Data Taxonomy

A hierarchical classification system that organizes analytics data into logical categories, defining how events, properties, and metrics relate to each other and to business concepts.

Also known as: data classification, analytics taxonomy

Why It Matters

A data taxonomy is the organizational structure that makes your analytics data navigable and understandable. Without it, you end up with hundreds of events and properties that only the person who created them can interpret. With it, anyone on the team can find and understand the data they need.

Taxonomies prevent naming chaos. When five developers independently track user engagement, you might end up with "user_engaged," "engagement_event," "feature_used," "active_usage," and "user_activity." A taxonomy establishes that all feature usage events follow the pattern "{feature_name}_used" and belong to the "product_engagement" category.

Good taxonomies also make analytics tools more useful. When events are logically organized, dashboards are easier to build, reports are easier to navigate, and automated insights can group related metrics meaningfully.

Industry Applications

E-commerce

A home goods retailer organizes their event taxonomy into five categories: Discovery (search, browse, filter), Evaluation (product view, comparison, review read), Purchase (cart, checkout, payment), Post-Purchase (delivery tracking, review write, return), and Account (login, profile update, preference set). This structure makes it easy for any analyst to find relevant events.

SaaS

A project management tool structures their taxonomy around the AARRR framework: Acquisition (signup, invite accepted), Activation (first project created, first task completed), Retention (daily login, weekly active), Revenue (upgrade, expansion), Referral (invite sent, referral converted). Each category has standardized naming and required properties.

How to Track in KISSmetrics

Build your data taxonomy before implementing tracking in KISSmetrics. Organize events into logical categories (acquisition, activation, engagement, retention, revenue) and establish naming conventions for each category. Document the taxonomy in a shared location and reference it during tracking implementation reviews.

Common Mistakes

  • -Creating a taxonomy that mirrors your org chart instead of your customer experience
  • -Making the taxonomy too deep (5+ levels) which becomes cumbersome to navigate and maintain
  • -Not enforcing the taxonomy through code reviews and validation, allowing organic drift over time
  • -Building the taxonomy in isolation without input from the teams who will use the data

Pro Tips

  • +Align your top-level taxonomy categories with your business model: acquisition, activation, engagement, revenue, referral
  • +Use consistent prefixes or namespaces for related events: "checkout_started," "checkout_step_completed," "checkout_completed"
  • +Maintain a visual diagram of your taxonomy that shows how categories relate to each other
  • +Review your taxonomy quarterly and prune unused categories to prevent bloat
  • +Include your taxonomy in your engineering onboarding materials so new developers track events consistently from their first contribution

Related Terms

See Data Taxonomy in action

KISSmetrics tracks every user across sessions and devices so you can measure what matters. Start free - no credit card required.