Marketing Mix Modeling, Done Properly
A statistical method that estimates how much each marketing channel contributes to sales using aggregated historical data rather than individual tracking.
Samir Haddad
The key figures
- Data
- aggregated spend, sales and external factors over time
- Privacy
- does not require user-level tracking
- Open-source tools
- Google's Meridian and Meta's Robyn
- Limitations
- needs enough historical variation in spend
Why this is worth getting right
As tracking becomes less complete, mix modeling offers a privacy-friendly way to guide budget allocation across channels.
Do this, not that
Do
- Collect clean weekly spend and sales data
- Include seasonality and external factors
- Validate models with experiments
- Refresh models regularly
- Use results to guide, not dictate, budgets
Don’t
- Too little history for reliable estimates
- Ignoring promotions and pricing changes
- Treating model outputs as certain
- Changing all budgets at once based on one model
When to bring in help
Our advice Bring in help when managing significant multi-channel budgets, or when attribution data has become unreliable.
Where this comes from
- Google for Developers — Meridian marketing mix model
- Meta Business Help Center — About marketing mix modeling
The figures and practices above come from the sources listed.
Working on something like this?
We take on Digital Marketing & CRO work for teams who want it done once, properly. Tell us what you are building and we will tell you honestly whether we are the right studio for it. Start a project.
Where to go next
Spotted something wrong? Report an error on this page. We correct on the page and say what changed.
