A measurement plan should be agreed before the work starts, not assembled from whatever the dashboard shows afterwards. The version that works is short: one business goal, one number that proves marketing moved it, a target someone can say no to, and the decision you'll make if it isn't working.
Below are the eleven sections, what each one is for, the mistake I see most, and a line from a filled-in plan for a made-up coffee subscription launch. You can see the whole plan here, or start your own. It's free and there's no sign-up.
The number the business actually cares about. Revenue, customers, retention, share. Not a marketing metric.
The usual mistake is writing a marketing metric here, like awareness or engagement. Write the number the business reports on.
Push harder: Would the CFO recognize this as a goal?
The behavior or belief that moves the goal, written as from, to.
"Drive consideration" isn't a change. From what, to what is. It also tells you what to count.
Push harder: What exactly do they do today, and what should they do instead?
The specific people. The rest of the plan only counts what they do.
If you can't find these people in your data, you can't measure what they did. Define them the way the report does.
Push harder: Who is left out on purpose?
One primary KPI, and exactly how it's counted.
Three KPIs means no KPI. Pick one, then write down exactly how it's counted so nobody argues in month two.
Push harder: Is it one number?
The current number, the date range, and where it came from.
Most plans skip this, and then every result looks good. Write the number, the dates, and the report it came from.
Push harder: Over what dates?
From what, to what, by when. And how you picked it.
A target without a reason gets negotiated down at the first bad week. Say how you picked it.
Push harder: Why this number and not half or double?
Two or three things you'd see in the first two weeks if it's working.
The KPI usually moves too slowly to steer by. Early signals are what you'll actually watch in the first two weeks.
Push harder: What would you see in week two that tells you to keep going?
Numbers that must not get worse while you chase the target.
Targets get hit in ways nobody wanted: discounts that wreck margin, sign-ups that cancel. Name the numbers that must hold.
Push harder: What could you break while hitting the target?
The tool or report for each number, and who pulls it.
When two tools disagree, the meeting becomes about the tools. Decide the source for each number now, and who pulls it.
Push harder: Who pulls each number, by name?
The meeting or report, the owner, and the rhythm.
Numbers nobody looks at stop being collected. Put them in a meeting that already exists, with an owner.
Push harder: Which meeting does this show up in?
The decision you agree on now, before the numbers come in.
This is the section that makes it a plan and not a report. Agree now what you'll change if the number is short by a date.
Push harder: What number, by what date, changes the plan?
Answer the eleven sections, run the check for the usual problems, and send it as a link. People can comment on the part they mean and approve it with their name.
Start a measurement plan See the example
Also on one page: the creative brief and the messaging framework.
From the same desk
A weekly note on account planning for agency strategists. One thing that changed, the tension in it, and what I'd do Monday. The first part of every issue is free.
Written by Cody Simmonds. I also write The Rationale, a weekly note for agency strategists.