Client Situation
•Client had been experiencing unbalanced per capita sales across states.
•Advertising and promotions had not been effective in lifting sales for underperforming states.
•Client sought answers to key questions:
What drives consumption growth in the category?
What action should to pursue to close gaps in target markets?
What early-warning signs could be monitored in order to identify upcoming drops or surges in demand?
Results
•Identified which factors were driving regional differences and should influence growth plans.
•Quantified the implications of macro demand trends.
•Built a comprehensive database incorporating client’s proprietary assets and data harvested off of the public domain.
•Implemented individual regional growth plans for sales expansion
•Reversed per capita underperformance in two-thirds of the underperforming states
SFC VALUE-ADD
When growth slows, improving sales in underperforming regions can be a huge opportunity for improving overall company performance. In many cases, it is much easier to improve performance in lagging regions than it is to further increase sales in overperforming regions.
For any modeling to be valid for extrapolation of strategy recommendations, the dataset that models are based on must be comprised of both over and underperforming regions and the models must test for a variety of hypothesized factors driving sales.
The biggest hurdle associated with properly estimating demand is the specification of the model. It is crucial to specify a model that captures the important factors driving sales trends. The challenge is knowing which factors are truly driving sales trends and should influence strategy and resource planning. Instead of narrowing factors subjectively, we cast a very wide and relied on Machine Learning to establish which ones truly matter and affect sales.
In addition to client's proprietary data, we harvested the public domain for broader economic activity data, regional macro and competitive data, as well as media, trade promotion and in-field resources for competitors.
Harmonizing these various data assets required extensive data merging and wrangling. The result was a large database available for model development and simulations.
The solution to identifying which factors truly drove sales was to assess all relevant factors simultaneously.
The result of our statistically-based estimation was a detailed understanding of the factors that drive sales at multiple levels of activation. With this knowledge, the client could more effectively allocate resources across initiatives, functions, and geographies.
Equipped with these relevant insights, we were well positioned to evaluate the manufacturer's growth levers based on statistically-derived contributions to sales.
These evaluations, in turn, allowed us to build simulations and conduct scenario assessment for longer term "what-if" analyses and planning.
Using our simulations, we developed tactical action plans. After assessing the impact of various initiatives and their cost/difficulty in execution, we could finalize operating and business plans to improve per capita sales in underperforming regions.
