We are a small research desk attached to a backlink-data API. Our job here is to test claims about AI search against link numbers, and to say when the numbers are silent.
AlgoLink research team ·
What we are
AlgoLink is the research-facing side of a backlink API. The same product sits behind it: a REST service with JSON responses, our own crawler, and our own link index of 181,132,097 domains.
The articles are written by the AlgoLink research team. We do not use invented personas, and we do not publish customer stories, because we have none we could stand behind.
Why this angle? People who ask how AI answers choose sources tend to want evidence, and link data is one kind of evidence we happen to own.
Our focus is AI search research for people who need to explain results to a client or a boss. Plain evidence travels better than hype in those rooms.
Corrections are welcome and we log each one. If a date, number or claim on this site is wrong, tell us which page and what the evidence says.
How we try to behave
We label guesses
Each claim on the site is stated, observed or guessed, and we try to say which.
We date things
Launches, updates and our own checks carry months, so you can judge age.
We name our limits
The index refreshes about monthly. We do not track keywords or ranks, and for that Ahrefs and Semrush are fuller.
We prefer small tests
A cheap experiment that you can repeat beats a big claim you must trust.
What we publish, and what we leave out
Every page here starts from a question people really ask about AI search: why one page gets cited and another does not, whether links still count, and how to check a claim without buying a tool.
We write the answer first, then the caveats. Where the honest reply is that nobody knows, we say so in the first lines instead of padding with advice that sounds plausible.
We do not sell rankings or citations. We do not run an agency, and we do not promise placement in ChatGPT, Perplexity, Claude or Google AI Overviews, because no outsider controls that.
The pages that carry most weight are the source claims table, which sorts statements by evidence, and the timeline, which anchors events to months. The glossary keeps our vocabulary fixed.
Using the data
You can look up any domain for free, with no key. Heavier pulls are billed from $0.0001 per credit, and a Google sign-in gives $100 of credit with no card.
If you want to start with a problem, the experiment list and the cited-domain pull are the two pages most readers use. Send corrections through the account page, and we will fix errors we can verify.
Where the link data comes from
Our own crawler walks the open web and our own index stores the result: 2,774,625,913 domain-to-domain links and about 798.5 billion backlinks in total. We do not resell anyone else's database.
The index is rebuilt roughly once a month. That makes it good for comparing sites and tracking slow change, and poor for catching a link the day it appears.
DR is our own 0 to 100 score from the count of referring domains, on a log scale. It resembles Ahrefs DR or Moz DA in spirit, but it is a different number, so do not mix them in one chart.
We do not run ads on these pages, and we do not accept paid placements in the articles. If that ever changes, the page will say so near the top.