Data-driven pricing strategy used to drive 2% increase in net profits for retailer
Updated: May 29
A national grocery retailer was looking for opportunities to drive profitability of it’s under-performing convenience store formats.
After discussing a few different options it came to light that their pricing strategy had never been reviewed in isolation of the larger store estate (convenience store prices were +5% over larger stores due to higher operating costs).
An approach was devised to review relative price position of various categories using 4 key drivers of profitability
Price perception – how aware customers were of the prices of various products
Price elasticity – the degree to which demand changed relative to price
Margin – the average margin of each product
Volume – how much of each product was sold per week
Solution & Results
Customer research was initially undertaken to validate the factors of the shopping experience that were most important to them (price, convenience, store experience, value, quality etc) and then to specifically understand how strong their price perception was by category (how competitively priced different types of product were) – this was used as input number 1 to the model. For input number 2, a machine learning (Random Forest) price elasticity model was built on historic sales and price data to provide an accurate view on how much sales volume was impacted by increases and decreases in price by category. These input were combined with volume and margin inputs to help determine the right mix of price positions across the store estate.
The new prices were initially implemented in a small set of trial/control stores for a 12-week period, where the new price strategy directly drove an increase in profits by 2%. The pricing strategy was subsequently rolled out across the entire 500+ convenience store estate.