Read country and device reports responsibly
Use country and device breakdowns to investigate traffic and conversion while accounting for small samples and derived context.
Country and device rows are useful for broad patterns, quality checks, and experience testing. They are derived dimensions, not verified facts about a person.
Audience breakdowns add context to traffic, content, goal, funnel, and revenue reports.
Use them for broad patterns and quality checks, not for identifying people.
Country reports
Country-level context can help with:
- Finding unexpected traffic spikes.
- Reviewing regional campaign reach.
- Prioritizing language or localization research.
- Checking whether a launch reached its intended market.
Country is not a verified residence. VPNs, relays, proxies, corporate networks, and mobile routing can change the derived location.
Device reports
Device categories help compare desktop, mobile, tablet, and other recorded device types.
Use device data to locate experiences worth testing. For example, a large mobile audience with weak funnel progression is a reason to inspect the mobile flow, not automatic proof that the device caused the difference.
Watch sample size
A segment with two conversions from ten visitors has a 20% conversion rate, but it is not more reliable than a segment with hundreds of visits.
Read the visitor count beside any rate and compare a sufficiently long period.
Connect the breakdown to action
When a row looks unusual:
- Confirm the traffic volume.
- Compare its sources and landing pages.
- Review relevant goals and funnel steps.
- Test the site on the device class or from the region where practical.
- Check whether tracking coverage differs.
Next steps
Use traffic sources and referrers to understand acquisition mix or build a funnel to compare progression.
FAQ
Is country the visitor's home country?
No. It is country-level context derived for the recorded request and can be affected by VPNs, relays, proxies, and network routing.
Should I optimize for the highest-converting device immediately?
Check traffic volume, source mix, and the actual experience first. Small samples can create unstable rates.