GA4 Filters and Segments: How to Slice Your Data
By Emily Redmond, Data Analyst at Emilytics · April 2026
TL;DR: Filters hide data (show only organic traffic); segments compare groups (converters vs. non-converters). Both are essential for focused analysis. Filters are in reports; segments are in explorations.
Raw data is overwhelming. You need to zoom in. GA4 gives you two tools: filters (isolate specific data) and segments (compare groups). Know when to use each, and your analysis gets infinitely sharper.
Filters vs. Segments at a Glance
| Aspect | Filters | Segments |
|---|---|---|
| Purpose | Show only certain data | Compare two groups |
| View | Hide the rest | Side-by-side comparison |
| Use Case | "Show me only mobile" | "Compare mobile vs. desktop" |
| Where | Standard reports, explorations | Explorations |
| Impact | Removes data from view | Keeps all data, labels it |
Filters: Zoom In
A filter shows only data that meets a condition. Use filters when you want to focus on a subset.
When to Use Filters
- Isolate a traffic source: "Show me only paid search"
- Focus on a device: "Show me only mobile"
- Segment by geography: "Show me only USA"
- Exclude test data: "Exclude internal IP addresses"
- Focus on a page: "Show me only /pricing page visitors"
Creating a Filter in a Standard Report
- Open any standard report (e.g., Acquisition → Source/Medium)
- Click Add filter (or find the filter icon)
- Choose a dimension to filter by (e.g., "Device category")
- Choose an operator (contains, equals, doesn't contain, etc.)
- Enter the value (e.g., "Mobile")
- Apply
Now the report shows only mobile data.
Creating a Filter in an Exploration
Same process:
- Create a freeform exploration
- Under the dimensions/metrics panel, find Filter
- Click Add filter
- Configure dimension, operator, value
- Apply
Filter Operators
GA4 offers several operators:
- = (Equals): Exact match
- ≠ (Doesn't equal): Anything except this
- ~ (Contains): Substring match
- !~ (Doesn't contain): Exclude substring
- Matches regex: Advanced pattern matching
- Begins with / Ends with: Prefix/suffix matching
Example: Filter "landing page" "contains" "product" to show all product-related pages.
Common Filters to Set Up
| Goal | Filter |
|---|---|
| Exclude internal traffic | Country ≠ [your country] OR IP ≠ [your office IP] |
| Show only organic search | Source = "organic" |
| Show only paid traffic | Medium = "cpc" or "cpm" or "display" |
| Show only mobile | Device category = "mobile" |
| Show only desktop | Device category = "desktop" |
| Show only a specific campaign | Campaign = [campaign name] |
| Show only conversions | Conversions > 0 |
| Show only a specific page | Landing page = "/pricing" |
💡 Emily's take: I always start an exploration with filters. "Show me organic traffic only, desktop only, from the USA." Narrow it down. Otherwise you're looking at a noisy average that doesn't tell you anything actionable.
Segments: Compare
A segment lets you compare two user groups side-by-side. Use segments when you want to see the difference between groups.
When to Use Segments
- Compare converters vs. non-converters: Do they behave differently?
- Compare traffic sources: Does organic traffic convert better than paid?
- Compare new vs. returning users: Which group is more valuable?
- Compare high-engagement vs. low-engagement: What's the difference?
Creating a Segment in an Exploration
Segments are only available in explorations (not standard reports).
- Create a freeform exploration
- In the Settings tab, find Segment comparison
- Click Add segment
- Define the condition (e.g., "Conversions > 0")
- Add another segment if comparing two groups
- Run
GA4 now shows metrics for each segment side-by-side.
Example: Converters vs. Non-Converters
Segment 1: Conversions > 0 (people who converted) Segment 2: Conversions = 0 (people who didn't)
Metrics: Users, average session duration, pages per session, scroll depth
Result:
- Converters: 8 min avg duration, 5 pages/session, 70% scroll depth
- Non-converters: 2 min avg duration, 2 pages/session, 25% scroll depth
Insight: Converters spend way more time, see more content, and scroll deeper. This means your content is working—it's a sales tool. People who don't see it don't convert.
Example: Organic vs. Paid
Segment 1: Source = "organic" Segment 2: Source = "google" (paid search)
Metrics: Users, conversion rate, average session duration, bounce rate
Result:
- Organic: 40% conversion rate, 5 min duration
- Paid: 15% conversion rate, 2 min duration
Insight: Organic visitors are 2.7x more likely to convert. They're more qualified. This affects your budget allocation—organic search might be your best channel despite lower volume.
Pre-Built Segments
GA4 includes some pre-built segments you can use without creating them:
- Converters: Users who completed a conversion
- Non-converters: Users who didn't convert
- Active users: Users with at least one engagement event
- Users who engaged with your site: Same as active users
You can use these directly in explorations without custom setup.
Filters vs. Segments: When to Use Each
Use a Filter When
You want to eliminate noise. Show only the data you care about.
Example: "Show me only mobile users who converted."
Use a Segment When
You want to understand differences. Compare two groups.
Example: "Compare mobile converters vs. mobile non-converters."
Combining Both
Often you'll use both: filter to focus, segment to compare.
Example:
- Filter: Device category = "mobile" (focus on mobile)
- Segment 1: Conversions > 0 (converters)
- Segment 2: Conversions = 0 (non-converters)
- Compare: See how mobile converters differ from mobile non-converters
Data Retention and Filtering
Important: Filters affect data availability.
Example: If you filter out internal traffic using a filter in GA4 (via Admin → Data filters), you're actually removing that data from your property. You can't get it back.
But filters in reports (like "Show only organic") don't remove data—they just hide it. You can remove the filter and see everything again.
Tip: Use Admin data filters for test traffic and internal traffic you never want to see. Use report filters for temporary analysis.
Advanced Filtering with Regex
For complex conditions, GA4 supports regex (regular expressions).
Example: Show pages with "product" OR "pricing" OR "plan":
Page path matches regex: ^/(product|pricing|plan)
Regex is powerful but requires some learning. Use the "Matches regex" operator and test carefully.
Frequently Asked Questions
Q: Can I save a filter so I don't have to set it up every time? A: Not directly for report filters. But you can save explorations with filters baked in, and re-open them.
Q: What happens to my data if I apply a filter? A: It depends. Report filters just hide data (you can remove them and see it again). Admin data filters actually remove data from your property (permanently).
Q: Can I combine multiple filters (AND logic)? A: Yes. Add multiple filters and GA4 applies AND logic (both conditions must be true).
Q: Can I combine multiple filters with OR logic? A: Not directly. You'd need regex or multiple explorations.
Q: Do filters affect my historical data? A: Report filters don't—they only affect what you see. Admin data filters affect all historical data in the property.
Q: Why is my segment empty? A: Your segment condition didn't match any users. Example: "Sessions with revenue > $1,000" might have zero matches if your average order value is lower. Check your condition.
The Bottom Line
Filters focus your view; segments compare groups. Both are essential for moving beyond surface-level metrics.
Master basic filtering first (device, traffic source, geography, page). Then learn segments for comparing groups. Once comfortable, explore regex for advanced conditions.
The sharper you can slice data, the clearer your insights. And clear insights lead to better decisions.
For building custom audiences based on similar logic, see GA4 Audiences: How to Build and Use Them for Retargeting.
Emily Redmond is a data analyst at Emilytics — the AI analytics agent that watches your GA4, Search Console, and Bing data around the clock so you never miss what matters. 8 years of experience helping founders and growth teams turn data noise into clear decisions. Say hi →