how we do it
Unlike many other forms of analysis technique, our approach allows all forms of marketing to be evaluated in a single analysis framework, regardless of type providing consistency of measurement across the wider marketing mix.
Based on a specific branch of multiple regression analysis, our approach does so by controlling for movements in a dependent variable (such as EPOS sales) via the inclusion of a series of “explanatory” data sets that allow us to effectively control for non-marketing related factors that may be impacting the sales of a brand.
“A statistical process that uses various data sources to assess the level of mutual correlation between a dependent (KPI) variable and multiple explanatory variables.”
In doing so not only do we remove the chance that sales uplifts caused by another factor are wrongly attributed to the influence of marketing activity but also provide an additional means to explain historical performance by way of the influence of non-marketing factors on the performance of marketing.
Depending on the dimensions of the data used, we can go down to quite deep levels of insight, with analysis taking place at the individual intervention or channel level by store by SKU by day.
Example equation:
