It is tempting to install an analytics SDK, switch on automatic tracking and assume the insights will follow. Usually they do not. You end up with thousands of screen views and button taps, and still cannot answer simple questions such as "why are people cancelling?" A small set of well-chosen events, defined before launch, will tell you far more.
Start with questions, not events
Before listing a single event, write down the decisions you need to make in the first few months after launch. For example:
- Are new users reaching the core feature?
- Which acquisition channel brings people who stay?
- Where do people abandon checkout or booking?
- Which features are used by the customers who renew?
Each event you track should help answer at least one of these. If an event answers nothing, leave it out for now.
Acquisition and install events
You need to know where users come from. Most stores and attribution tools capture installs automatically, but make sure you also record the first open and, if relevant, the campaign or referral that brought the person in. This lets you compare channels on quality, not just volume.
Activation events
Activation is the moment a new user gets real value. Define it specifically for your product, such as "created first project" or "completed first booking". Then track the steps leading to it:
- Account created
- Onboarding steps completed or skipped
- Key permission granted, if the core feature depends on one
- First core action completed
With these in place you can see exactly where new users stall.
Engagement and retention events
Raw session counts tell you little on their own. More useful are events tied to your core value: messages sent, items added, reports viewed, routes completed. Pair them with a retention view showing what share of each week's new users are still active later on. Watching retention by sign-up week shows whether product changes are actually helping.
Feature adoption
When you ship a new feature, add an event for its first use and for repeated use. One-off curiosity looks very different from a habit, and only the latter justifies further investment.
Revenue events
If the app makes money, track the full path, not just the final purchase:
- Viewed pricing or paywall
- Started checkout or trial
- Completed purchase or subscription
- Renewed, upgraded, downgraded or cancelled
- Refund requested
Store-side data from Apple and Google is useful, but your own events let you link revenue back to behaviour inside the app.
Friction and error events
Some of the most valuable events record things going wrong: failed payments, validation errors on forms, search queries with no results, timeouts and crashes. These point directly at fixes. Crash reporting tools handle crashes well, but logged business errors, like "address not recognised", often need custom events.
Name and document everything
Inconsistent names quietly ruin analytics. Agree a simple convention, such as object followed by action in past tense: "order_placed", "profile_updated", "search_performed". Add properties for detail rather than creating new events for every variation. Keep a shared tracking plan listing each event, its properties, when it fires and who owns it. Review it whenever features change.
Respect privacy
Collect only what you need, avoid sending personal details like names or emails as event properties, and make sure your app's privacy declarations and consent flows match what your analytics tools record. In the UK and EU in particular, consent rules for tracking deserve careful attention.
FAQ
Which analytics tool should we use?
The right tool depends on your budget, privacy needs and how much product analysis you plan to do. More important than the tool is having a clear tracking plan, since good events can usually be moved between platforms.
How many events should we start with?
Enough to cover acquisition, activation, core usage, revenue and errors. For most apps that is a short, manageable list rather than hundreds of events.
Can we add events after launch?
Yes, but you lose history for anything you did not track from the start. That is why activation and revenue events are worth defining before release.
Next step
Analytics only helps when it is designed alongside the product. Our mobile app and digital marketing teams can build a tracking plan that answers your real questions. Get in touch to start the conversation.