Guide 70 · Reports · GA4 + Tag Manager
Segments versus audiences in GA4: which one to build, and where each one lives
A segment is a temporary lens for analysis; an audience is a saved group that other Google products can use. Build a segment to ask a question, an audience to act on the answer.
After step 5, you can state which of the two you built, where it lives, who can see it, and what it can be exported to.
Download the report reference: the Segment vs audience reference (JSON) (1369 bytes, MD5 b447c229048810ae4b34bc2c08366943). It holds the report or exploration definition and the worked example with the recomputed figures shown on this page.
What you need
- Viewer access or above to build a segment; the Editor or Administrator role to create and manage audiences in Admin.
- The group defined in one sentence — the same sentence for both, so the comparison is fair.
- The decision made before you build: is this group for analysis, or for remarketing. That decision picks the tool.
The trace
- Write the group down once. For example: users who viewed the pricing page and did not purchase within 7 days. One sentence, one group. Both tools will be built from this sentence.
- Build it as a segment first. In Explore, create a user segment from the sentence and apply it to a free-form exploration. The segment is a lens: it changes what the table shows for you, in this exploration, right now. It does not change anything for anyone else.
- Read what the segment tells you. Read the segment's sessions, engagement rate, and key event rate beside All users. The segment answers whether the group behaves differently. If it does not, stop — an audience built from a group that behaves like everyone else is a list with no leverage.
- Build it as an audience only if you will act on it. In Admin, then Data display, then Audiences, create an audience from the same conditions. An audience is saved to the property and can be exported to linked advertising products — for example, as a remarketing list in Google Ads — so it is a commitment to use the group, not just to look at it.
- Check who can see each one. A segment lives in explorations and is yours to apply. An audience lives in the property and, once exported, appears in the linked advertising product. Sharing an exploration gives a colleague a read-only view; it does not give them your segment to edit.
Dead ends
DEAD END — BUILDING AN AUDIENCE TO ANSWER A QUESTION. An audience is for action in another product. A question belongs to a segment. Building an audience to read a table is the wrong tool and leaves a list behind that someone may use as a targeting list.
DEAD END — EXPORTING AN AUDIENCE BUILT ON DEMOGRAPHICS. Audiences that contain demographics or interests data are not exported to advertising products. A segment can be built on demographics; an audience built the same way cannot leave the property.
DEAD END — ASSUMING A SHARED EXPLORATION SHARES ITS SEGMENTS. Sharing an exploration gives a read-only view. The colleague must duplicate or copy the shared exploration to edit it, and the segment travels with the copy, not as an editable shared object.
End state
Segments are for analysis; audiences are for action. The worked example below recomputes the rates for the group in question, so the decision to build an audience rests on numbers, not on a hunch.
Worked example — stated inputs, recomputed results. The groups below are stated inputs for the example. Every rate in the table was recomputed from them.
| Row | Sessions | Engaged sessions | Engagement rate | Key events | Key event rate | Revenue |
|---|---|---|---|---|---|---|
| Viewed pricing, no purchase | 920 | 644 | 70.0% | 0 | 0.0% | $0.00 |
| All users | 4,390 | 2,551 | 58.1% | 260 | 5.9% | $5,420.50 |
| Total | 5,310 | 3,195 | 60.2% | 260 | 4.9% | $5,420.50 |
Revenue per session is $5,420.50 ÷ 5,310 sessions = $1.02. Engagement rate is 3,195 ÷ 5,310 = 60.2%.
Sources
- Analytics Help — Get started with Explorations — explorations let you create segments and audiences; shared explorations can be viewed but not edited, and must be duplicated to edit
- Analytics Help — Confirm that you are collecting data — many reports and explorations can take 24 to 48 hours to process data
Checked 11 November 2026. The click paths above are the ones in Google's help pages, opened or read 11 November 2026 (the Explorations and detail-report pages were opened in full; the report, library, navigation and thresholds pages were read from Google's help results the same day, after the browser hit a rate limit). No step in this guide was run inside a live Analytics property, because that needs a Google sign-in and this site does not sign in as anyone. The worked-example figures on this page were recomputed in Python with decimal arithmetic, and the downloadable JSON file carries the same recomputed figures. Google moves menu items without warning; if a path on this page no longer matches, the linked help page is the current map.