The most common causes
Different date ranges
The experience view and the experiment view may not be using the same reporting period. One surface can show a broader or all-time total while the experiment is filtered to a specific date range. Always align the date range before comparing the counts.Traffic reached the experience directly
If someone uses the direct experience URL instead of the experiment’s Smart Link, those sessions can appear on the experience without belonging to the experiment variant. This commonly happens when:- An ad agency uses the direct landing-page URL
- A preview or QA link is shared broadly
- An old campaign still points directly to the experience
- The experience was live before it was added to the experiment
Internal QA traffic
Manual testing can also create small differences, especially immediately after launch.How to diagnose the difference
1
Align the reporting period
Set the same start and end dates on the experience and experiment views.
2
Check the campaign destination
Confirm every live ad or campaign uses the intended experiment Smart Link rather than the direct experience URL.
3
Compare the size of the difference
A one- or two-session difference may be QA traffic. A large or growing difference usually points to another traffic source.
4
Check older links
Review active ads, emails, creator links, and redirects that may still reference the underlying experience directly.
If the experience is only intended to receive experiment traffic, the direct page URL should not be used as a live campaign destination.
Related guides
- How Experiment Traffic Is Assigned
- Smart Links
- Analytics Overview
- Filtering Data on Experiences