.svg)
Segmentation for retention campaigns in Braze starts with three decisions made before anything sends: who counts as drifting away, who must be left alone, and how you will measure the outcome. Build the audience from recency and engagement filters, then protect it with exclusion groups. Check its reachable size and judge the result with retention reports.
Key Takeaways
- A retention segment needs a definition of drift, a list of people to protect and a measure of success.
- Exclusion groups connect to filter groups with an and-not operator and override segment criteria, so write your protections there.
- Using the or operator across negative filters on one attribute targets every user.
- Estimated segment counts come from a sample, so calculate exact statistics and check channel reachability before you send.
- Campaign retention reports track a retention event you choose over the 30 days after receipt.
- Segment Extensions reach back two years (730 days) for behaviour that runs on a long cycle.
What does segmentation for retention campaigns in Braze involve?
Braze describes segments as dynamic groups of users who fit criteria you define, such as user attributes, user behaviour and custom events. For retention work that matters, because the audience is never a fixed list. People move in and out as they use the product.
Braze updates a user's segment membership as data is processed, typically instantaneously. Its documentation gives a lapsed-user example: a user in a lapsed segment moves out as soon as the new session is processed. Someone who has just come back shouldn't get a message written for someone who left.
A retention segment is a statement about behaviour over time. In our view it needs a definition of drift, a list of people to protect, and a measure of success. We recommend writing all three down before you open the segment builder.
Which filters show that a user is drifting away?
Braze documents a range of recency, messaging and predictive filters. The table lists four that sit at the centre of retention audiences. Each row gives the documented description and our reading of how to use it.
| Filter | What Braze documents | How we would use it |
|---|---|---|
| Last Used App | Segments users by the latest time they opened your app | The base of a drift definition |
| Last Received Email | Segments users by the last time they received one of your email messages | Keeps a win-back from landing on top of last week's email |
| Churn Risk Category | Segments users by churn risk category according to a specific prediction | Useful once a prediction exists for your workspace |
| Segment Membership | Filters on segment membership anywhere filters are used | Reuses an audience you already trust |
We recommend defining drift against the product's own rhythm. A grocery app and an annual-filing app have different natural gaps between visits. One shared inactivity threshold will be wrong for at least one of them.
Churn Risk Category depends on a prediction, so check one exists before you build on it. If your workspace has none yet, CustomerIK's marketing analytics service covers predictive churn modelling.
One boundary is worth stating early. Profiles are created once a user has used the app for the first time, so a retention segment can't include people who have never opened it. Those users belong to an activation programme instead.
How do you build a retention segment step by step?
Follow the order below. It puts the protective steps before the launch, which is where retention audiences tend to go wrong.
- Go to Audience, then Segments, and create the segment. Name it after the user you intend to target, because vague segment titles can be confusing.
- Choose the apps or platforms the segment covers.
- Add your recency filter, then one or two engagement filters that describe the behaviour you want to see again.
- Add an exclusion group for anyone who should not hear from you.
- Select View funnel statistics and read how each filter changes the count.
- Test the logic by looking up a known user. The documentation says to search for a user's external_id or braze_id in the User Lookup section.
- Calculate exact statistics before the audience is attached to a campaign or Canvas.
The funnel view rewards patience. Braze notes that adding filters that document your user flow shows the points where users fall off. Joining filters with and shrinks the count, while joining them with or grows it.
Why do exclusion groups matter more in retention than in other programmes?
Exclusion groups always connect to your filter groups with an and-not operator, and they override segment criteria. If a user falls into the exclusion criteria, that user is out, even when every inclusion filter matches.
That makes the exclusion group the safest place to encode your promises. Anyone you must never message goes there once, so it never has to be written into every inclusion filter by hand.
In our view, retention messages carry extra risk because they go to people who have already shown low interest. A badly aimed message does more damage to that audience than to a loyal one. Our suggestion is to write the exclusions first and the inclusions second.
How can the OR operator undo a retention segment?
The documentation warns about one pattern in particular. Take the negative operators: is not, does not equal, not included and does not match regex. Use any of them with the or operator two or more times against the same attribute, and users with all values for that attribute will be targeted.
The reason is simple logic. Take an attribute that holds one value per user, so a user with value A passes the is-not-B filter and a user with value B passes the is-not-A filter. Everyone passes, so use and between those filters or move the negative conditions into an exclusion group.
How accurate is the segment size before you send?
Estimated counts come from a sample, and Braze says estimated statistics are cached for 15 minutes unless you edit the segment. Very small segments can show an estimated range that includes 0, so a niche win-back audience may read as empty when it isn't.
