Media mix modelling (MMM) uses historical data and statistical regression to estimate how much each marketing channel contributed to business outcomes such as sales or revenue. Unlike attribution models which work at the individual user level, MMM operates at the aggregate level — using time series of spend and outcomes by channel to separate each channel contribution. MMM naturally handles non-clickable channels like TV and CTV because it does not require individual-level identity linkage — it observes that when CTV spend went up, sales tended to go up by a certain amount, all else equal.
MMM is increasingly relevant for India CTV because it sidesteps the identity and deduplication problems that plague individual-level attribution. For brands with 2+ years of marketing spend data and robust sales tracking, MMM can quantify CTV ROI alongside linear TV, digital, and other channels in a single model. Limitations: requires aggregated data at sufficient granularity (weekly spend and outcome data minimum), takes 8-12 weeks to build, and cannot provide real-time optimisation signals. Bayesian MMM approaches (Meta Robyn, Google LightweightMMM) are accessible to mid-sized advertisers. For India CTV, MMM is the most methodologically honest way to answer what CTV is contributing to the business when identity-based attribution cannot do so reliably.
Full guide
For a complete explanation, read: Media mix modelling for CTV: how MMM quantifies contribution to business outcomes