Site traffic estimator: five things the number is made of

A site traffic estimator gives you one big number for any website. That number is assembled from smaller ones, and each piece has its own weak spot. Here are the five things it is made of, in the order they are stacked.

1. A list of rankings

The estimator starts by finding every Google search the site shows up for, and where it shows up. A site might rank first for a few phrases, somewhere on page one for hundreds, and on page two or three for thousands.

These positions come from snapshots of search results taken from time to time. Rankings move daily, so the list is always a little out of date.

Why it matters: most clicks go to the top three results. A site with a huge keyword count that sits mostly below position 10 gets far less traffic than the count suggests. seogeoaeo.ai's Domain Overview shows this split directly: first place, 2 to 3, 4 to 10 and below.

2. Search volumes

Next, each phrase gets a number for how many people search it per month. Those numbers are averages across months, so a seasonal phrase looks steadier than it is.

Why it matters: brand names are a weak spot. Searches for a company's own name are small and noisy in these datasets, so estimators often undercount a site's branded visits, which are usually its best visitors.

3. A click curve

Then the estimator decides how many searchers click each position. It uses an average curve: first place gets a large share, third gets less, tenth gets little.

A worked example with round numbers: third place on a phrase searched 10,000 times a month, at about 10% of clicks, comes out as 1,000 visits.

Why it matters: real results pages are crowded with ads, AI Overviews and rich results that push organic listings down. An average curve does not see any of that.

4. A sum

All those modeled visits are added together. That total is the headline figure you see.

Why it matters: each input above has its own error, and adding thousands of them does not cancel them out. The total is best treated as a size class, useful for saying "this site is about twice as big as that one", not "this site had 41,000 visitors".

5. A blind spot

The last ingredient is what is missing. A ranking-based estimator only models organic Google search. It cannot see:

• direct visits and bookmarks
• social media, email and newsletters
• visits from AI assistants such as ChatGPT, Perplexity or Gemini
• paid clicks, except as a separate ad estimate

Why it matters: a brand with a loyal audience can look small and still be busy.

How to use the number anyway

Three habits make an estimate useful:

• Test it on yourself. Put your own site into the same estimator on the same day and compare with your real Google clicks in Search Console. Now you know how far off it runs.
• Follow the trend. A figure that doubled in six months means real ranking gains.
• Stick to one tool. Different tools use different rankings, volumes and curves.

Where to run one

Our pick is Domain Overview on seogeoaeo.ai. You enter a domain and one of 20 countries and get estimated Google visits, ranking keywords by position, what those visits would cost as ads, the site's own ad spending estimate, and its link profile. One run is 50 credits, and plans start at $29 a month for 2,000 credits.

If you only want a free first look, Ahrefs has a free traffic checker for organic estimates. And for your own site, skip estimators completely: your analytics and Search Console count the real thing.