71am
David Krug David Krug May 21, 2026

Analytics You'll Actually Look At

Most analytics setups measure everything and tell you nothing. A useful dashboard tracks a handful of numbers tied to money — and you can read it in ten seconds.

analyticsgrowthprocess

Open most companies’ analytics and you’ll find a wall of charts nobody checks. Pageviews, bounce rate, sessions by device, a real-time map with three dots on it. It’s a lot of measurement and almost no insight. The dashboard exists, but it never changes a decision.

A useful setup does the opposite. It tracks a short list of numbers that connect to revenue, and it’s legible in about ten seconds. If you can’t glance at it and know whether the week was good, it’s not working.

Measure the steps to money, not the noise

Pick the path a visitor takes to becoming a customer and instrument that. For most sites it’s a handful of events:

  • A visit to a money page (pricing, product, key landing page)
  • A meaningful action (started a form, booked a call, added to cart)
  • The conversion itself (signup, purchase, qualified lead)
  • The follow-through (activated, paid, retained)

Four numbers like these tell you where people fall out. Pageviews don’t. The goal is a funnel you can read, not a museum of metrics.

Tag events so they survive a redesign

The reason analytics rot is that events get wired to brittle things — a button’s text, a page URL that later changes. Six months on, half the tracking is silently broken and nobody noticed.

Anchor events to stable attributes instead:

<button data-event="book_call" data-source="pricing">
  Book a call
</button>

Then your tracking listens for data-event, not the label or position. The copy can change, the page can move, and the number keeps counting. Durable instrumentation is the difference between a dashboard you trust and one you relearn every quarter.

One question per chart

Every chart should answer a specific question you’d actually ask out loud:

  1. Are more of the right people arriving?
  2. Are they taking the first step once they’re here?
  3. Are first steps turning into customers?

If a chart doesn’t map to a question that would change what you do next, it’s decoration. Delete it. A dashboard with three honest charts beats one with thirty you ignore.

The point is the decision

Analytics isn’t a reporting exercise. It exists to make the next move obvious — which page to fix, which step is leaking, where the next executable should go. When the numbers are this clear, the work plans itself: find the biggest drop-off, ship the fix, watch the number move, repeat.

That loop — measure, fix, measure — is the whole game. Everything else on the dashboard is just something pretty to scroll past.