
To measure customer engagement in Braze, match each question to the report built to answer it: campaign and Canvas analytics for each message, funnel reports for drop-off, retention reports for return visits, and Report Builder for comparisons. Choose the engagement event before launch. Every report is only as useful as the event it counts.
Key Takeaways
- Match each engagement question to the report built for it: campaign and Canvas analytics, funnel, retention, Report Builder.
- Canvas analytics assigns a conversion to the latest message the user received, so step counts record timing, not cause.
- Funnel reports filter users by entry date and enforce event order. Canvas analytics filters events by when they occurred.
- Rolling retention can exceed 100%, because a user who repeats the event is counted in several time frames.
- Funnel and retention reports aren't available for API campaigns, so plan another route to measure those sends.
- Choose one engagement event before launch and use it for the conversion goal, the funnel and the retention check.
How do you measure customer engagement in Braze?
Braze groups its measurement tools by the depth of the question. Its summary: use "dashboards for at-a-glance monitoring, reports for deeper analysis across campaigns and Canvases, and tracking settings to configure how Braze captures user actions."
The reports section carries a comparison table to help you pick. We have condensed it below and added the question we tend to ask first for each report.
| Report | What the documentation says it does | The question we ask first |
|---|---|---|
| Campaign analytics | Real-time results for each campaign, broken down by messaging channel | Did this message get seen and acted on? |
| Canvas analytics | Top-line stats, variant performance and step-level metrics for a Canvas | Which step or variant carried the result? |
| Funnel reports | The journeys customers take after receiving a campaign or Canvas, and where drop-offs occur | Where did people stop? |
| Retention reports | The rate at which recipients perform a retention event over the 30 days after receiving the message | Did anyone come back, and for how long? |
| Report Builder | Results across multiple campaigns or Canvases in one view, with customisable metrics and drilldowns | How do these messages compare? |
Nothing in that table defines engagement for you. That is a product decision.
What do campaign and Canvas analytics tell you first?
Start here, because these pages show what happened before any deeper report takes over. Canvas analytics adds step-level detail and a breakdown by variant, including conversion totals, conversion rates, uplift against the control variant and statistical confidence for each conversion event.
The attribution rules deserve a careful read. A user can only convert once per conversion event per Canvas entry, and conversions are assigned to the latest message that user received for that entry.
Braze gives a clear example. Take one user who opens the app before the first message and another who opens it after every push.
The Canvas summary shows two conversions while the individual steps show one conversion on the first step and none after that. Treat step counts as a record of the latest message at the time, which is a weaker claim than cause.
Revenue follows its own rule. Purchases count toward Total Revenue even when the user hasn't performed the configured primary conversion event, provided the purchase falls within the attribution rules for the window.
Why do funnel reports and Canvas analytics show different counts?
They answer different questions, and the documentation says so plainly. Canvas step analytics can show higher counts than the funnel for the same calendar dates. Step analytics include broader engagement and conversions, while the funnel enforces event order and timing rules.
Two mechanics explain the gap. The date range in Canvas analytics filters events by when they occurred, while the date range in a funnel report filters users by when they entered the Canvas. Take a user who enters on the first of the month and converts on the eighth.
One view shows one entry and zero conversions. The other shows one entry and one conversion.
The second mechanic is order. Funnel reports require events to occur in the specified order and count each user only once. A user leaves the funnel at the first step where they don't perform the next event in the sequence you configured.
When two dashboards disagree in a review meeting, we suggest writing down both date-range rules before anyone debates the numbers. Such arguments often come down to one filter applied to events and the other to entrants.
What should you know before you build a funnel report?
Four documented details shape what a funnel can tell you. Funnel reports aren't available for API campaigns, and the follow-up window runs for a maximum of 30 days. If you set the window to one day, the event must occur within 24 hours of message receipt.
Identity matters as well. Funnel reports track follow-up events that match the user ID at the time of entry. An anonymous user who later becomes identified gets a new Braze ID, so conversions after that point won't appear in the funnel.
If your Canvas audience includes anonymous users who sign up mid-journey, expect the funnel to understate what happened after sign-up. Say so in the report notes.
How do you read retention reports without fooling yourself?
Retention reports measure the rate at which a recipient performs a retention event over the 30 days after receiving the campaign or Canvas. You choose the event, and you can change it and re-run the report to see the effect on different outcomes.
There are two views. Rolling retention counts users who do the event on or after each day listed across the top. Range retention counts users who come back within each range of days.
Here is the catch with rolling retention. A user who completes the event several times can be counted in several time frames. In Braze's own words, that "could result in the retention rate exceeding 100%".
It isn't an error. It's how the measure works, so label it before it reaches a slide.
Two practical limits apply. Retention reports aren't available for API-triggered campaigns. Once a report is generated, it can't be re-run with the same retention event for 24 hours.
The documentation suggests checking day-of-week trends for recurring campaigns and asking where the impact starts to decline. It also points to the control group's retention data, which shows what retention looks like without a message.
How does Braze show whether a variant really won?
Both the Canvas analytics variant table and the retention report by variant report a confidence figure. For retention, Braze compares each variant's conversion rate against the control's with a statistical procedure called a Z Test to calculate a confidence percentage.
