Event tracking schemas: the decision that decides what questions you can ever answer

An event is a record of something a user did: "viewed a product," "added to cart." An event schema is the consistent shape every one of those records gets logged into. Getting this shape right, early, determines what questions your data can ever answer later.

🎯 Explain Like I'm Hired An event schema is a standard form every "thing a user did" gets filled out on. Every form has the same required fields (who, when, what happened) plus a flexible section for details specific to that type of event. Example: if "add to cart" logs the product under a field called item, but "checkout" logs the same concept as product_ids (a list, differently named), you can't cleanly connect the two events without writing custom one-off logic, and multiplied across hundreds of event types, that's how "just pull the numbers" quietly becomes a two-week project.

CREATE TABLE events (
  event_id      uuid PRIMARY KEY,
  user_id       uuid NOT NULL,
  session_id    uuid NOT NULL,
  event_name    text NOT NULL,      -- 'product_viewed', not 'ProductViewed' or 'view'
  event_time    timestamptz NOT NULL,
  properties    jsonb,              -- details specific to this event type
  context       jsonb               -- device, app version, locale (same on every event)
);