How to Track Website Traffic in 2026 (And Why Your Tools Never Agree)

That last part is not a bug in your setup, and understanding it is most of what separates people who read analytics from people who look at analytics.

Step 1: Install GA4 properly

Create a property at analytics.google.com and install the tag. You have two ways to do it, and the choice matters more than it looks.

Pasting the tag directly into your site header works and takes five minutes. Installing it through Google Tag Manager takes twenty minutes and means every future tracking change happens without touching your site code. If you expect to add anything at all later, conversion tracking, a Meta pixel, a call tracking script, do it through GTM. We cover what Google Tag Manager is and when you genuinely need it separately, including the cases where it is overkill.

Then verify it. Open DebugView, load your own site, and watch the events arrive. Skipping this step is how people end up analyzing three months of data from a tag that only fired on the homepage.

If you are starting from a completely blank account, the settings to fix before you read anything are in how to use Google Analytics from a standing start.

Step 2: Connect Search Console

Google Search Console is free, separate, and the single most underused tool in this list. Verify your domain, then link it to GA4 so search queries appear alongside behavior data.

What Search Console gives you that GA4 cannot: the actual queries people typed, how often you appeared, and where you ranked. GA4 knows somebody arrived from Google. Only Search Console knows what they searched for.

Step 3: Read the report that answers the question

The report you want is Reports, Acquisition, Traffic acquisition. By default it groups sessions into channels. Change the dimension dropdown to Session source / medium and you get the granular view, which reads like this:

What you see	What it means
google / organic	Found you through an unpaid Google search result
google / cpc	Clicked one of your Google Ads
facebook.com / referral	Clicked a link on Facebook that you did not tag
(direct) / (none)	Google could not determine a source
newsletter / email	Came from a link you tagged yourself

If you used Universal Analytics years ago, this is the Source/Medium report. It still exists, it moved, and it answers the same question it always did. The full map of where every old report went is in what each GA4 report actually shows.

Step 4: Fix your Direct traffic before you trust anything

Here is the most common problem in small business analytics, and it makes every other number wrong.

(direct) / (none) is supposed to mean somebody typed your URL or used a bookmark. In practice it is a landfill. Untagged email links, links from apps, PDFs, QR codes, some redirects, and anything where the referrer got stripped all pile up there. We routinely see accounts where 40% of traffic is Direct, and almost none of it is what the label claims.

The fix is tagging every link you control with UTM parameters. It is thirty minutes of work once you have a naming convention, and it moves that traffic into channels you can actually evaluate. Our guide to UTM parameters includes the convention we use, which matters more than the tags themselves.

Until Direct is under control, any conclusion you draw about which channel works is a guess.

Why your tools never show the same number

This is the part that sends people looking for a second opinion, so let me put the actual reasons in one place.

Tool	What it counts	Why it reads low or high
GA4	Sessions, client-side via JavaScript	Reads LOW. Ad blockers, consent rejections and users who leave before the script loads are never counted.
Search Console	Clicks from Google search results, server-side	Reads HIGHER than GA4 for organic, because Google counts the click it sent regardless of whether your tag fired. Only covers Google search.
Server logs / host dashboard	Every HTTP request	Reads WILDLY HIGH. Bots, crawlers, uptime monitors and scrapers are all in there.
Third-party checkers	Modeled estimates	Not your data at all. Directional only.

A 10% to 30% gap between GA4 sessions and Search Console clicks for the same period is normal and healthy. A gap of 60% means something is broken, usually a tag that is not firing on part of the site or a consent banner blocking analytics by default.

The practical rule: pick one tool per question and stop reconciling. Search Console is the truth for search performance. GA4 is the truth for on-site behavior and conversions. Never quote a number from one to answer the other's question.

What about checking someone else's traffic

Half the results for this search are tools promising to check any website's traffic. Ahrefs, Semrush, Similarweb and SpyFu all offer one.

They work by modeling: clickstream panels, search volume data, and extrapolation. For a large site the estimates land in the right neighborhood. For a local business doing a few thousand visits a month they are frequently off by multiples, and sometimes report zero for sites that are doing fine.

Use them to compare competitors against each other, which is what they are genuinely good at. Do not use them to judge whether a competitor is beating you in absolute terms, and never use them on your own site when you have GA4 installed.

The metric worth watching

Once tracking is clean, the number to build a habit around is not sessions.

It is engaged sessions by channel. GA4 counts a session as engaged if it lasts more than ten seconds, fires a key event, or reaches two or more pages. Sessions tell you how many people arrived. Engaged sessions tell you how many arrived and did not immediately leave, which is the difference between a traffic report and a marketing report.

Run it monthly, by channel, and you will spot a failing campaign weeks before the leads stop.

One caveat to hold alongside it. Even a clean setup only sees the touchpoints that happened on your site, which is why your conversion path looks simpler than it was. The report is a partial view, and treating it as a complete one is how demand-building channels get cut to fund the one that captured the demand they built.

Behavior tools sit on top of this rather than replacing it. When the numbers say a page loses people and you want to see why, heatmaps show behavior analytics cannot.

Common questions

Why don't Google Analytics and Search Console match?

They measure different events with different methods. Search Console counts clicks Google sent you, recorded on Google's side. GA4 counts sessions recorded by JavaScript on your side, which misses anyone with an ad blocker or a rejected consent banner. A 10% to 30% difference is expected.

Can you see who visits your website?

Not as individuals, and not legally in most contexts. You can see aggregate location, device, source and behavior. Tools claiming to identify individual visitors by name are doing IP-to-company matching, which works poorly for consumers and carries real privacy exposure.

How do I check a competitor's website traffic?

With a third-party estimator such as Similarweb, Semrush or Ahrefs. Treat the output as a comparison between competitors rather than a measurement of any one of them. There is no way to see another site's real analytics.

Is Google Analytics free?

Yes. GA4 is free for the volumes any small or mid-sized business will produce. The paid tier, Analytics 360, starts in the tens of thousands per year and exists for enterprises hitting data limits.

What is direct traffic really?

Officially, people who typed your URL or used a bookmark. Actually, a catch-all for any visit where the referrer was lost, which includes untagged email, app links, PDFs and QR codes. If Direct is more than about 20% of your traffic, you have a tagging problem rather than a loyal audience.

Where most setups go wrong

Almost every broken analytics setup we open has the same three problems: the tag is missing on some templates, nothing internal is excluded so the team's own visits are in the data, and campaign links were never tagged so half the traffic sits in Direct.

None of those are hard to fix. All of them silently invalidate every decision made from the data in the meantime, which is the expensive part. A year of confident conclusions from bad numbers costs more than a year of no conclusions.

We audit this as part of our SEO service in Hialeah, because a ranking report built on broken measurement is worth nothing to anybody.

Written by J Raydel Sanchez, Founder and CEO of tamer, a digital marketing agency in Hialeah, Florida.