Marketing Mix Modeling is a statistical analysis technique used by marketers to understand the individual and combined contributions that multi-media marketing investments have on business results.  Thus, marketing mix models help estimate the ROI associated with historical marketing spending, as well as forecast the prospective business results that future spending will generate.  These models are particularly applicable where direct attribution of sales to marketing stimuli is weak or where consumer response is fairly long and complex. 

Marketing mix models use time series data in a structural modeling approach to determine all the relevant independent marketing and non-marketing variables that affect the dependent variable being examined (such as sales).  Heavily dependent on the amount and granularity of the underlying data, marketing mix models try to account for advertising carryover and wear out, diminishing returns, and interaction effects.

In most instances, marketing mix models are developed to analyze what has been achieved from past marketing spending.  Acting as a microscope, the models look retrospectively at past performance and try to explain as much as they can about why certain results happened.  However, because markets change - along with advertising effectiveness - it is advisable to update these models on a regular basis so that their coefficients (the impact factors associated with each independent variable) most closely represent recent experience. 

Not to be overlooked, marketing mix models can also play an important role in marketing planning.  Figuratively turned around and acting as a telescope, these models can be used prospectively to predict what is likely to occur given planned spending and assumptions about future market conditions.  In fact, to some the best thing about marketing mix models is their scenario simulation capability.

For more information, please refer the paper “Quantifying the Effects of Advertising" or read our Marketing Mix Modeling Case Studies.