Select Calculate exact statistics for the real figure. Braze states that exact segment membership is always calculated before a message from a campaign or Canvas affects the segment.
Total reachable users includes everyone who matches your filters, even people who are no longer subscribed to a channel. The documentation suggests adding a filter such as push enabled for iOS to align the total with subscribed users. It also lists failed channel reachability checks at send time as one reason send counts can fall below the estimate.
When do Segment Extensions help with retention work?
Segment Extensions let you build precise segments over an extended period of a user's history. You can reach users who completed a custom or purchase event any number of times in the past two years (730 days).
The documentation calls Segment Extensions an enhancement and says they are optional. For retention, we suggest reaching for them when the behaviour you care about runs on a long cycle, such as an annual renewal or a seasonal purchase. For a recency gap measured in weeks, the standard filters are the simpler tool.
What should you measure after the campaign sends?
Campaign retention in Braze shows the rate at which users who received the campaign performed a retention event you specify over the 30 days from receipt. You choose between two types. Rolling retention counts users who return on or after each listed day, while range retention counts users who return within each listed range.
We suggest picking a retention event that proves the person came back to the product, such as a session or a purchase. A message open shows the email was seen, which is a different question.
Retention reports aren't available for API-triggered campaigns, so decide how you will measure those before you build them. For a broader view, Segment Insights compares one segment against a baseline across pre-selected KPIs, including the average days since the segment's users' last session. Its statistics are estimated by default.
What does a worked example look like?
Take a hypothetical subscription retailer that wants to win back customers who have stopped opening its app. This is an illustration. It is not client work and it carries no results.
The retailer first defines drift against its own reorder cycle and avoids a round number. It then builds one filter group: the Last Used App gap is longer than that cycle, and the customer has purchased before. It adds an exclusion group for anyone who received one of its emails recently, using the Last Received Email filter.
Funnel statistics show how much each filter removes. The team calculates exact statistics, then reads the reachable users by channel. If push reach is thin among the lapsed group, the team chooses email as the lead channel and decides that before writing copy.
After sending, it reads the campaign retention report with a purchase as the retention event. It compares the segment against the all-users baseline in Segment Insights. A returning user leaves the lapsed segment on their next processed session, so the audience shrinks as the campaign works.
How does this connect to the rest of the programme?
Segments are the foundation, and the surrounding decisions decide whether they pay off. Our guide to segmenting your audience for better engagement covers the wider audience model that retention segments sit inside.
Timing is the next question, and triggering retention messages at the right time deals with it directly. For the automation side, see reducing churn with retention automation.
Measurement has its own post on retention metrics worth tracking. The audience you built here feeds directly into win-back campaigns. The wider set of tactics is in customer retention strategies using Braze.
Retention audiences also pair with creative mechanics. A case study on re-engaging lapsing users with gamification covers how an interactive reward was built into automated lifecycle communications.
As a Braze implementation partner, CustomerIK recommends writing the exclusion group before the inclusion filters. We also recommend naming every retention segment after the behaviour it targets. Both habits make the segment easier to audit later.
Frequently Asked Questions
1. How does Braze keep a retention segment up to date?
Segments update in real time as data changes, and membership is updated as user data is processed, typically instantaneously. A user who returns leaves a lapsed segment once the session is processed.
2. Can you target users who have never opened the app?
No. Profiles are created once a user has used the app for the first time, so those users can't be targeted by a segment.
3. What happens if you archive a retention segment?
Any campaigns or Canvases using it are archived too, even when the segment sits in a single Canvas component. Braze shows a warning that lists what will be archived first.
4. Should you use Segment Extensions for retention?
Use them when the behaviour you need sits further back than standard filters reach, up to two years (730 days). The documentation describes them as an enhancement and not a requirement.
5. Why might a user get a message after no longer matching the segment?
A user may have been in the segment at send time, then changed behaviour on any of their devices afterwards. Membership at send time is what counts.
Sources
- Braze, Create a segment. Retrieved 28 Oct 2026.
- Braze, Measure segment size. Retrieved 28 Oct 2026.
- Braze, Segment Insights. Retrieved 28 Oct 2026.
- Braze, Segment Extensions. Retrieved 28 Oct 2026.
- Braze, Segmentation filters. Retrieved 28 Oct 2026.
- Braze, Retention reports. Retrieved 28 Oct 2026.
- Braze, Segments. Retrieved 28 Oct 2026.
- Written by CustomerIK, a Braze implementation partner.
CustomerIK is a Braze implementation partner working across onboarding, technical integration, marketing operations and customer data management.
If you want your retention segments reviewed before the next win-back goes out, let's talk.
.svg)
