Case study: Price optimisation for the retail sector2019-02-12T11:11:56+10:00

Price Optimisation for retailer

Improve revenue and profit margin through Advanced Analytics

Problem

Pricing optimisation is one of the most common challenges in retail. One needs to account for a range of facets such as promotional timing, market shifts and others in order to establish an optimal pricing strategy that maximises growth or profitability. In addition, anticipations of competitor actions are needed given the complexity of demand dynamics in modern markets.

An FTSE listed FMCG company was looking to re-invent how they approached their pricing strategy. This involved three different components:

  • They wanted to pivot from their pricing being guided by gut feelings to a scientific, data-driven methodology;
  • They wanted to pursue a market share acquisition strategy on their high visibility, competitive products;
  • But, they also wanted to retain higher margins (and hence profitability) on their less competitive offerings.

Solution

Given the dynamic nature of most markets, we undertook an extensive approach of modelling the underlying drivers of demand by building a price recommendation engine that automates the derivation of an optimal price based on the business objective- quantifying crucial components of the problem such as expected demand, prices elasticities, cross-product cannibalisation, and halo effects. This exceeded traditional approaches that often fail to respond adequately to changing market conditions and where accuracy is only maintained through frequent recalibrations of the initial model.

The price recommendation engine is rooted in a foundation of sophisticated maths and analytic techniques. An initial stage pre-processes and cleanses the data into an appropriate granularity and purity. Systematic trends and seasonalities are then decomposed and removed in order to isolate the effective demand and cross-product elasticities. A nonparametric model was subsequently fitted on the data in an iterative process that improves model generalisation accuracy. Lastly, numerical optimisation techniques were applied to optimise prices across all products at the same time and in accordance with cannibalisation, other cross-product effects and the appropriate pricing strategy. Moreover, the ability for an end-user to specify expected competitor actions equips the price recommendation engine to adjust the optimal price point accordingly and enables the user to actionably respond to competitor actions with foresight.

Result: Our price recommendation engine facilitated an automated and responsive way for which optimal pricing strategies can be derived to maximise revenue or profit. This resulted in greater market share and profitability for our client.

BOOK A MEETING WITH US TODAY.