How to check website traffic (and what each number means)

How to check website traffic exactly on a site you own and estimate it on a competitor's, with the sources, the numbers and the limits.

·5 min read

Website traffic is two different measurements wearing one name. On a site you own, you can read real visits from analytics, server logs and Google Search Console. On any other site you can only estimate organic traffic from the keywords it ranks for, the searches those keywords get and the share of clicks each position usually earns. The first number is measured. The second is modeled, and treating it as a visitor count leads to bad decisions.

Both are useful. They answer different questions, and the rest of this guide keeps them apart.

Which source answers which question

Question Source Accuracy
How many people visited my site? Analytics such as GA4, Plausible or Matomo Measured, but undercounts ad blockers and declined consent
How many clicks did Google send? Search Console Performance report Measured for Google Search, 16 months of history
Which pages and files were requested? Server or CDN logs Measured, and includes bots
How much organic traffic does a competitor get? A traffic estimator built on keyword data Modeled, directional only
How much traffic does one competitor page get? Page-level estimate in a domain tool Modeled, weakest for small sites

If the question is about your own site, start with analytics. If it is about someone else's, start with a keyword-based estimate and read it as a range, not a number.

How to check traffic on a site you own

Three sources cover nearly every question. Use them in this order.

  1. Open analytics and set the date range. Look at users, sessions and views separately. A user is a person, a session is one visit, and a view is one page load. One person can create three sessions and nine views.
  2. Split by channel. Organic search, direct, referral, social, email and paid. A traffic drop in one channel is a different problem from a drop across all of them.
  3. Compare with Search Console. Search Console counts clicks from Google Search only. Analytics counts sessions from every source. They will never match exactly, and reconciling them to the visit is a waste of time. Pick one source for one question.
  4. Read the landing pages report. It shows which pages start sessions rather than which pages get seen. A page with many views and few entrances is a page people reach internally, not from Google.
  5. Check the logs when a number looks wrong. Logs show every request, including bots that never run JavaScript. They settle whether a spike is real traffic or a crawler.

Our guide to reading search performance walks through connecting Search Console, Bing Webmaster Tools and analytics without double-counting.

Clicks, sessions and users are not interchangeable

Metric Counts Where it lives
Click A Google result click Search Console
Session A visit, however many pages it reads Analytics
User A person, deduplicated by cookie or login Analytics
Pageview One page load Analytics and logs
Request One file or document fetched Server logs

A page can gain clicks while sessions fall, because the same person clicks more often or because a tracking change broke the session count. Check two metrics before believing a trend.

How to check traffic on someone else's site

You cannot see a competitor's analytics. Every tool that claims otherwise is modeling, usually from the same three inputs: the keywords the domain ranks for, the monthly search volume of each keyword, and an estimated click-through rate for its position.

  1. Confirm you have the right domain. Include the subdomain that actually ranks. A blog on blog.example.com and a store on shop.example.com are separate hosts with separate profiles in most tools.
  2. Pick one market. Search volume and rankings differ by country. Mixing a US estimate with a UK keyword set produces a number that describes neither.
  3. Pull the domain overview. Read estimated organic traffic, the number of ranking keywords, traffic value and the top pages together. One number alone is easy to misread.
  4. Look at the trend, not the total. A site at 40,000 monthly visits and falling is a different target from one at 12,000 and doubling every quarter.
  5. Check the top pages. A competitor whose traffic comes from one calculator or one viral post is not strong across the board. A competitor with fifty pages sharing the traffic is.

Our website traffic checker guide explains how the estimate is built and where measured and modeled traffic disagree.

What the estimate cannot see

Traffic estimators are blind to most of the internet's actual visits.

  • Direct and branded traffic. People who type the domain or search the brand are often the largest segment for an established business.
  • Email and newsletters. A site with 200,000 subscribers can get more visits from one send than from a month of Google.
  • Paid search and social. Ads do not appear in organic estimates unless the tool tracks paid placements separately.
  • Social and community traffic. Reddit, LinkedIn and YouTube referrals are invisible to keyword models.
  • Logged-in and app traffic. A product used behind a login may serve millions of sessions with almost no indexable pages.
  • Small sites. Below a few hundred ranking keywords, the model has too little data and the error grows.

A site can look small in a traffic checker and still be a serious business. A site can look large and be entirely paid.

Sanity-check a traffic estimate before quoting it

Use two independent signals. If they disagree, trust the weaker claim.

  • Brand search volume. A brand with real demand has people searching its name. A brand with none is usually bought traffic.
  • Ad presence. Check the Google Ads Transparency Center and Meta Ad Library. Heavy ad spend with low estimated organic traffic suggests paid acquisition.
  • Indexed pages and content depth. site:example.com in Google gives a rough page count. A domain with 40 pages rarely earns six figures of organic visits.
  • Social and community footprint. A company with 2,000 followers and no engagement is not running a large audience.
  • Hiring and funding. A team of three and a seed round sets a different ceiling from a 300-person company.

Which number to use for which decision

Decision Use Do not use
Fix a traffic drop on my site Analytics plus Search Console, by page and query A third-party estimate
Find content gaps Competitor keyword and page data Their total visit count
Size a market Search volume and CPC for the category One competitor's traffic
Judge an acquisition target Their analytics, under NDA Public traffic checkers
Report to a client Analytics and Search Console A modeled number presented as fact

The common mistake is mixing the two. A modeled competitor figure belongs in a slide labeled estimate. A measured analytics figure belongs in a report with a date range and a source.

Check a site's traffic without confusing the two

Our Domain Overview returns estimated organic traffic, ranking keywords, traffic value and top pages for any domain, in one market you choose. Use it to compare competitors and find pages worth studying, then use your own analytics and Search Console for every number you have to defend. If you want to see how a single rival compares with you, read how to check a competitor's website traffic next.

Related guidesFrom the same topic, then the newest guides.