Quickstart for data teams
This quickstart is for data governance owners, data engineers, and implementation leads. Follow these steps to create a tracking plan, validate your event data, and organize your team's instrumentation work.
You'll set up
Before you start
An Amplitude account with access to Amplitude Data — go to Create a new account if you don't have one.
An Admin or Manager role in your organization (required to create and publish tracking plans).
Alignment with your product and engineering teams on what to track — go to Plan your implementation to work through those decisions first.
Starting from scratch? Go to Instrumentation prework to build alignment on taxonomy, user identity, and naming conventions before you create your tracking plan.
Create
Build a tracking plan
A tracking plan is a single source of truth that defines every event and property your product sends to Amplitude. It gives engineers a spec to implement against and gives analysts confidence in the data they query.
Go to Amplitude Data and open the Tracking Plan tab. Create a new plan and add your events — start with the five to ten actions most critical to your product's success, such as sign-up, activation, and conversion events. For each event, define its required properties and their expected types.
Publish the plan so it's visible to your engineering and product teams.
Validate
Check incoming events
Once engineering instruments against your plan, use event validation to check that real events match your spec. Amplitude Data flags events with missing required properties, unexpected property types, or names that don't match your plan.
Open the Observe tab in Amplitude Data to see a live stream of incoming events alongside any validation warnings. Resolve issues by updating the tracking plan, fixing the instrumentation, or both — the resolution workflow keeps a record of what changed and why.
Organize
Set your team's workflow
A clean tracking plan only stays clean if everyone follows the same process. Go to Implementation team organization for a recommended workflow: who owns the tracking plan, how engineers request changes, and how to handle schema evolution over time.
Define your team's review process before implementation starts so that new events go through the tracking plan before they hit production.
Next steps
Was this helpful?