Website traffic rankings are all estimates. They don't have to be.
Every public ranking of website traffic is a guess.
SimilarWeb models it from panel data and clickstream. Ahrefs infers it from keyword positions. Every "fastest-growing sites" list is built on one of those two, and the people who publish them will tell you as much if you ask.
The real number exists. It sits in Google Search Console, in each site owner's account, and nobody has ever had a reason to open it.
That's the whole idea behind realrank.lol: a public leaderboard where your position comes from clicks Google actually recorded, not from anyone's model of them. You connect Search Console, choose which verified properties to publish, and the data does the ranking.
Two design decisions worth stating, because they're the arguable parts.
It ranks on momentum rather than volume recent clicks measured against the preceding period. Ranking by raw size would produce a list of large sites that never changes, which is a directory, not a leaderboard. Momentum means a small site growing fast can sit above a large one standing still. It also means the top moves every week, which I think is the point.
And publication is the site owner's call. Verified data only works if people hand it over voluntarily, which means the whole thing lives or dies on whether being measured accurately is worth more to people than being ranked favourably.
I don't know yet whether it is. That's the interesting question and it's why the thing exists.
realrank.lol — if you've got a site with Search Console access, you can connect it and see where you land.
The instinct is right, but I’d narrow one part of the claim to make it bulletproof: GSC clicks aren’t ground truth for total traffic; they’re an exact measurement of one traffic channel which is Google Search.
They don’t include direct, referral, or social traffic, and GSC’s privacy thresholding can quietly remove some long-tail queries from the reported rows. The number can also differ depending on whether you’re looking at a domain property or a url-prefix property.
That’s still a big upgrade over modeled estimates. I’d just label it “measured Google Search clicks” and put that definition right next to the leaderboard rather than burying it in a footnote.
The second issue matters even more for the ranking itself: momentum heavily rewards small denominators. A site going from 4 → 12 clicks has tripled and can outrank one going from 40,000 → 44,000, even though the latter gained 4,000 clicks.
I’d publish the base alongside the growth rate. Otherwise the leaderboard risks turning into a small-site lottery.
And the voluntary publication point you touched on is actually pretty interesting. You’re not really ranking “websites” you’re ranking the subset of sites whose owners are willing to be measured and publish the numbers. I’d say that explicitly on the page too. It’s a selection effect, but arguably a useful one (it filters for exactly the kind of people you want participating).