MMM data readiness checklist

What's inside

• A 20-point checklist across history, spend, outcomes, controls, structure and evidence, with a readiness score and a verdict
• A spend variation check: paste up to two years of weekly spend for up to eight channels and see which vary enough to measure
• A channel correlation check that highlights channels moving together too closely to separate
• A data dictionary listing the columns a modelling dataset needs
• Illustrative example data, so every check shows how it works before you add your own

Who it's for

Marketing and analytics teams considering a mix model, and anyone briefing an agency or vendor to build one.

How to use it

1. Answer the checklist honestly; “Not sure” counts as a gap.
2. Paste your weekly spend over the example data.
3. Fix anything marked “No” or “Barely varies” before modelling, or accept wider ranges for those channels.

Related reading

• The weekly dataset nobody wants to fund
• If your spend never changes, no model can measure it
• Channels that move together can't be told apart

Free to use within your organisation. Please don't resell or republish it. Download terms