Skip to main content
Next Step PM
Back to the library

Build · Core

Plan your analytics before you build

Decide what data you need to measure success before the code is written. Adding tracking afterwards is slower, and you lose the baseline you need to prove the change worked.

About 1-2 daysEvergreenLast reviewed 5 November 2026

Who this applies to

Skills it builds
Data and measurementTechnical and AI fluency
PM types
All
Product types
All
Starting points
All
Ways of working
All
Experience
All
Size of organisation
All

Steps, by experience

Beginner (under 1 year)

  1. Start from your success metrics and guardrails. For each, ask: what user actions do we need to record to calculate it?
  2. List those actions as events with clear names, such as 'Order Completed', and the details each needs, such as order value.
  3. Check what is already tracked so you do not duplicate it.
  4. Check with your privacy or legal colleague that you are not collecting personal data you do not need.
  5. Add the tracking work to the user stories so it is built and tested with the feature.
  6. After release, check the events arrive correctly before relying on them.

Intermediate (1 to 4 years)

  1. Map each metric to the events and properties needed.
  2. Write a tracking plan with naming conventions and owners.
  3. Review for privacy and add tracking to story acceptance criteria.
  4. Validate the data in a test environment and after release.

Experienced (5+ years)

  1. Maintain one shared tracking plan so teams use consistent definitions.
  2. Design events around user intent so they stay useful as the interface changes.
  3. Build data quality checks so broken tracking is caught within a day.
  4. Plan the experiment and attribution needs up front.

Who can do this

Owner: Product manager

Can be delegated to

  • Data Analyst or Performance Analyst: Write the tracking plan and check the data after release
  • Frontend Engineer or Full-stack Engineer: Add the events and check they fire correctly
  • Policy / Legal / Compliance: Check consent and data minimisation

Reviewed by: Tech Lead, Policy / Legal / Compliance

Tool examples

Google Analytics 4
Plan custom events and parameters, and check them in DebugView before release.
Amplitude
Use the data planning features to define events so unplanned ones are flagged.
Mixpanel
Use Lexicon to document each event so everyone reads the data the same way.
Jira
Add 'events fire as described in the tracking plan' to the acceptance criteria.

Definition of done

  • Every success metric mapped to events
  • Tracking plan reviewed for privacy
  • Tracking included in stories and checked after release

Common pitfalls

  • Tracking everything and understanding nothing
  • Vague event names that mean different things to different people
  • Collecting personal data you do not need

If you lead the team

What good looks like: Tracking is planned before building, so success can be measured from day one.

Ask your PM

  • Will we be able to measure our success metrics on launch day?
  • Who checked the tracking for privacy?

Warning signs

  • Tracking is added after launch.
  • Events are inconsistent across teams.

Do these first

Sources