Definition
Media mix modelling (also marketing mix modelling, MMM) is a top-down statistical method that estimates how each marketing channel, plus non-marketing factors, contributed to an outcome like registrations or revenue over time. It regresses the outcome against historical spend and activity by channel while controlling for seasonality, promotions, sporting-event calendars, price, competitor activity and macro trends, and it models diminishing returns and lagged or carryover effects for each channel.
Because MMM works on aggregated time-series data, it needs no user-level tracking, cookies or device IDs, which makes it resilient to privacy changes that have degraded click-based attribution. Its trade-offs are the opposite of attribution's: it captures hard-to-track and offline channels and gives a strategic view of channel efficiency and saturation, but it is coarse, needs a long and varied history to be reliable, cannot guide day-to-day optimisation, and can confuse correlation with cause if a channel's spend always moved with something else.
In context
Large iGaming operators use MMM to answer strategic questions that user-level attribution cannot: what is the true combined contribution of affiliates as a channel versus paid social versus TV or sponsorship; where is each channel on its diminishing-returns curve; how much do sporting-event calendars and promotions drive the baseline; and what would happen to registrations if a channel's budget were cut or doubled. It is particularly useful for valuing brand and sponsorship activity and for cross-checking whether the sum of channel-level attribution claims is even plausible against total results.
MMM and the bottom-up methods are complementary. A common setup is a triangulation: MMM for the top-down strategic allocation across channels, multi-touch attribution for tactical optimisation within trackable digital channels, and incrementality experiments to calibrate both and settle disputes with causal evidence.
For affiliate strategy, MMM can show whether the affiliate channel as a whole is still delivering incremental value at its current spend level or has saturated, which informs programme-wide decisions about commission levels, tier expansion and how hard to recruit - decisions that per-affiliate last-click reports cannot support.
Worked example
An operator's MMM shows the affiliate channel is near saturation at current spend - additional commission outlay would buy few extra incremental registrations - while paid social is well below its diminishing-returns point. Combined with incrementality tests, this leads the operator to hold affiliate budget flat, tighten low-incrementality partner deals, and grow paid social instead.
Related terms
Frequently asked questions
Browse more iGaming terms in our glossary.
← Back to glossary