Sentrana
The Science to Lead Markets™
Case Studies
Top 5 Diversified Retailer in Canada
OBJECTIVE: New U.S. based competitors entering the Canadian retail space with more efficient cost structures and supply chain capabilities forced this retailer to move from Everyday Low Pricing (EDLP) to a High-Low Price (HLP) retail strategy. To succeed at an HLP strategy, the client needed to accurately predict increases in consumer demand and profitability resulting from price-based promotions to avoid eroding profit margins due to over- and under-stocking.SOLUTION: To predict consumer demand precisely, we mathematically modeled several trillion Bytes of historical data, including advertising expenditures and sales data, and the resulting model was incorporated into a pricing decision support system.
IMPACT: Demand forecasting accuracy was improved from 62% to 84% (i.e., 84% of the time, we were within 12% of actual demand), labor required to manually reconcile pricing conflicts from disparate pricing systems was reduced by 46% for each Store-SKU combination, and gross profit margin for the modeled Store-SKU combinations improved by 17%.
Fortune 50 Internet & Media Services Provider
OBJECTIVE: The advent of multiple channels for accessing Internet content rapidly transformed the once specialized service of online access into a commodity service. The unique, competitively differentiated service provided by our client (the dominant company) in this industry came under swift and vigorous attack from rapidly proliferating entrants capable of providing a similar service at deep discount. This caused a rapid increase in this Fortune 50 Company’s customer defection and dramatically eroded its pricing power.SOLUTION: To optimally execute preemptive retention marketing, statistical modeling was undertaken to predict the churn probability of each customer (note: the prediction goal was at a finer level of detail than simply predicting the churn probability of a customer segment). Also, the modeling sought to uncover which specific marketing message would be most aligned with each of these customers.
IMPACT: This solution enabled the client to reduce its annual churn rate from 9.2% to 6.3%, which resulted in a revenue savings of $109 MM.
Fortune 200 Hardware Distributor & Retailer
OBJECTIVE: A global Home Improvement & Hardware retailer with annual sales exceeding $13B, and with more than 4,000 stores across 70 countries, has come under increasing competitive pressures from retailers that have larger assortments and lower cost structures. Each store is run as an independent proprietorship – consistent SKU prices cannot be deployed across all stores since each proprietor has the flexibility to charge their own price. Although each store can exploit local micro-market knowledge to set prices, the central warehouse cannot accurately predict demand, nor can the retailer execute a unified pricing strategy to compete. Varying levels of IT sophistication and maturity from store to store – makes centralized operations management and marketing effectiveness difficult to optimize.SOLUTION: Determined price elasticity at the SKU-Store level. We used SKU-level elasticity, instead of category level elasticity, to pricing for each store, combined subject matter expertise from local store and national managers with econometrics to yield pricing and assortment models that maximize market share growth and profitability, and determined optimal cross-category promotions to increase size of customer purchases.
IMPACT: Increase revenue by 3.2% and profits by 7.9% for the product categories and stores that were modeled.