How to Build a Retention-First Culture with Braze

Hand pointing at the green stripes at the foot of a tram stop timetable, showing customers coming back as the return journey in a retention programme.

A retention-first culture means a team judges each campaign by whether customers come back, with opens mattering less. In Braze, that habit rests on four documented tools: retention reports, conversion events, Canvas exit criteria and Predictive Churn. This post sets out the order to configure them and the review habits that keep the numbers honest.

Key Takeaways

  • A retention-first culture judges each campaign by whether customers come back.
  • Define churn first: what a user does, or stops doing, in what time frame.
  • Read the control group line in the retention report to see what happened with no message.
  • Set conversion events deliberately: up to four per campaign, a window of up to 30 days, and a default of three days.
  • Add an exit event so that returning customers stop receiving the journey.
  • Treat a Churn Risk Score between 0 and 100 as a ranking, and set the cost of a false positive per campaign.

What does a retention-first culture mean in practice?

A retention-first culture is a team habit, and you can't switch it on with a feature. It shows up in the campaign review. People ask whether customers came back, and what would have happened if we had sent nothing.

Braze gives the team an instrument for each question. Retention reports measure return behaviour, and conversion events record the action you wanted. Exit criteria stop messaging that no longer fits, and Predictive Churn scores who is drifting away.

We recommend settling the order of setup before any single feature. The order matters more than any one tool.

Why should the team define churn before building anything?

Predictive Churn begins with a definition. What does a user have to do, in what time frame, for you to consider them churned? The documentation asks you to think of churn as something a user either does once or stops doing.

The churn window can be set for up to 60 days, depending on the available data. That choice shapes which users the model learns from, so it deserves a decision rather than a default.

We recommend settling the churn definition first, because every later choice leans on it. If the retention event, the conversion event and the exit event each describe a different behaviour, the campaign review becomes an argument about definitions.

What should a retention report actually measure?

Braze lets you measure user retention on the Analytics page of a campaign or Canvas. The retention reports documentation covers the detail.

Retention shows the rate at which users performed a retention event you specify. The window is the 30 days from the time they received the campaign or Canvas.

The report offers both a rolling and a range retention formula. Its by-variant view includes the control group's retention data. That shows what retention looks like if you didn't send a message.

We suggest reading the control group line first, because it alone answers whether the message changed behaviour. The documentation also asks where the impact starts to decline. That point may call for a new campaign or Canvas aimed at that moment.

Which Braze tool answers which review question?

Each tool answers a different question, and the review works best when every campaign is checked against all of them. The table maps each question to the feature and to what the Braze documentation says it does.

Review questionBraze featureWhat the documentation says
Did customers return after the message?Retention reportShows the rate at which users performed a retention event over the 30 days after receiving a campaign or Canvas
What would have happened with no message?Retention report by variantLets you view retention for the control group
Did they take the action we wanted?Conversion eventsCounts a conversion within a window of up to 30 days
Who should stop receiving this?Exit criteriaRemoves users from the journey when they perform a specific action
Who is likely to leave next?Predictive ChurnAssigns a Churn Risk Score between 0 and 100

How do conversion events keep the review honest?

A conversion event tracks a high-value action after a recipient gets your messaging. A campaign can carry up to four conversion events. You can set a window of up to 30 days, and if you select none, the time defaults to three days.

A user counts as converted if they performed the event in the window. They needn't open or click the message. For a campaign aimed at lapsing users, the documentation suggests adding a secondary conversion event beside the primary Starts Session event.

We recommend choosing a conversion window that fits the behaviour your churn definition names. Write the reason in the campaign brief, so the next reviewer can tell a deliberate window from a leftover default.

When should a journey stop messaging someone?

By adding exception events to the Canvas entry rules, you can remove users from the journey when they perform a specific action. Braze marks a user to exit as soon as they perform the exit event, and they don't advance to any later steps. A user in a Delay step remains there until the delay period ends.

The documentation notes that exception events keep your messaging relevant and personalised. The frequency capping documentation adds the cost of getting it wrong.

A user who receives too many messages within a short time will feel over-encumbered. They either turn off push notifications or uninstall the app.

