Share of Search + Difference in Difference (DID) = Faster Validation of MMM
Share of search can be a powerful metric to validate MMM
At Aryma Labs, we rely on a lot of causal as well as quasi casual experiments to prove the efficacy of Marketing Mix Models (MMM).
In our whitepaper titled "proving efficacy of MMM through Difference in Difference method", we had showcased how we could prove the efficacy of MMM. Link for the same is in resources section.
For quite some time we were also researching on Share of search (SOS). Initially we were researching on SOS as we thought it could be a valid proxy for Brand Equity measures.
SOS more or less translates to brand's mental availability. Les Binet has done some great work with regards to Share of Search (links to his papers in resources section).
Because brand's mental availability generally precedes sales, we thought that Share of search could provide quicker feed back on the efficacy of changes made as a result of MMM.
This of course comes with a lot of caveats - such as:
1. Making sure that the changes made as a result of MMM (such as increasing spends on YouTube ads, TV ads, or Instagram ads)
are the only changes being affected on the brand. Basically we have to control for all factors affecting the KPI (e.g. Sales).
2. Other factors like price sensitive nature of the industry where the sales could be heavily driven by price points rather than the media activity by the brand.
If we take care of the caveats, then, SOS can provide quicker feedback on MMM's efficacy.
We just currently wrapped up this experiment and will publish a paper on this soon.
But through our experiment we found out the following:
🎯 Share of Search can be used as an additional validation for proving MMMs efficacy.
🎯 Share of search can provide feedback at a faster cadence in comparison to sales numbers and market share numbers.
🎯 Because Share of Search precedes sales (most often than not), DID which tests for incremental lift in sales (later in the time) would also be indirectly validating the delta observed in SOS.
🎯 Thus DID not only would be causally proving the incremental lift, it could also technically prove the incrementality in SOS.
📌 In summary:
We believe Share of Search along with DID can lead to faster and more reliable validation of MMM.
Resources:
Les Binet’s 's papers on Share of Search
https://ipa.co.uk/news/binet-presents-fast-cheap-predictive-share-of-search-metric/
Kantar's article : https://www.kantar.com/uki/inspiration/advertising-media/share-of-search-your-moment-has-arrived
Les Binet's video :
DID whitepaper: https://www.arymalabs.com/proving-efficacy-of-mmm-through-difference-in-difference-did/
Thanks for reading.
For consulting and help with MMM implementation, Click here
Stay tuned for more articles on MMM.