Research
Open methods, published in full
Most visibility claims cannot be checked. These can. Each piece below is either a protocol you can run yourself or a catalogue you can verify yourself — with the version, the date, the limits, and the sources stated on the page.
The Local Visibility Signal Catalogue
A working reference for the 36 signals that decide whether a service business is found, understood, and chosen. Each entry states what the signal is, what it costs you when it fails, the exact steps to check it with free tools, and what a passing result looks like.
Read the full methodThe AI Visibility Benchmark
Everyone sells AI visibility. Almost nobody publishes how they measure it. This is the full protocol: 15 buyer prompts, three assistants, a five-point rubric, and the session rules that make two runs comparable. Free to run, on any business, including ours.
Read the full methodEditorial rules
What we will and will not publish here
Research is only worth linking to if the reader can check it. These four rules are what make that possible.
Runnable by anyone
Every method here can be executed by the reader with free tools. If a check needed a paid product, it was rewritten or left out.
No invented numbers
Where a figure would make a piece more impressive, it is either measured and dated or it is absent. A fabricated benchmark is exactly the defect our audits exist to find.
Limits published, not buried
Each study states what it cannot tell you before anyone has to ask. A method that claims more precision than it has produces confident nonsense on the next run.
Free to reuse, including against us
Quote it, adapt it, run it on a competitor — or on Found Clearly. Attribution and a link back is the only condition.
Prefer we run it on your site?
The snapshot applies the same catalogue to your own pages and shows you the first gap worth investigating — with the evidence behind it.
Evidence-based scores, not guarantees.