Rank sites by traffic: how the web rank tools build their numbers and which to use

To rank a set of sites by traffic, use one estimating tool for all of them and compare the sites against each other rather than reading any single number as fact. Similarweb, Semrush and Ahrefs each publish a traffic estimate and a rank for any domain, Cloudflare Radar publishes a popularity ranking built from its own resolver data, and Tranco publishes a research grade list that combines several sources. Only your own analytics and Search Console report measured traffic for a site you own, so use those for your own numbers and the estimators for everybody else's.

That distinction is the whole skill. Someone ranking sites by traffic is usually doing one of three jobs: sizing up competitors, choosing where to buy advertising or a backlink, or building a market map. All three work well on relative numbers, and all three go wrong when somebody quotes an estimate as though it came from the site's own logs.

Where each tool gets its data

Tool	Data source	Gives you	Free access
Similarweb	Panel and clickstream data, direct measurement from partners, modelled to a total	Visits, global and country and category rank, traffic by channel, audience overlap	Limited free view per domain
Semrush	Clickstream plus its own keyword index	Organic and paid traffic estimate, traffic by keyword, Authority Score	A few lookups a day
Ahrefs	Its own crawler and keyword index	Organic traffic estimate, Domain Rating, top pages	Free Website Authority Checker and Webmaster Tools for your own site
Cloudflare Radar	Cloudflare's public DNS resolver	A domain popularity ranking and trends	Yes, open
Tranco	Combines several public ranking lists over a rolling window	A stable research list of top sites	Yes, open
Chrome UX Report	Real Chrome users who opted in	Performance by origin, and a popularity bucket	Yes, via BigQuery and the API
Google Search Console	Google's own logs for a site you verify	Measured clicks, impressions, queries and pages	Yes, your own sites only
Google Analytics	Your own page tags	Measured sessions, sources, behaviour	Yes, your own sites only

Alexa, which is the rank most people remember, was retired by Amazon in 2022. Any article still quoting an Alexa rank is out of date, and any tool still displaying one is showing you a frozen number.

The two kinds of rank, and why they disagree

A traffic rank orders every site on the web by estimated visits, so position one is the largest site measured and the numbers run into the millions. An authority rank, such as Ahrefs Domain Rating, Semrush Authority Score or Moz Domain Authority, scores a site out of one hundred based on its backlink profile. They measure different things and they routinely disagree.

A regional news site can have enormous traffic and a modest authority score. A technical documentation site can have a very high authority score and little traffic. When somebody asks you to rank sites by traffic, confirm which of the two they want, because buying a link usually calls for authority plus topical fit, while buying advertising calls for traffic plus audience.

A method that holds up

1. Write the list of domains down first, in a spreadsheet, root domains only. Mixing a subdomain and a root domain in the same list is the most common error in a market map.
2. Pick one tool and use it for every row. Mixing a Similarweb visits figure with a Semrush organic figure produces a league table that means nothing. Note in the sheet which tool and which date.
3. Record the same two columns for all of them: total visits for the same month, and organic share of those visits. The second column is where the story usually is.
4. Sanity check three rows you know. Include your own site, because you have real figures for it, and see how far the estimate sits from the truth. That gap is roughly the gap on every other row.
5. Compare trends, not absolutes. Pull three or six months. A competitor up forty per cent over a quarter is a finding you can act on, whichever way the absolute number is biased.
6. Add the channel split before you decide anything. Two sites with identical visits are completely different businesses if one is ninety per cent organic and the other is ninety per cent paid.

Why the estimates differ from each other

Each vendor models a different population up to a whole. Panels skew by country, by device and by age. Clickstream data captures browser traffic better than it captures apps. A tool built on a keyword index estimates organic search traffic well and sees very little of the direct, email and social traffic a brand gets. Small sites are hardest of all, because there is less signal to model from.

Practical rule: trust the estimators most for large sites, for organic search traffic, and for comparisons between sites measured the same way on the same date. Use your own analytics whenever the question is how much traffic a site you control received.

This page covers ranking and comparing sites. For reading your own traffic in detail, see how to measure and read organic traffic, and for a repeatable competitive process see competitor analysis that ends in decisions.

Questions buyers ask about rank sites by traffic

How accurate are website traffic estimates?

Accurate enough to rank sites against each other, and closest on large sites with mostly organic search traffic. Check one site whose real figures you have, note how far the estimate sits from it, and apply that sense of scale to the rest of your list. For anything you need to be exact, use the site's own analytics.

What is the best free way to check a site's traffic?

Use the free lookups from Similarweb, Semrush and Ahrefs, note all three, and treat the middle one as your working figure. Cloudflare Radar and Tranco are fully open if what you need is a popularity ranking rather than a visit count. For your own site, Search Console and Analytics are free and measured.

Is there still a list of the top sites in the world?

Yes, several. Similarweb publishes global, country and category rankings. Cloudflare Radar publishes its own top domains. Tranco publishes a list designed for research that is deliberately stable from day to day. They order sites differently because they measure different things, so name the source whenever you quote a position.

What counts as good traffic for a small business site?

The useful comparison is against yourself and against the sites you compete with, not against an industry figure. Pull your own last twelve months and the same period a year earlier, then pull three competitors from one estimator. A site taking enough qualified enquiries to fill the diary is doing its job at any volume.

Why does a competitor show high traffic and no visible rankings?

Their traffic is coming from somewhere other than organic search. Look at the channel split: it will usually be direct traffic from a known brand, paid search, paid social, email, or an app. That is a useful finding, because it tells you the channel to study rather than the keyword list.

One ranking none of these tools covers yet

Every tool above ranks sites by people arriving at them. A share of buyers now never arrives anywhere. They ask ChatGPT, Gemini, Perplexity or Claude, or they read the AI Overview at the top of Google, and they act on the three or four companies named in the answer. A site that is never named in those answers can hold a strong traffic rank and still be absent from the moment of choice.

There is no shared ranking behind it, so a strong position in one assistant tells you nothing about the others, and each one has to be checked separately. Write down the five questions a buyer asks before choosing in your category, ask them in two or three assistants, and note which companies get named and which pages get cited. Keep your own brand name out of the question so the answer comes back clean. AI Knows Us runs that check across the main assistants on a schedule and reports which sources each answer was built from.