We would read that figure as a prompt to keep looking. A variant that wins on the first conversion event can still lose on the retention event a few weeks later. Check both views before you roll the winner out.
When does Report Builder earn its place?
Use Report Builder when the question spans messages. You choose what the rows represent (campaigns, Canvases, channels, tags, segments), then add drilldowns and the metrics you want. The Rows selection impacts the metrics you can view, so settle it first.
Two features suit recurring reviews. You can add campaigns automatically by setting rules for which campaigns or Canvases to include.
The report picks up new matches on future runs. You can also add a finished report to a dashboard.
If your campaigns carry a consistent tagging scheme, one automatic rule gives you a standing comparison without a monthly rebuild.
Three behaviours are worth knowing. Report download links expire after one hour.
Report Builder reports don't support team assignment the way campaigns do. In the same report, Unique Recipients can be higher than Unique Impressions when a message variant was deleted after send.
The release notes for 28 May 2026 added Banner as a channel and RCS as a sub-category under SMS. Both can now be measured in custom reports alongside every other Braze channel.
What do engagement reports and custom events add?
Engagement reports send statistics for chosen campaigns and Canvases to your inbox at a time you choose. They aggregate data per campaign or Canvas, not at the workspace level, and they aren't saved in the Braze dashboard. Use them to feed a recurring review, and use Report Builder for workspace-wide questions.
The custom events report is the closest thing to a direct engagement trend. You can break results down by segment, and a per-active-user option normalises event volume against your monthly active user count. That normalisation stops a growing audience from looking like growing engagement.
One integration detail is easy to miss. Custom events sent through the users track endpoint without an app identifier are recorded. They won't appear in the graph when an app filter is applied.
If you want these reports turned into executive-level dashboards, our marketing analytics service covers that work.
How should you choose the engagement event?
Every report above counts an event you chose, so the choice decides the answer. We recommend picking one event that represents a customer getting real value.
Write it down before launch. Then use the same event for the conversion goal, the funnel's final step and the retention check.
Opens and clicks are useful diagnostics. They rarely tell you whether the customer did the thing your product exists for, so let one product event carry the verdict.
The failure we see in planning reviews is a different event in each report. The Canvas converts on a session start, the funnel ends on a purchase, and retention looks for a login. Three reports then answer three questions, and nobody can say whether the programme worked.
What does this look like in a worked example?
Take a hypothetical marketplace app that sends a three-step re-engagement Canvas to people who haven't browsed recently. It is an illustration with no client data and no real results.
The team picks one engagement event: a saved search that the customer returns to. They set it as the Canvas conversion goal and as the retention event, and they make it the last step in the funnel.
After launch, Canvas analytics shows the second step carrying the largest share of conversions. The funnel shows the drop-off sits between the message and the first return visit, which points at the message itself. Retention by variant shows both variants ahead of the control group in week one.
The team checks week four before choosing a winner, because an early leader may fade. They also add a caveat that anonymous visitors who sign up mid-journey fall out of the funnel.
The same event runs through every view, so the reports agree about what engagement means even when the counts differ.
How does this connect to the rest of the programme?
Measurement works best when it sits beside the work that feeds it. Our guide to advanced reporting strategies covers how to structure reviews once the basics are in place, and what to track in email reporting goes deeper on one channel.
If you work mostly in the app, see how to measure in-app message engagement. For warehouse and BI needs, read about connecting Braze to external analytics platforms.
The reports tell you whether a change worked, but the change has to come from somewhere. Our posts on building a customer engagement strategy and A/B testing email campaigns cover that side.
As a Braze implementation partner, CustomerIK recommends fixing the engagement event and the reporting calendar before the first send. A report built after launch tends to measure whatever was easiest to count.
Frequently Asked Questions
1. Why don't my funnel and Canvas analytics numbers match?
Canvas analytics filters events by when they occurred, while a funnel filters users by when they entered the Canvas. Funnels also enforce event order and count each user only once.
2. Can retention rate exceed 100%?
In rolling retention it can, because a user who completes the event several times is counted in several time frames.
3. Can I run funnel or retention reports for API campaigns?
The documentation says both report types aren't available for API campaigns, so plan another route to measure those sends.
4. What is the difference between Report Builder and engagement reports?
Report Builder compares results across campaigns and Canvases in one view, including workspace-wide questions. Engagement reports aggregate per campaign or Canvas and arrive by scheduled email.
5. Does this post settle how to define engagement?
No. The Braze pages document the reports and give no definition, so the choice of event is yours.
Sources
- Braze, Analytics. Retrieved 10 Nov 2026.
- Braze, Reports. Retrieved 10 Nov 2026.
- Braze, Canvas analytics. Retrieved 10 Nov 2026.
- Braze, Funnel reports. Retrieved 10 Nov 2026.
- Braze, Retention reports. Retrieved 10 Nov 2026.
- Braze, Report Builder. Retrieved 10 Nov 2026.
- Braze, Engagement reports. Retrieved 10 Nov 2026.
- Braze, Custom events report. Retrieved 10 Nov 2026.
- Braze, Release notes, 28 May 2026. Retrieved 10 Nov 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 your reports disagree about what engagement means, let's talk.




