How much traffic does a website get? What you can know, and what you can't

You want to know how many people visit a website: a competitor, a shop you’re about to buy from, a site you’re thinking of acquiring. Search for it and you’ll find tools that print an exact-looking number: “1.2M visits last month”.

Here’s the thing to keep in mind: only the site’s owner knows its real traffic, from its own analytics. Everyone else is estimating. Some estimates are good; none are measurements. Knowing where each number comes from is the difference between a useful signal and false precision.

Where outside numbers come from

There are three kinds of outside source, and they measure different things.

Public popularity lists

Researchers and browser makers publish ranked lists of the most popular sites. Two are especially useful because they are free, documented and updated regularly:

• Tranco is a research ranking created by academics to fix a problem with older top-site lists: they were easy to manipulate and changed wildly from day to day. Tranco combines several underlying rankings over a 30-day window, which makes it steadier and harder to game. It ranks the top million sites and is published daily.
• The Chrome UX Report comes from Google and is based on real Chrome users who have opted in to share usage statistics. Instead of exact positions it puts sites into buckets by order of magnitude: top 1,000, top 5,000, top 10,000, and so on up to the millions. It’s published monthly.

These lists rank sites against each other. They don’t say how many visits a site gets.

Commercial estimates

Tools such as Similarweb, Semrush and Ahrefs publish traffic estimates. They build them in different ways: from panels of users who share browsing data, from data partners, or, for search traffic, by combining a site’s search rankings with how many people search for each term. They can be very useful for comparing sites in the same market. But they are models, and different tools often disagree by large margins about the same site, especially smaller ones.

The site itself

Some sites publish their numbers: “two million readers a month”, a media kit, a screenshot of analytics. Treat these like any claim a business makes about itself: possibly true, impossible to check from the outside. If you’re buying a site, ask for read-only access to its analytics rather than screenshots.

Why we show bands, not visit counts

When we check a site’s popularity, we show the band it falls into, the best of the public lists we use:

Next to the band we show each source, its rank or bucket and the date of the list. We never turn a rank into a visit count: none of these lists measures visits, and inventing a precise number from a rank would only look more accurate than it is.

How to read the band

• A top band is strong evidence of a big audience. Being in the top 10,000 of a list built from real browsing data doesn’t happen by accident.
• “Outside top 1M” is not “no visitors”. The lists cover roughly the top million sites. A small shop, a local business or a new site can have a loyal audience and still be outside them.
• Lists rank whole sites. A page deep inside a big site shares that site’s rank. A blog on a big platform’s subdomain may or may not be ranked on its own.
• Lists lag. A site that took off last week won’t show up yet; a site that collapsed recently may still be listed.

A sensible way to size up a site

1. Check its band in public popularity lists, and when each list was published.
2. If you need a number, compare two commercial estimates and treat the range as the answer, not either figure.
3. Look for signs of a real audience: comments, an active community, mentions on other sites.
4. If money depends on it, such as buying the site or a large advertising deal, ask for direct access to its analytics.

See where any site ranks: the Website Traffic Checker shows its band in Tranco and the Chrome UX Report, with each list’s date.