GA4 Segments: Building Reusable Audience Segments That Actually Work (2026)

How do GA4 segments work?

GA4 segments filter users, sessions, or events to a subset you define, allowing comparison against all users or other segments in Explorations. There are three segment types: user segments (all sessions from users who meet the criteria, ever), session segments (only sessions where the criteria were met), and event segments (only events matching the criteria). Scope matters enormously. A user segment for "users who purchased" includes all sessions from those users — including sessions before and after the purchase.

A session segment for "sessions containing a purchase" includes only the sessions where a purchase occurred. Same users, very different numbers. The right scope depends on your question: use user segments for lifetime behaviour analysis, session segments for visit-level analysis, event segments for granular event debugging.

The three segment types in detail

User segments

Captures all users who meet your criteria, then includes all their activity — not just the sessions where the criteria were met.

Example: A user segment for "purchased at least once" includes the user's pre-purchase research sessions, their purchase session, and all post-purchase sessions. This is the right scope for:

• Understanding the full research-to-purchase journey
• Cohort analysis (what did high-LTV users do in their first sessions?)
• Audience building for remarketing

Condition timing options:

• Across all sessions: The condition must have been met at any point in the user's history
• Within the same session: The condition must have been met in a single session
• Within the same event: The condition must appear on a single event

Session segments

Captures only sessions where your criteria were met. Users appear only when their session-level behaviour matches.

Example: A session segment for "sessions with a purchase" includes only the sessions containing purchase events — not the user's other sessions.

Use for:

• Comparing converting sessions vs non-converting sessions
• Analysing what distinguishes sessions that resulted in a key action
• Channel-level behaviour analysis (sessions from paid search vs organic)

Event segments

Captures individual events matching criteria. Rarely used in Explorations (most questions are session or user level), but useful for:

• Isolating specific parameter values in event-level analysis
• Debugging event firing patterns

8 high-value segment recipes

Segment 1 — High-intent non-converters (user)

Users who visited a product page and added to cart but did not purchase.

Configuration:

• Condition 1: Event add_to_cart — at least 1 time
• Condition 2: Event purchase — 0 times
• Scope: User

Use: Remarketing audiences, UX analysis of abandonment behaviour, email re-engagement.

Segment 2 — Paid traffic sessions (session)

Sessions from paid channels only.

Configuration:

• Condition: Session default channel group exactly matches Paid Search OR Paid Social OR Display
• Scope: Session

Use: Compare paid traffic quality (engagement, conversion) vs organic. Identify if paid sessions convert at a fundamentally different rate.

Segment 3 — Highly engaged users (user)

Users with above-average engagement time and multiple sessions.

Configuration:

• Condition 1: Session count > 2 (multiple sessions)
• Condition 2: Total engagement time > 300 seconds (5 minutes cumulative)
• Scope: User

Use: Understanding what engaged power users do differently. Product roadmap insight: if engaged users use feature X consistently, feature X may be underexposed to new users.

Segment 4 — Mobile non-converters (session)

Sessions from mobile devices that did not result in a key event.

Configuration:

• Condition 1: Device category = mobile
• Condition 2: Key events = 0
• Scope: Session

Use: Identify mobile UX problems. Compare this segment's path and funnel behaviour vs the same segment on desktop. High mobile non-conversion vs desktop indicates mobile CRO opportunity.

Segment 5 — New users from organic search (user)

Users whose first session came from organic search.

Configuration:

• Condition: First user default channel group = Organic Search
• Scope: User

Use: Understanding the quality of SEO-acquired users. Compare LTV and conversion rates vs paid acquisition.

Segment 6 — Recent purchasers (user)

Users who purchased within the last 30 days.

Configuration:

• Condition: Event purchase — at least 1 time — within last 30 days
• Scope: User

Use: Build a remarketing audience for upsell/cross-sell. Analyse what recent purchasers do post-purchase (do they return? Do they engage with content?).

Segment 7 — Bounce sessions (session)

Sessions with a single page view and no engagement.

Configuration:

• Condition 1: Session engagement time < 10 seconds
• Condition 2: Session page views = 1 (add this as an additional condition in a Free Form exploration by filtering the Screen page views metric)
• Scope: Session

Note: In GA4, "non-engaged sessions" (not Engaged sessions) is a proxy for bounced sessions. Use the calculated 1 - Engagement rate metric as a simpler alternative in standard reports.

Segment 8 — Returning high-value users (user)

Users who have had more than 3 sessions AND purchased at least twice.

Configuration:

• Condition 1: Session count > 3
• Condition 2: Event purchase — at least 2 times
• Scope: User

Use: Your most loyal customers. Understand their behaviour to inform retention strategy. What content do they consume between purchases? What channels do they use?

Segments vs audiences: the key difference

Segments exist only within Explorations. They filter data in the analysis interface and are not exported to Google Ads or other platforms.

Audiences are segment-like definitions in GA4 Admin that build real user lists over time. Audiences:

• Populate continuously as new users meet the criteria
• Can be published to Google Ads for remarketing
• Appear in GA4 standard reports as a dimension
• Have a maximum membership duration (up to 540 days)

You can convert a segment to an audience in GA4: after building a segment in Explorations, click the three-dot menu next to the segment name → "Build audience." This creates an equivalent audience definition in GA4 Admin.

Key distinction: A segment on "users who purchased in the last 30 days" in an Exploration shows historical data. An equivalent audience in GA4 Admin builds a real-time list of users who have purchased in the last 30 days, continuously updated.

Common segment configuration errors

Error 1 — Using user scope when session scope is needed If you want to compare "sessions from paid search" vs "sessions from organic," use session segments, not user segments. User segments for "users who came via paid search" include all those users' organic sessions too, producing misleading comparisons.

Error 2 — Zero results from overly specific conditions Combining too many AND conditions with exact parameter matches can produce zero results. Test each condition individually in a Free Form exploration before combining them. A condition returning zero results in isolation means the event name or parameter value doesn't match what's actually in GA4 (check capitalisation, spelling, and parameter key names).

Error 3 — Condition timing mismatch A segment for "users who added to cart AND purchased in the same session" requires both conditions to have timing set to "Within the same session." If timing is set to "Across all sessions," the segment captures users who added to cart in any session and purchased in any session — including users who abandoned cart and returned months later.

Error 4 — Sampling invalidating comparisons When comparing two large segments in a single Exploration, sampling may apply unevenly — one segment might be sampled at 70%, the other at 90%. The resulting comparison is distorted. Always check the sampling indicator for each segment in the exploration header.