When you look at website data, the hardest part isn’t finding numbers—it’s trusting them. Maybe an overseas bot farm spiked sessions, or a partner’s QA clicks inflated conversions. This guide shows you how to filter out “unwanted” traffic and compare the segments that matter—by country, device, campaign group, or anything GA4 can dimension—without breaking your baseline reporting.
Below you’ll find a clear workflow, plus exactly where to add screenshots and what each one should display so you can publish this as a step-by-step tutorial.
What you’ll learn
- The modern way to filter and compare traffic in GA4 (and why it’s safer than permanent filtering).
- How to use Comparisons in standard reports to include/exclude countries, devices, channels.
- How to build segments in Explore to compare audiences side by side across any metric.
- When to use property-level data filters (and when not to).
Quick context: In Universal Analytics, many teams relied on permanent View Filters. In GA4, everyday analysis is done with Comparisons and Segments—non-destructive, easy to duplicate, and perfect for “what-if” analysis.
The toolbox at a glance
- Comparisons (Reports UI): Fast, per-report includes/excludes (e.g., “Device category = mobile” vs. “desktop”). Great for quick answers and screenshots for stakeholders.
- Segments (Explore): Reusable logic (user/session/event scopes) for deeper analysis, cohorts, sequences, funnels.
- Data Filters (Admin): Property-level rules that typically exclude internal or developer traffic from all future processing.
Use Comparisons and Segments for analysis and decisions; reserve Data Filters for long-term hygiene (e.g., excluding your team).
Part 1 — Quick filtering with Comparisons (Reports)
Best for: “Show me how mobile traffic compares to desktop,” “Exclude Country X,” “Only see Paid Social.”
Steps
- Go to Reports → Acquisition → Traffic acquisition (or any standard report).
- At the top right, click Add comparison.
- Choose a dimension (e.g., Country, Device category, Default channel group).
- Add a condition:
- Include Country = United States
- Exclude Country = India
- Include Device category = mobile
- Add a second comparison to show the contrasting group (e.g., desktop).
- Apply. You now get side-by-side cards and tables for each slice.
Where to place Screenshot #1 — “Comparison Builder”
Place directly under these steps.
Show: The Add comparison panel with Dimension: Device category, Include: mobile, then a second comparison Include: desktop. Visually highlight that two comparisons can run at once.
Where to place Screenshot #2 — “Comparisons Applied”
Place right after the first image.
Show: The report with two color-coded comparison chips at the top and the table/cards showing split metrics (sessions, engagement rate, conversions) for each slice.
Pro tips
- Stack comparisons sparingly—two or three max—so charts remain readable.
- Save the page URL with comparisons applied for quick sharing (the chips persist in the link).
- Use this technique to sanity-check suspected anomalies in specific countries or devices.
- If you’re looking for a broad explanation of legacy permanent filters, look up what is a view filter google analytics—but remember, GA4 leans on safer, non-destructive tools for day-to-day analysis.
Part 2 — Deeper analysis with Segments (Explore)
Best for: Rich, reusable logic—e.g., “Users from Tier-1 countries on mobile who viewed pricing,” “Sessions from Paid Search with high engagement,” “Event-based segments like ‘Viewed product AND Added to cart’.”
Steps (Free-form exploration)
- Go to Explore → + Blank → Free form.
- In the left panel, click Segments → + to create a User, Session, or Event segment.
- Add conditions such as:
- User segment: Country in {US, CA, UK} AND Device category = mobile
- Session segment: Default channel group = Organic Social AND Session medium contains “social”
- Event segment: Event name = view_item AND item_category = “Shoes”
- Click Apply to add your segment(s) to the canvas.
- Drag metrics (Sessions, Engaged sessions, Conversions, Revenue) into the Values area; drag Dimensions (Page path, Landing page, Country, Device category) into Rows/Columns.
- Add a second segment to compare side by side (e.g., Tier-1 Mobile vs. Tier-2 Desktop).
Where to place Screenshot #3 — “Segment Builder”
Place under the steps.
Show: The segment configuration screen with scope (e.g., User), conditions (Country set, Device category = mobile), and Preview counts on the right.
Where to place Screenshot #4 — “Segmented Table”
Place after the builder screenshot.
