a fixed list of subreddits. The pipeline identifies relevant communities based on the pain you described. It searches both broad startup and SaaS communities and the niche-specific subreddits where your market is most likely talking. This means different ideas surface different subreddits, which is part of what makes the evidence specific to your market rather than generic.
Can Proven search LinkedIn or X (Twitter)?
Not currently. Reddit, Hacker News, and App Store reviews are the highest signal-to-noise public sources for market pain. LinkedIn conversations are largely promotional and network-driven. X is high-volume but low-specificity for validation purposes. Proven uses these sources because they're where founders, developers, and early adopters describe problems in detail, which is the signal that matters for validation.
How current is the data in my brief?
Proven fetches posts live at the moment you submit, not from a cache or a snapshot. The evidence in your brief reflects what was on Reddit, Hacker News, and the App Store on the day you ran it. This means briefs for the same idea submitted months apart can return different evidence, which is useful for tracking how a market is evolving.
Does Proven search my exact idea or related topics too?
Both. Proven searches your specific framing and expands into related pain signals, competitor names, and adjacent topics the pipeline finds relevant. The goal is to surface what your future customers are actually saying, which is often in different language than how founders describe their own idea. A description about 'async project updates' might surface posts about 'too many meetings' and 'status update fatigue'. The same pain, different vocabulary.
Learn what's inside a Founder's Brief or understand how the Pain Score is calculated.
Written by Dairon Canel with AI assistance for research and structure.