Transcript
Even before user-level attribution was broken beyond repair, Multi-Touch Attribution (MTA) has always been incapable of optimizing beyond low-funnel performance marketing. Media Mix Modeling (MMM) has promised to solve this problem for over a decade. It has not. Don't be fooled by shiny product marketing making false promises.
MMM uses mathematical techniques like ad stocks and multi-level models to try to force marketing’s impact from two weeks, to four, to six weeks, or more. This approach is more art than science. It's based on a presupposed decay shape and length for different parts of your media mix.
The reality is that mixed models are doing gymnastics to produce the effect your experts want to see. TV has a six week impact because it's given a six week lag effect on the model. Pinterest has five weeks, Meta gets four weeks.
Data is not driving the outcome; it's finding the effect the model wants it to find based on how the model builder expects media to operate in the marketing funnel. Beware of any media model that tells you it has quantified the long-term impact of your full-funnel marketing strategy.
MMM on its own cannot simply, reliably, and accurately inform the long-term impact of marketing. However, when MMM is combined with incrementality testing, marketers have a credible way to ensure your brand is capturing the full impact of your entire media mix. This is based on causality. Ground truth experimentation. Incrementality gives you the ability to train mix models, tune ad stocks, and quantify the impact of marketing across multi-level models.
To find out more about our unique triangulated media measurement offering in the Measured platform, schedule an expert demo today.