How to write a metric tree for an ecommerce ad account

A metric tree starts from one number the client cares about and breaks it into the parts that move it. It shows where to look when that number changes.

Measurement and reporting · 5 min read · Updated 2026-10-07

Start at the top

  • Pick one top metric the client already signs off on. Often this is contribution margin or MER against a target, not raw revenue.
  • Write its formula in one line, with the units. A metric without a formula starts a debate instead of a review.

Split it into drivers

  • Each branch should be a part you can measure or change, such as order value, margin rate, discount depth, shipping cost and media cost per order.
  • Check that the branches add up to the top. Multiplicative trees need the same check at every level.
  • Stop at the level where a person can act. A branch nobody can change is context, not a driver.

Attach an owner and a check

  • For each branch, write the owner, the source, the definition from the definitions sheet and how often it is read.
  • Mark each branch as measured, estimated or not measured yet. Never show an unmeasured branch as zero.

Use the tree when a number moves

  • When the top metric moves, read down one level at a time. Stop at the branch that explains most of the change.
  • Record the branch and the finding in the change log, so the next review starts from the same place.
Made-up scenario: a made-up store's contribution per order fell. The tree showed order value unchanged and a deeper discount on one product group. The agency did not change the ads yet. It raised the discount question with the client first.

A tree on one page is more useful than a tree in a long deck. Keep it to one screen and review it once a month.

Built for shop owners, creators and agencies. Stock photos, not customers.

Run the next client review from one screen.

We open one agency at a time, with your own accounts connected read-only first.

Request access

How we plan to price it