Show: A Free-form table with two segments applied (e.g., “Tier-1 Mobile Users” vs. “Desktop All Countries”) in columns, and Landing page in rows, plus metrics (Engaged sessions, CVR, Revenue). The contrast should be obvious at a glance.
Use the phrase ga4 segments once when introducing this capability, then rely on “segments” thereafter. If you prefer the older naming, many marketers still say google analytics custom segments—once is enough to map the vocabulary.
When (and when not) to use Admin-level Data Filters
Data filters in GA4 live under Admin → Data settings → Data filters. They’re designed to permanently exclude traffic types like internal or developer traffic. They are not meant for routine country/device filtering because they change processed data going forward.
Where to place Screenshot #5 — “Data Filters List”
Place here.
Show: Admin → Data settings → Data filters, with one Internal filter in Active state and one Developer filter in Testing state, so readers see the difference.
If you’re coming from UA and searching for google analytics filters, think of GA4’s Admin filters as surgical (internal/dev) rather than flexible. For everyday slicing (countries, devices), favor Comparisons and Segments so you never lose data.
Practical scenarios (repeatable recipes)
A) Remove noise from non-target countries
- Goal: Your business sells only in NA/EU, but you see spikes from elsewhere.
- Do this: Add two comparisons—Include Countries {US, CA, UK, DE, FR} vs. Exclude them—to visualize impact on engagement and conversions.
B) Device-specific UX checks
- Goal: Mobile conversions dipped after a layout change.
- Do this: In Explore, create two Session segments: Device = mobile vs. desktop; add Landing page rows and Conversion rate to values. Look for pages where mobile CVR lags most.
C) Channel mix quality
- Goal: Compare bounce-y Paid Social clicks to high-intent Organic Search.
- Do this: Use Comparisons on Default channel group; then confirm in Explore with Session segments and add Average engagement time, Scroll depth (if tracked), and CVR.
D) Botty burst from one market
- Goal: Sudden traffic spike from a single country with near-zero engagement.
- Do this: Comparison by Country (include the spike country vs. all others). If confirmed, adjust campaign targeting and annotate your report.
Interpreting what you see (and avoiding traps)
- Non-destructive first: Start with Comparisons/Segments. If a pattern is consistently noise (e.g., your own office), graduate to an Admin Data filter.
- Scope matters: User vs. Session vs. Event segments can tell different stories. A user might visit from multiple devices over time; a session captures one visit; event segments are great for precise behaviors.
- Small segments = volatile rates: When you slice narrowly (e.g., mobile users in one small country), expect more noise. Look at absolute counts alongside rates.
- Share context, not just charts: Always annotate comparisons (what’s included/excluded) before sending screenshots to stakeholders.
Suggested structure for your article’s screenshots (recap)
- Comparison Builder — Add comparison → Include Device category = mobile; second chip for desktop.
- Comparisons Applied — Report with two colored chips and split metrics in cards/table.
- Segment Builder (Explore) — User segment with Country set + Device category = mobile; preview counts.
- Segmented Table (Explore) — Two segments side by side with Landing page rows and conversion metrics.
- Data Filters (Admin) — Internal (Active) and Developer (Testing) to demonstrate the limited, surgical use of property-level filters.
Governance & collaboration tips
- Name things clearly: “Mobile-Tier1-Users” beats “Seg-3.” You’ll thank yourself in a month.
- Keep a segment library: Store common slices (Target Markets, Key Devices, High-Intent Referrers) and reuse across explorations.
- Document comparisons used in reports: A one-line note (“Comparisons: US+CA vs. Rest, Desktop vs. Mobile”) prevents misreads.
- Review quarterly: Markets evolve. Retire stale segments and update country lists as you expand.
Final word
If you treat your analytics like a lab—clean inputs (Admin filters for internal traffic), controlled experiments (Comparisons), and repeatable analyses (Segments)—you’ll quickly separate meaningful performance changes from background noise. Start simple with a couple of comparisons, validate with segments in Explore, then decide whether any rule deserves to be permanent.
One-time keyword usage (for clarity mapping)
- what is a view filter google analytics — legacy UA concept for permanent view-level filtering.
- ga4 segments — build and compare audiences in Explore.
- google analytics custom segments — alternate phrasing marketers still use for GA4 segments.
- google analytics filters — broad concept; in GA4, reserve Admin filters for internal/dev traffic.