A figure is only as good as its source, so each one gets a page. What the source is, who runs it, why it can be relied on, exactly which numbers are taken from it, and what it cannot tell me. Nothing in the statistics is a number that was made up, and nothing is a number that cannot be pointed at.
Written by Miles Foster, Owner and tester, RankDude.
01
Search visits, visits by country, searches by name, popularity, links and Google positions, pulled from the same sources in the same order every month in every category, so brands can be compared instead of described. That is the category monitor.
02
In the first category, YouTube transcripts, the Reddit threads Google surfaces and the press about each brand were read in full and classified. Two of the sources below are for that; the other categories carry the measured figures only.
03
None of these sources grades anything. RankDude ranks only the tools I have tested myself, with the procedure on each category's methodology page; the statistics monitor each category around them.
Visits, searches, popularity, links and Google positions. Each source is pulled on the same schedule for every tracked brand in every category, and every figure in the statistics names the source and the month it covers.
DataForSEO
Search data API
The search data behind most numbers on this site: traffic estimates, visits by country, referring domains and live Google results.
Google Ads keyword data
Search volume data
Google's own count of how many people search a term each month, which is how I measure demand.
Tranco
Domain popularity ranking
A research-grade ranking of the internet's most popular domains, built to resist manipulation, used here as the all-traffic signal.
Google Chrome UX Report
Real-user dataset
Google's own popularity tiers, taken from how real Chrome users browse, used as a cross-check on Tranco.
Ahrefs
Backlink index and crawler
One of the largest independent web crawlers and backlink indexes. I use exactly one number from it: Domain Rating.
SerpApi
Search results API
Fetches what Google and YouTube actually showed: Google Trends for every category, and for AI girlfriend the videos, news and threads about each brand.
Google Trends
Relative search interest
The shape of interest over time: when a term rose, when it peaked, where it is searched most.
Not tools so much as sources. For AI girlfriend brands, the most-watched videos were read through their transcripts and the Reddit threads Google surfaces with their top comments, each classified after reading it in full and recorded in the brand’s data file. The other categories do not carry this reading yet.
YouTube
Primary source, user-published
Where most hands-on reviews of AI girlfriend products live. For that category I read the transcripts, not the titles.
Primary source, user-published
What users say when no one is selling anything. For AI girlfriend I read the threads directly, with the votes and comments attached.
RankDude is not affiliated with any of these providers. Ahrefs data is limited to Domain Rating under its public licence. Every category’s statistics pages use these sources; see them in use on the AI girlfriend statistics pages, for example. Written by Miles Foster.
