Contributing to the field of Marketing Science.

i4i is one of the world’s foremost practitioners of Bayesian statistics.

Most analytics partners use modeling approaches that leverage regression. What makes i4i different is our expertise applying the principles of Hierarchical Bayesian (HB) modeling that gets to a level of granularity in products, outlets, and media channels that can support strategies and enable tactics.

What exactly is Hierarchical Bayes?

For those unfamiliar, the technical explanation is that Hierarchical Bayes is a statistical model written in multiple levels (hierarchical form) that estimates the parameters of the posterior distribution using the Bayesian method. Hierarchical modeling is used when information is available on the observational units (such as product, outlet, and geography), where there is a need or desire to pool data, where data quality varies a lot across units, and where different observational units may have different predictors.

More concretely, what does it give you?

Hierarchical Bayes based analytics enable the right decisions based on more accurate and actionable media KPIs and forecasts.

      • It goes beyond high-level averages to shed light on how media works at the level executed and provides precise indicators of the odds your individual KPIs are correct
      • It systematically leverages both previous experience and current data to ground results in business reality – much less reliant on subjective analyst judgment
      • It provides highly granular results that can be rolled up & reported at whichever levels are most applicable to media decisions

Hierarchical Bayes provides increased returns on your marketing investments

      • It makes better use of all available inputs – sharing across all products, markets, and media to make good predictions even when data are sparse or when tactics are spotty & infrequent
      • It can apply category, brand, market, campaign, and target audience attributes to explain differences in critical media KPIs

Hierarchical Bayes enables a wider range of marketing effectiveness understanding and insight

      • It’s able to capture effects of many more media and other sales drivers than traditional approaches – even in cases with limited or collinear data
      • It can handle portfolio level MGA – providing stable, actionable media KPIs across all brands and markets in one pass, not one at a time

 

As a client faced with an increasingly complex array of decisions to make across media channels and consumer touchpoints, hierarchical bayesian analytics delivered by i4i is the right choice … now.

To learn more, contact Stuart Schwartz sschwartz@in4ins.com.