Most software directories look the same. A logo, a star rating, a wall of user reviews, and a number that claims to tell you exactly how good the product is. 8.7 out of 10. 4.3 stars. "Ease of use: 9.2."
We think that's a broken way to help people choose software. Worse, it's dangerous for the founders who get ranked by those numbers.
This post is for anyone who has ever looked at a directory listing and wondered: where did that score come from? And it's for founders who are tired of watching their product get boiled down to a decimal point they can't control.
The three tiers of truth
At directree, every single field on every listing carries a provenance label. There are only three kinds of information we show, and we mark each one clearly so you always know what you're looking at.
Observed — the facts we can verify
Observed data is anything we can crawl, check, or measure directly from a tool's website or public properties. Does the pricing page list a free plan? That's Observed. Are there screenshots on the landing page? Observed. Social links, integration logos, feature lists, demo videos — if we can point to a URL and confirm it, it goes in this bucket.
We render Observed facts in a monospace typeface. That isn't a design quirk. It's a signal. When you see mono text on a directree listing, you know you're looking at something we pulled straight from the source and did not embellish.
This tier matters because it's the bedrock. If we say a tool has a free trial, it's because we saw a "Start free trial" button. Not because an AI guessed. Not because a salesperson told us. We looked, we recorded, we labeled it.
AI-inferred — helpful, but never disguised as fact
The second tier is where our enrichment pipeline does its work. When someone pastes a URL into directree, our system builds a structured listing in about thirty seconds. A lot of that structure comes from reading the page, but some of it comes from interpretation.
What is this tool best for? Who should probably look elsewhere? What are its main strengths and weaknesses? Which competitors does it most often get compared to?
These answers are useful. They help buyers navigate a crowded market. But they are not facts. They are guesses — educated guesses, grounded in the text of the site, but guesses nonetheless.
So we label them AI-inferred, always, and we use a rose-tinted badge to make the distinction impossible to miss. We never let an inferred summary sit next to an Observed fact without a label. That would be lying by layout.
This is where a lot of directories go wrong. They let AI-generated prose blend into the page as if it were editorial copy written by a human who tested the product. It wasn't. At directree, we think you deserve to know that.
Founder-edited — the highest trust tier
The third tier is what happens when a founder claims their listing and verifies ownership. They get editing rights. They can correct mistakes, fill in gaps, and add details our crawl missed.
When a founder edits a field, it gets a gold Founder-edited badge. That badge means a real person with verified control of the product has stood behind this information.
This is the strongest signal on the platform. Not because founders are inherently more honest than crawlers — though they usually know their own product better — but because a verified claim creates accountability. If the pricing is wrong and the founder hasn't fixed it, that's a choice. If an AI gets the pricing wrong, that's a bug. The two are not equivalent.
Why we refuse fake precision scores
Here's where we part ways with most of the industry.
We do not assign numeric scores. No 8.7. No 4.3 stars. No "ease of use" deciles. We don't do it because the precision is fake and the harm is real.
A single number implies a level of measurement rigor that almost no directory actually has. Did the reviewer run a controlled benchmark? Did they survey a representative sample of users across company sizes and use cases? Did they weight the features by what matters to each buyer persona? Of course not. The number is an opinion dressed in math.
For buyers, fake precision creates false confidence. Someone picks the 8.7 over the 8.3 and feels like they made a data-driven decision. They didn't. They picked the louder marketing budget.
For founders, it's worse. A young product with five reviews can get buried beneath an incumbent with five hundred, not because it's worse, but because the scoring algorithm was built to reward volume and seniority. The number becomes a moat for entrenched players.
We use qualitative bands instead. A tool is either a strong fit for a given use case, or it isn't. The community can upvote tools they like. Our surveys let the crowd pick favourites in specific categories. But nobody gets reduced to a decimal.
What we learned building this way
We've been building directree in public since day one, and the honesty model has already saved us from at least one bad decision.
Earlier this week we were building a GEO checker — a tool that tells founders whether ChatGPT and other AI search engines are likely to recommend their product. Our first version almost shipped with leading questions designed to push every result toward a glowing 100% score. It would have felt good for users. It would have been useless.
We tore it out. The live version now gives an honest, nuanced read based on real signals: whether your site is indexed, whether your structured data is clean, whether your product appears in relevant AI responses. Some scores are high. Some are low. All of them mean something.
That same instinct runs through everything we build. Our Product Hunt auto-lister doesn't scrape maker emails or spam founders. It reads public launch data, runs it through our own enrichment pipeline, and creates an honest listing with the same three-tier labels as every other tool on the site. If the founder wants to claim it, they can. If they don't, the listing still carries the truth labels.
Even our compare pages follow the rule. We only let search engines index comparison pairs that pass a real demand filter. Every other comparison still works for users — you can compare any two tools in the same category — but we don't pretend a zero-traffic page deserves Google's attention. Honesty applies to SEO too.
Where directree fits in your stack
If you're a founder trying to get discovered, you already know the playbook: build something useful, write about it, earn backlinks, show up in AI answers. We've written about SEO for startups and getting found on AI search in detail.
A directree listing is one concrete step in that playbook. Paste your URL, get a structured page in about thirty seconds, and every fact on it is labeled honestly. If you claim and verify your listing, you get a do-follow backlink badge — real SEO value, not a no-follow shrug. And because our pages are structured and labeled, they show up in AI answer engines too.
It's free. It takes less than a minute. And it doesn't ask you to game a fake scoring system.
FAQ
Q: Can I trust AI-inferred information on a directree listing?
A: You can trust that it's clearly labeled and grounded in the source text we crawled. It's a starting point, not a verdict. Always click through to the tool's own site before making a decision.
Q: What happens if a founder edits their listing to hide something?
A: Founder-edited fields are labeled as such, so you know who wrote them. Observed facts from our crawl remain visible alongside the founder's version. If there's a conflict, you can see both sides. We also have a report path for misleading claims.
Q: Why don't you show user review scores like other directories?
A: Because we haven't built a review system we trust yet. Forced reviews create garbage data. Incentivised reviews create bias. When and if we add reviews, they'll follow the same honesty rules: transparent sourcing, no hidden weighting, no fake precision. Until then, we use upvotes and community surveys as cleaner signals.
Q: Does the honesty model hurt directree's own SEO?
A: We think it helps. Google has been explicit that low-quality, templated content gets demoted in search. A directory where every fact is sourced and labeled is the opposite of that. Structured, honest data also performs better in AI search answers — which is why we built our GEO checker in the first place.
Q: How do I claim my listing?
A: Find your tool on directree, click the claim button, and verify ownership via a domain badge or email. Once verified, you can edit any field and your changes get the Founder-edited label. It takes about two minutes.
Want to see the honesty model in action? Submit your tool and watch how we label every field — or browse any listing on the site and look for the badges.