How to check website traffic: your own site, and any other site

How do you check website traffic?

To check website traffic, use Google Search Console and your analytics for your own site, and a website traffic checker for anyone else's. The two answers are different kinds of number. Your own traffic is measured from your own site's data. Another site's traffic is a model's guess from the outside, because nobody but the owner can see it.

How to check your own search traffic in Search Console

Open Google Search Console, go to Performance → Search results, and read Total clicks: that is the number of times people clicked through to your site from Google Search. The Performance report help defines a click as a user clicking your site from Google Search results, and an impression as your site appearing in the results.

1. Verify the site

   Add your domain as a property in Search Console and verify it.

2. Pick the date range

   Use the last 28 days for a monthly figure, or compare two periods to see the trend. Turn on the Impressions, Click rate and Average position metrics too.

3. Break it down

   The Queries tab shows which searches bring clicks, the Pages tab shows which pages earn them, and the Countries and Devices tabs show where they come from.

4. Check the other search types

   Search Console splits Web, Image, Video and News results. Switch the search type to see clicks you would otherwise miss.

If your clicks fell, our checklist on why website traffic drops walks through tracking problems, Google updates and lost rankings in order.

How to check total website traffic in analytics

For all your visits, not just search, open your analytics tool: in Google Analytics 4, go to Reports → Acquisition → Traffic acquisition. Google's Traffic acquisition help says the report groups sessions by channel, such as Organic Search, Direct, Organic Social, Paid Social and Email, and that you can search for a source such as “google / organic” to see that traffic alone.

Expect Analytics sessions and Search Console clicks to disagree. Google's guide to using Search Console and Analytics together lists the reasons:

• Different units. Search Console counts clicks from Google Search; Analytics counts sessions, from any source.
• Time zones. Search Console reports in Pacific Time; Analytics uses the time zone you set.
• URLs. Search Console reports the canonical URL Google chose; Analytics counts every URL that has the tracking code.
• Bots and consent. Analytics excludes known bots, and visitors who decline tracking or block scripts are missing from it.
• Files. Search Console counts clicks to PDFs and other non-HTML files, which Analytics may not measure.

Google's guidance is clear on which to trust: Search Console is the source of truth for search performance, and Analytics for what visitors do on your site. Bing has the same kind of report for its own results in Bing Webmaster Tools Search Performance.

How to estimate a competitor's website traffic

To estimate competitor website traffic, enter their domain in a website traffic checker and read the result as a range, then compare it with other sites in the same tool. No checker can see a competitor's Search Console or analytics, so each one builds a model from data it can collect from outside. The three common approaches:

Each approach is fair for some questions. Ahrefs and Semrush have far larger keyword and panel datasets than MrKeyword, and Similarweb's sources include sites that share their own analytics with it, so for big sites and channel breakdowns they will usually be the better choice. A link-based estimate's strengths are that it is free, covers any well-linked domain and says exactly how it was made.

How accurate are website traffic checkers?

Every third-party traffic number is a model, and for any one website it can be off by a large factor; the vendors themselves say their numbers will not match the site's own data. Ahrefs' help article on why its numbers differ from Analytics and Search Console gives three reasons: it does not track every keyword, its search volumes are estimated, and its click-through rates are modelled, adding that only Google knows the real numbers of clicks. Semrush's Traffic Analytics article says its data is approximated from clickstream behaviour and is not interchangeable with Google Analytics. Similarweb describes its figures as estimates and says differences from Google Analytics are expected.

Errors stack up. A keyword-based estimate multiplies a rounded search volume by an average click rate for a position that changes daily, then adds thousands of those guesses together. A panel-based estimate scales a sample of users up to the whole internet. Either can land close for one site and far off for the next.

How to use estimates well:

• Compare inside one tool. If a checker says competitor A gets three times the traffic of competitor B, that ratio is more reliable than either absolute number.
• Check the tool against a site you know. Enter your own domain and compare the estimate with your Search Console clicks. The gap tells you how far to trust that tool in your niche.
• Watch direction, not level. A competitor's estimate rising steadily over months says more than this month's figure.
• Match the metric. Organic clicks, visits and unique visitors are different numbers. A tool estimating organic clicks will always be below an analytics total for all channels.

How MrKeyword's website traffic checker estimates traffic

MrKeyword's free website traffic checker estimates a domain's monthly organic search clicks from its position in the MrKeyword web graph, and shows every figure labelled Estimated with a low-to-high range and the model that produced it. Enter up to five websites, with or without https://; results appear on the page with no account, and each has a link you can share.

1. Position in the web graph

   The MrKeyword web graph ranks about 135 million domains by PageRank, that is, by how the rest of the web links to them, and keeps the top 50 million. Well-linked sites tend to get more search traffic, so position is the input.

2. A documented baseline curve

   With no calibration, organic clicks are spread over domains by position on a Zipf curve (a site's share falls in proportion to its position), scaled to Google's estimated monthly organic clicks. The tool states this formula, marks the result low confidence and gives a ±3x range.

3. Calibrated on real clicks

   Sites connected to MrKeyword supply their measured Search Console clicks from the last 28 days. The curve is shifted by the average gap between those real clicks and the curve, cautiously while sites are few, and only near the positions where connected sites actually sit. The ±3x range narrows only after eight sites calibrate it.

4. A fitted model, when it earns it

   Once at least 20 connected sites are available for training, a regression on web-graph position, linking domains and hosts is fitted daily. One site in five is held out, and the fit replaces the curve only if at least half of the held-out sites land within 3x of their real clicks. Its range comes from that held-out error, never narrower than 1.5x either way.

Each result says which of these models produced it and how many connected sites calibrated it. Figures are rounded to two significant digits so they never look more exact than they are. Ranking keyword counts appear only when the model has enough connected sites reporting queries to estimate them. Alongside traffic you see the domain's authority; the domain authority checker and backlink checker go deeper on links.

For your own site, connect Search Console on a free account and MrKeyword shows measured clicks, positions and keywords instead of an estimate, in the free rank tracker, with traffic drop alerts when clicks fall. To size the demand behind a single keyword rather than a whole site, see how to find keyword search volume.

What MrKeyword does not do

• It does not measure anyone's traffic. Only the site's owner can, in Search Console or analytics. Every figure in the checker is labelled Estimated.
• It estimates organic search clicks only, not direct, social, paid, email or referral visits, so it will always be lower than a site's total visits.
• It does not use Similarweb, Semrush, Ahrefs or any paid traffic or SEO data API, and it does not use a clickstream panel or scrape Google.
• It does not break traffic down by keyword, page or country for sites you do not own.
• Websites outside the top 50 million domains of its web graph, usually new or barely linked sites, show “No data” rather than an invented number.
• Its range is wide on purpose. Until enough connected sites train the fitted model, treat a single site's figure as an order of magnitude.

Questions and answers