Tranco, Cloudflare Radar and CrUX: free top-site rankings

Tranco, Cloudflare Radar and the Chrome UX Report (CrUX) are the three most useful free sources for ranking websites by popularity. They measure different things, publish on different schedules and fail in different ways. Knowing how each works lets you use them for competitor research without drawing the wrong conclusion.

Why free rankings matter now

For years the standard free reference was a single well-known ranking that has since been retired. Today there is no single replacement, which is why researchers and analysts rely on several lists. Each one is a lens on the same question: how many people reach this domain, compared with other domains?

None of them give you visits directly. All of them give you a position, and positions are the raw material for traffic models. If you want a primer on turning positions into numbers, see our guide to how rankings translate into monthly visits.

Tranco

Tranco was created to give researchers a ranking that is stable and hard to manipulate. The project describes its standard list as one million domains obtained by averaging four rankings over the past 30 days. Those four inputs are Cisco Umbrella, Majestic, the Chrome UX Report and Cloudflare Radar.

Two details are worth knowing:

• Tranco applies filters, including Google Safe Browsing and a CrUX filter, so that malicious sites are removed and only sites regularly visited in Chrome remain.
• It is published daily, and you can download a specific day's list, which makes comparisons reproducible.

What Tranco is good at

Tranco is the best single starting point. Because it averages four independent signals over a month, a short spike or a quirk in one input does not move a domain far. It is ideal when you need a stable answer to "roughly how popular is this site?"

Where Tranco can mislead

It is an average, so it hides disagreement between its inputs. A domain that is huge in DNS lookups and nearly absent from link data will land somewhere in the middle. The result is a ranking, not a visit count, and the gaps between ranks grow wider as you go down the list.

Cloudflare Radar

Cloudflare Radar's domain rankings come from aggregated, anonymized data from Cloudflare's public 1.1.1.1 DNS resolver. According to the company's explanation, popularity is estimated as the relative size of the user population looking up a domain over a period of time, not raw traffic volume. Automated traffic is filtered out.

It publishes two kinds of output:

• An ordered top 100, globally and per country, covering the last 24 hours.
• Unordered buckets, from the top 200 up to the top million, covering the last 7 days and updated weekly.

What Radar is good at

Radar is timely. Because it refreshes daily at the top and weekly in the buckets, it reacts to change faster than a monthly source. It is also useful for country-level views, since it ranks domains per country.

Where Radar can mislead

Radar sees only people using Cloudflare's resolver. Users on other resolvers, corporate networks or mobile carriers with their own DNS are outside its view. And the buckets are deliberately unordered inside each band, so you cannot tell whether a domain is first or last in its bucket. A DNS lookup also is not a visit, so domains used by apps and services can rank above what their human audience suggests.

Chrome UX Report (CrUX)

CrUX is a Google dataset that reflects how real Chrome users experience popular websites. It is part of the Web Vitals program and mainly exists to help site owners understand performance, but it also reveals popularity, because only sites with enough visitors appear.

The methodology documentation is specific about who counts. A user must have usage statistics reporting enabled, sync their browser history and use a supported platform such as desktop Chrome or Android Chrome. Chrome on iOS and other Chromium browsers do not contribute. An origin must be publicly discoverable and have at least an undisclosed minimum number of visitors.

On the data side, Google documents that per-country tables exist alongside the global data, and that monthly data is released on the second Tuesday of the following month.

What CrUX is good at

CrUX is the best free source of geographic information about real visitors. Because it is built from actual browsers, it is harder to game than a link-based list. Its eligibility threshold also acts as a useful filter. If a site has data in CrUX, it has real traffic.

Where CrUX can mislead

It covers a slice of Chrome users, and Chrome does not publish what proportion of users that is. It will skew toward people who sync history and keep telemetry on. Sites mostly visited from iPhones, which means Safari or Chrome on iOS, are under-represented. Sites with too little traffic are absent entirely, which leaves you blind exactly where small startups live.

How the three differ

	Tranco	Cloudflare Radar	CrUX
What it observes	Average of four lists	DNS lookups via one resolver	Opted-in Chrome users
Update rhythm	Daily, 30-day average	Daily top 100, weekly buckets	Monthly
Output	Ordered top million	Ordered top 100, unordered buckets	Popularity and country data
Best use	Stable overall rank	Fresh signals and country views	Real-user country mix
Main risk	Hides disagreement	Single-resolver view	Chrome-only slice

How to use them together

The reliable approach is triangulation. Check a domain in more than one source and ask whether they tell the same story.

1. They agree. If Tranco, Radar and CrUX all place a site in the same band, you can trust the scale.
2. Radar high, CrUX absent. The domain may be API or app driven, with few people browsing it.
3. Tranco decent, Radar weak. The site may rely on link authority or an older audience.
4. CrUX strong in one country. The audience is regional. Use that to refine your market view.

StartupsIntel does this combination for you. The traffic checker pulls from five free lists, adds Chrome country data and shows a modelled visit range. The monthly top traffic rankings apply the same approach across thousands of domains, and the methodology page lists each input.

Practical tips

• Download the same list on the same schedule, so changes reflect the site rather than your timing.
• Record the rank, the date and the source name. Without those three, a number cannot be compared later.
• Do not compare a Radar bucket to a Tranco rank directly. Convert both into a rough band first.
• Be especially careful with subdomains. Lists sometimes merge them into the parent domain and sometimes treat them separately.

Other lists worth knowing

Two more sources deserve a mention because they feed the aggregate lists. Cisco Umbrella publishes a popularity list based on DNS queries seen by its resolvers, and the research literature on top lists, such as the paper A Long Way to the Top, explains how these lists can be manipulated and why averaging several of them improves stability. Majestic, as noted above, ranks by link data.

The lesson from that research is simple. Any single list can be wrong or gamed, and a combination is harder to fool. That is the reason to look at several rather than favor one.

FAQ

Which free ranking is the best?

Tranco is the best single starting point because it averages four sources over 30 days. For country-level questions, Cloudflare Radar and the Chrome UX Report add detail that Tranco cannot.

Do these rankings show actual visit numbers?

No. They show position, not volume. Visit estimates require a model that converts position into a range, and those estimates should always be labelled as such.

Why is my website missing from these lists?

Most likely because it does not have enough traffic. CrUX requires an undisclosed minimum number of visitors, and the other lists only publish the top million domains. Absence means small, not broken.

How often should I check these rankings?

Monthly is enough for most research. Check weekly only if you are watching a launch or a campaign, and compare like with like.