Marketing mix model
Also called MMM, media mix model, marketing mix modeling.
A statistical model that estimates each channel's effect on sales from aggregate spend and sales over time.
A marketing mix model (MMM) looks at totals over time: spend per channel per week, sales per week, and outside factors such as season, price and promotions. It fits a model that explains sales from those inputs, then reads each channel's contribution and its diminishing returns.
It needs no cookies, click ids or user-level data, which makes it durable as tracking gets harder. Two open-source models are widely used: Google's Meridian (Google) and Meta's Robyn (Meta).
The limits: it needs a long history of weekly data with real variation in spend, and it answers budget questions per channel, not per ad. Its estimates are only as good as its assumptions, which is why both projects support calibrating the model with experiments such as lift tests. For per-campaign decisions, multi-touch attribution and click data still do the daily work.
All 58 terms ยท Spotted a mistake? Email hello@sporetag.com.