Differences Between Omniconvert & Analytics Tools
Overview
It's normal to see small or large differences between the number of conversions reported in Google Analytics (or another analytics tool) versus your Explore dashboard. Explore isn't an analytics tool - it's a CRO platform - and it deliberately tracks things differently for good reasons.
Fail Safe
Explore applies live changes to your website, and needs to do so without slowing down the page for visitors. As a safety mechanism, if an experiment can't be applied within 2 seconds, it's canceled for that visitor.
Example: a visitor on a poor connection may not retrieve data from Explore's servers in time - in that case, they're deliberately excluded from the experiment rather than shown a slow or broken page.
This is closely related to the Loading Tolerance setting - see Configure Website Settings for the full explanation of that mechanism.
Relevant Data
Explore only tracks conversions that were actually influenced by an experiment. If you allocate just 50% of your traffic to an experiment, conversions from the untested 50% are ignored entirely - they were never part of the test.
This is one of the most common reasons Explore's numbers won't match your analytics tool's total conversion count, which typically counts everyone regardless of experiment participation.
Maintenance Time
Explore occasionally pauses parts of the system to improve performance or roll out new features. Conversions can't be tracked during these windows.
When working with a CRO tool, base your decisions on relative data - percentages and averages - rather than absolute conversion counts. Absolute numbers will always be somewhat sensitive to the reasons above.
Tablet Detection Differs from Google Analytics
Explore consistently reports more tablet users than Google Analytics. The cause: iPads.
How iPads "hide" from GA
Since iPadOS 13+, Safari on iPad sends a desktop User-Agent string:
Mozilla/5.0 (Macintosh; Intel Mac OS X 10_15_7) AppleWebKit/605.1.15 ...Nothing in that string identifies the device as an iPad. Google Analytics determines device category purely from the User-Agent, sees "Macintosh," and classifies the visit as Desktop - systematically undercounting iPad tablets.
How Explore detects it correctly
Explore's tracking runs client-side JavaScript that checks navigator.maxTouchPoints. iPads report maxTouchPoints = 5, while Mac desktops report 0 or 1 - letting Explore correctly identify the device as a tablet before sending device_type=tablet to the server.
This is a genuine detection difference, not an error on either tool's part - Explore is simply able to use a signal (maxTouchPoints) that GA's User-Agent-only approach doesn't check.