Skip to content

Why two analytics tools never show the same numbers

Updated October 19, 2026

Put two analytics tools on the same site and they will disagree, sometimes by a few percent, sometimes by half. Neither is necessarily wrong; they measure slightly different things.

Consent

A tool that sets cookies only runs after a visitor accepts them. If 40 % of your visitors reject or ignore the banner, a cookie-based tool simply does not see them. A cookieless tool that needs no consent counts them. This is usually the largest difference.

Blockers

Browser extensions and some browsers block requests to analytics domains they know. Each tool is on different lists, so each loses a different share of visitors, typically larger on technical audiences. Serving the script from your own domain avoids most of it: see Serve the script from your own domain.

Bots

Tools filter automated traffic with different rules. A crawler that one tool drops can show up as a visitor in another, especially on small sites where a single bot is a visible share.

What counts as a visitor

  • A cookie-based tool counts a browser that keeps its cookie as one user for months, until it is deleted.
  • Statsdot counts a visitor per day: someone who comes back tomorrow counts again, because the visitor code changes every day.
  • Over a month, "visitors" in Statsdot is closer to "daily users added up" than to "monthly unique users".

Time zones

Statsdot counts in UTC days. A tool set to your local time draws the line between days a few hours away, so day-by-day numbers shift even when the weekly total agrees.

When the script loads

Scripts loaded late (after other heavy scripts, or by a tag manager) miss visitors who leave in the first second. Scripts loaded with defer in the page head catch almost everyone.

Single-page apps

A site that changes pages without reloading needs the tool to notice the address change. If one tool counts those navigations and the other does not, pageviews differ while visitors agree.

So which one is right?

Pick one tool as the reference, and compare trends rather than totals: if both show the same week-on-week change, they agree on what matters. When you switch tools, run both for a couple of weeks to learn the ratio on your own site.