Learn analytics
Filter website analytics
Narrow dashboard reports by acquisition, location or device without changing metric meaning.
What it is
An analytics filter limits the dashboard to visits that match a selected condition. Premely supports source, UTM, country, region, city, device, browser and operating-system context in the live dashboard.
Why use it
Use a filter to test a specific explanation, such as whether a traffic increase came from one campaign source or one country.
Before you start
- Read the unfiltered result first so you understand the overall change before narrowing it.
- Choose one filter that directly relates to the question. Multiple unrelated slices can make a small sample look more important than it is.
How to use it
- 1
Choose a filter group
Open Filter and choose Source or a
UTMfield for acquisition, Country, Region or City for location, or Device, Browser or Operating system for technology context.Product screenshot placeholderDashboard - Filters
Capture focus: Visit-scoped filter, operator and selected value
1Filter field2Operator3Applied filterReplace this slot with the matching production surface and numbered callouts before publishing. - 2
Choose the operator
Use
isoris notfor exact inclusion or exclusion. Usecontainsordoes not containfor supported text fields when an exact value is too narrow. Location filters use exact inclusion or exclusion. - 3
Enter one meaningful value
Use the value recorded by the report, such as a source, campaign tag or country. Keep campaign naming consistent so related visits can be filtered reliably.
- 4
Read the applied filter
Confirm the filter pill above the dashboard before interpreting the metrics. The headline cards and supporting reports now describe the matching visit context.
- 5
Remove the filter and compare
Return to the full result to understand the filtered group's contribution. If the filter does not change the decision, remove it rather than carrying it into later analysis.
What to expect
Session metrics keep their meaning
Views per visit, bounce rate and visit duration are calculated from complete visit context. Premely rejects unsupported event-level combinations instead of treating a matching event as if it were the whole session.
A filter reduces the population
Filtered counts and rates can move sharply when few visits match. Read the absolute values before treating a percentage change as a general pattern.
Useful ways to apply this
Check a campaign source
Filter by UTM source or campaign, then compare visitors, engagement and the matching conversion goal for that visit context.
Investigate a device issue
Filter by device after a release, then compare engagement and conversions with the unfiltered baseline before deciding whether the issue is device-specific.