How CitedIndex evaluates tools

What we believe

Google has been the front door for software discovery for two decades. That door is closing. ChatGPT, Claude, Perplexity, and Google's own AI Mode now answer a growing share of “what's the best tool for X” queries themselves, picking a small set of sources rather than handing back a list of links.

Most software directories haven't adapted. They optimize for volume. Long lists of tools, shallow information per entry, no editorial judgment. AI search engines tend to skip directories like that. They lean toward sources that read like trusted reference works.

CitedIndex is a directory built around how AI engines pick those sources. A listing here adds an authority signal to your tool. It says the tool was checked against the same public bar as every other listing on the site, and published in a structure those engines can extract.

Who we list

CitedIndex lists software products and the businesses behind them. The product has to be maintained, have a working signup or download flow, and either public pricing or a clearly free tier.

The full inclusion and exclusion rules live on the scope page. We don't list info-products, agencies, gambling, crypto tokens, or physical-goods e-commerce. The scope is narrow on purpose.

What we do, and what we don't

We don't pretend to have personally used every tool we list. The index is large enough that one editor cannot run hands-on trials of everything in it, and saying otherwise would be the kind of thin credibility that AI engines and careful buyers see straight through.

What we do instead is build each listing from a structured read of the vendor's own pages: pricing, docs, customer logos, the changelog, and the FAQ section. Every fact-bearing claim traces back to a URL we read, and the source list sits at the bottom of the listing. The structure of the listing is the work product: what the tool does, who it's for, what it costs, where it falls short, and how it compares to the typical buyer's alternatives.

Where we have hands-on knowledge, we say so on the listing (“Editor's read” line, plus the “currently testing” surface on author pages). Where we don't, the editorial judgment is anchored in the data we read, not in invented experience. That bar applies whether the listing is free, Verified, or Premium.

The editorial bar

Every listing is built against the same editorial template designed for AI search citation. The exact template is proprietary, but every tool in the index meets the same bar:

• About the tool, not the company.Listings cover what the tool does, who it's for, what it costs, and how it works. We don't chase decorative info like founder names or funding rounds.
• Verified against the live site.Pricing, features, and product details on a listing match the vendor's own website at the time of our last check.
• Specific over vague.Real pricing tiers, real features listed verbatim, real integrations, real customer logos pulled from the vendor's own pages. No marketing fluff.
• Sourced from the vendor's site.Every fact-bearing statement traces back to a page we read on the vendor's own domain. The full source list sits at the bottom of every listing.
• Editor byline, not a faceless brand. Listings carry a real human byline. No synthetic personas.

Re-evaluation

Listings are re-evaluated when the methodology updates or when a founder writes to [email protected] asking for a re-check. Every published listing carries a machine-readable last-verified timestamp, so AI engines and humans can see exactly when the data was last confirmed against the live site.

Editorial independence

The site earns money from paid listings, Featured placement in category pages, and outbound affiliate links where they exist. None of that decides whether a tool gets included, where it ranks editorially within a category, or what its listing actually says.

Free listings get the same editorial write-up as paid ones. Verified adds a do-follow backlink (also unlockable for free by embedding our badge). Featured adds priority placement in the category and a dedicated blog post. Featured placements are clearly labeled as such.

Outbound affiliate clicks are disclosed in every listing's trust footer.

Versioning

The methodology is versioned. When the bar moves, older listings get flagged for refresh. They are either re-evaluated against the new version or marked as legacy. Every listing carries the methodology version it was last evaluated under, so you can always tell what bar a given listing met.

Questions about the methodology or a specific listing? [email protected].