MMM template
Media mix modeling has a reputation as a data team project. The uncomfortable truth: the modeling is the easy part, the discipline around it is what fails. A usable MMM template is mostly about data preparation and honest interpretation. Here is what it should contain and where to start.
What a complete MMM template covers
- Data spec sheet: exactly which columns, granularity and history you need before modeling
- Spreadsheet MMM with adstock and saturation transforms, for smaller budgets
- Output reading guide: contribution charts, response curves, and what not to over-read
- Budget reallocation framework: turning model output into next quarter's budget
- When to graduate to Robyn or Meridian, and the notebooks to do it
Start free
The working templates
The MMM data spec sheet, the spreadsheet model with formulas, the output reading guide and the reallocation framework ship inside The Marketing Triangulation Playbook (Module 3: MMM Lite), with the weekly steps that use them. One-time purchase, lifetime updates.
See what's in the Playbook →Frequently asked questions
Can you really do media mix modeling in a spreadsheet?
For smaller budgets, yes, with honest caveats. A spreadsheet MMM with adstock and saturation transforms won't match a full Bayesian model, but it beats platform attribution for budget allocation and teaches you what the big models actually do. Above roughly 5-10 channels or with strong seasonality, move to Robyn or Meridian.
What data does an MMM template need?
Weekly (or daily) spend per channel, the outcome variable (revenue or conversions from your own records, not platform-attributed), and control variables: promotions, pricing changes, seasonality markers. Most teams need 18-24 months of history for stable estimates.
Are Robyn and Meridian free?
Yes, Meta's Robyn and Google's Meridian are open source. The models are free; the work is in preparing your data and reading the output honestly, which is exactly what a good template and spec sheet are for.