Problem Statement

Identify Winning Wholesalers, Understand why they are Winning and Distort Investments

Solution

1. Identified winning wholesalers

   • Growing over 2019-2022 and (2023 tri2->tri3 acceleration from prior year) (for below statistics)
   • Measuring growth on revenue and share
   • Measuring execution and investment by price, distribution, investments to understand impact.
   • When we initially looked, growth in volume was tied to regional political conservatives, especially for short term recovery from crisis. Therefore, we compared against local sets such as state and region to lessen this effect.

2. We analyzed which brands were contributing heavily to the share growth and then after which wholesalers and execution levers are leading the growth

3. We analyzed the investment opportunities and realized local media investments or sponsorships could be made at the DMA level, which is the closest to wholesaler we can get quickly (and a single wholesaler is often in 1-3 DMAs, and single DMA usually has 2-8 wholesalers)

4. So, we pulled many relevant statistics at a DMA level to build a statistical story

   • Market share, share by brand
   • Distribution
   • Investment
   • Price
   • Growth of all the above
   • Model-estimated ROIs of incremental spend/local ROIs (more below)

5. There were two problems with trusting the model to pick which brand/DMA we should add investment into National media is population spread based on zip code population even if a wholesaler has a tiny fraction of that zip code. This has most of our national media assigned to California/NYC, and regions with more overlapping wholesalers, and under-estimates investments into non-wholesaler-overlapping regions. It would be much more reasonable to assume we spend so much on national TV ads for football because of the central republicans who watch football and drink beer, but most of the investment gets mapped to California due to the higher population even though they watch less football and drink less beer.

   • This specifically hurts regression models because we end up with wholesalers with investment amounts comparable to their total revenue.
   • This is an open issue we have been requesting for CP&L and we are now starting to implement collaboratively with them.

6. Our Regression Model doesn’t take co-saturation into account.

7. Trusting all the above issues and solutions, we compiled the below types of information to come up with recommendations (Media or Sponsorship)

   • Wholesaler Brand parent vehicle Local ROI (from model)
   • Wholesaler Brand parent vehicle Total ROI (from model)
   • Wholesaler market share (we don’t want to invest where we will cannibalize our own brands)
   • Wholesaler Revenue% coming from “winners”.
     • We wanted to distort into brands growing within a DMA. Therefore, for winning wholesalers, we defined an overperforming brand contributing to that growth to brands that were growing in share% more than the state average 2019-2022.
     • We considered the % of DMA total revenue that was made up of specific brands in wholesalers those brands were overperforming in, to make sure we selected brand x DMAs that were made up of wholesalers actively working to grow those brands.

8. We recommended

   • Oklahoma City
     • BHL – media
     • MUL – sponsorship (they sponsor the Oklahoma City bucks)