Due to the high proportion of manufacturer brands, multi-label retail offers a particularly high level of price transparency. In contrast to pure online retailers, the range of services offered by brick-and-mortar formats is not only reduced to price, but also criteria such as service and shopping experience.
In a project with a traditional multi-label retailer, price algorithms were to be developed that take into account not only internal price factors but also context factors such as the strength of the retail brand. The particular challenge in this case was the complex data basis, which first had to be reduced.
In the initial workshop, the retailer’s structures, processes and price rules were first clarified. A uniform target picture was developed across all departments and once again it became clear – price management is multi-faceted! A first data analysis in our Data Lake provided insight into the complex data basis. After the data had been normalized, the focus was therefore on extensive preparation. Formatting, cleansing, linking and transforming were the methodical focal points of our Data Scientists. On this basis, the ideal article groups were formed through a two-stage cluster process. A time series analysis highlighted the seasonal effects and price elasticities by location, product group and brand, providing information on the different price sensitivities of customers. In addition, the available customer traffic data were modeled as valuable parameters. The final validation of the predictive model was carried out by means of a back test. Subsequent simulations with different pricing strategies and optimized framework conditions (including rules and regulations) uncovered the specific pricing potentials and formed the last step in the project phase.
“We knew that our database would make it difficult to develop intelligent algorithms for price management. The results of aifora really impressed us! We not only have far greater transparency over our data, but are now in a position to offer our customers prices that are in line with the market at all times”.
The potentials calculated in the business case illustrated the enormous impact of optimized pricing (4.2% increase in earnings with 8% higher sell-through ratio). Together with our Data Scientists, the retailer is already working on the next step: the integration of the Intelligent Price Advisor.