We recommend treating an exit rule as a retention tool. A win-back journey that keeps messaging a customer who has already returned spends goodwill for no gain.

How should the team act on churn risk scores?

Each user in the prediction audience gets a Churn Risk Score between 0 and 100. A higher score means a greater likelihood of churn.

Scores from 0 to 50 sit in Low Risk, 50 to 75 in Medium Risk, and 75 to 100 in High Risk. Braze retrains the prediction every two weeks.

The documentation measures Prediction Quality as lift quality, where 0% means no better than random guessing and 100% means perfect knowledge. It also states that all machine learning models make errors, so some users with a high score won't end up churning. Braze encourages you to decide how many churners to capture and what a false positive costs your business.

We suggest treating the score as a way to rank who gets attention, not as a verdict on a customer. Set the cost of a false positive per campaign: a discount spends margin, while a reminder spends very little. Teams that want help with churn prediction and reporting can read about CustomerIK's marketing analytics service.

What does a worked example look like?

Imagine a hypothetical subscription app whose team wants to cut cancellations. This is an illustration, not client work. It carries no figures, because real ones would have to come from the team's own data.

The team defines churn as a cancellation custom event and builds the prediction audience from current subscribers. It sets the primary conversion event to the behaviour it wants to protect, such as starting a session.

It writes down why it chose that window. A renewal becomes the exit event, so a subscriber who renews stops receiving the nudge.

At the next review, the control group in the retention report shows whether the nudge changed return behaviour. The team decides to keep it, change it or stop it, and records that decision beside the churn definition.

How does this connect to the rest of the programme?

Measurement sits at the centre, so start with our guide to retention metrics every marketer should track in Braze. On the automation side, see reducing churn with retention automation tactics.

Audience choices feed the same loop. Read using segmentation for smarter retention campaigns for targeting, and see how to trigger retention messages at the right time for timing.

When a customer has gone quiet, see how to build effective win-back campaigns in Braze. Before launch, check any plan against retention campaign mistakes to avoid in Braze Canvas. A CustomerIK case study on re-engaging lapsing retail users covers an interactive reward built into automated lifecycle journeys.

What order should the team set these up in?

  1. Write the churn definition in one sentence that names a behaviour and a time frame.
  2. Choose the retention event that matches that behaviour.
  3. Set the conversion events and the window, and record the reason for the window.
  4. Add the exit event so that returning customers stop receiving the journey.
  5. Keep a control group, so the review can read retention without a message.

As a Braze implementation partner, CustomerIK recommends keeping these five answers on one page. Review that page before any new retention campaign launches. It is short enough to read in a meeting, and it stops each campaign from redefining retention on its own.

Frequently Asked Questions

1. What is a retention event in Braze?

It is the event you specify in a retention report. The report shows the rate at which users who received the campaign or Canvas performed it over the 30 days from receipt.

2. Do I need to buy Predictive Churn to start?

Prior to purchase, the feature is available in preview mode. Full access means contacting your account manager. The preview does not let you target users for messaging according to churn risk.

3. How many conversion events can a campaign have?

A campaign can have up to four conversion events, with a window of up to 30 days. If you select none, the time defaults to three days.

4. Can a user convert without opening the message?

Yes, they can. The documentation says a user counts as converted if they performed the event in the window, even if they did not open or click the message.

5. Does this post settle questions about consent and data use?

It does not. Churn prediction works on user behaviour data, so check consent, data use and any cross-border questions with your own privacy team.

Sources

CustomerIK is a Braze implementation partner working across onboarding, technical integration, marketing operations and customer data management.

If you want help turning one churn definition into retention reports, conversion events and exit rules in Braze, let's talk.

Preferences

Privacy is important to us, so you have the option of disabling certain types of storage that may not be necessary for the basic functioning of the website. Blocking categories may impact your experience on the website.

Accept all cookies
Accept all cookies

These items are required to enable basic website functionality.

Always active

These items are used to deliver advertising that is more relevant to you and your interests.

These items allow the website to remember choices you make (such as your user name, language, or the region you are in) and provide enhanced, more personal features.

These items help the website operator understand how its website performs, how visitors interact with the site, and whether there may be technical issues.

Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.