The handful of events actually worth tracking

Dear brand owner,

I once inherited an account with sixty-three custom events. Nobody could tell me what half of them meant, and not one of them had ever changed a decision. That's the trap with event tracking: it feels productive to add more, but each one you can't act on is just noise wearing a badge.

Start from the questions

An event earns its place when it answers something you'd actually ask. Work backwards from the decision, not forwards from the button.

• Did they add to cart? Tells you if the product page did its job
• Did they begin checkout? Marks the line between browsing and intent
• Did they add a discount or fail one? Hints at price sensitivity
• Did site search return nothing? Shows demand you're not meeting
• Did they start but abandon a form? Points at friction you can fix

Name them like a human

Future you has to read these. A tidy naming scheme is a gift you give yourself later.

1. Pick a consistent verb-noun style, like add_to_cart or begin_checkout
2. Keep the same casing everywhere, no exceptions
3. Write down what each event means the day you create it

An untracked event you'll never analyse costs you nothing. A tracked one you can't interpret costs you every time you open the report.

Before you build the next event, ask what you'd do differently depending on the answer. If nothing changes, don't track it. If you want a second pair of eyes on which events are actually earning their keep, book advisory time and we'll trim the list together.