GA4, Hotjar and Clarity Show Different User Numbers: Why This Is Normal in Some Cases
When GA4, Hotjar and Microsoft Clarity show different user or session numbers, the instinct is to find which one is wrong. Often none of them are wrong. Each platform uses a different measurement model, applies different consent and coverage rules and handles browser behavior differently. Forcing them to match wastes time. The useful work is identifying which differences are expected and which are unexpected.
Table of Contents
- 01.Why three platforms report different numbers
- 02.What to check first
- 03.Expected vs unexpected differences
- 04.How to test the fix
- 05.How to validate the final result
Why three platforms report different numbers
GA4 counts users and sessions through its tracking implementation. Hotjar counts sessions or users through its own tracking. Clarity counts sessions through its script. Each defines a session or user differently, each loads under different consent conditions and each is blocked by browsers to a different degree. Three different measurement models produce three different numbers. That is the expected starting point.
| Cause | What happens | How to check |
|---|---|---|
| Different model | User vs session vs visitor definitions | Confirm what each platform counts |
| Consent | Different consent coverage | Compare consent setup across all three |
| Coverage | Different pages or traffic tracked | Confirm what each tracks |
| Browser blocking | Ad blockers hit one more than others | Check script blocking |
| Sampling or processing | Platform-specific collection rules | Confirm each platform's rules |
What to check first
- Confirm all three platforms track the same pages and traffic.
- Match the date range and time zone across all three.
- Compare the consent setup of each platform.
- Confirm none of the scripts is blocked more than the others.
Expected vs unexpected differences
A stable gap between the three platforms is expected. GA4 may report fewer users than Clarity because of consent or blocking. Hotjar may report fewer sessions because of sampling or plan limits. These differences are structural. An unexpected difference is a sudden change in one platform that the others do not share. That points at a tracking failure in the platform that moved, not a measurement difference.
How to test the fix
- 1.Confirm all three scripts fire on the same pages.
- 2.Match the date range and time zone exactly.
- 3.Compare consent handling across all three.
- 4.If one platform dropped suddenly, repair that installation.
How to validate the final result
Validation is not identical numbers across all three. It is a stable, explainable relationship between them. If one platform diverges suddenly and the others do not, that platform has a tracking problem. If all three move together at different levels, the differences are structural and expected.
Frequently Asked Questions
Should GA4, Hotjar and Clarity show the same numbers?
No. Each uses a different measurement model and different consent and coverage rules. A difference is normal. Investigate only sudden or unexplained changes in one platform.
Need help applying this to your business?
PPC, Conversion Tracking, CRM and Automation Specialist. Helping businesses generate qualified leads with Google Ads, accurate tracking and automated follow-up.
Related Guides
Microsoft Clarity and GA4 Data Do Not Match: Why Session Numbers Can Be Different
Clarity and GA4 define sessions differently and apply different consent and coverage rules. A difference is normal. Investigate only sudden or unexplained changes.
Hotjar Heatmap Clicks Do Not Match GA4 Events: Why the Numbers Can Differ
Hotjar counts visual clicks and GA4 counts events sent through tracking. They measure different things under different consent and coverage rules, so a difference is usually expected.
Website Analytics Audit Checklist: GA4, Looker Studio, Hotjar, Clarity and CRM
A full analytics audit from tracking installation to final reporting across GA4, GTM, Looker Studio, Hotjar, Clarity and CRM. Find gaps, duplicates and mismatches across the whole stack.