- Like Resource
A leading UK-based retail chain wanted to scientifically assess the impact of campaigns, and store initiatives such as format and layout changes, on footfall, sales and profits.
The client engaged The Smart Cube to provide specialist data analytics skills, sector experience and tailored tools to deliver a robust solution framework for estimating impact of trial initiatives at a store and category level.
The Smart Cube solution
The Merchandising Analytics strategy involved devising a test and control methodology to meet numerous client requirements, including measuring the effectiveness of campaigns, receptiveness of new and innovative products, and success of changes to store formats.
The study’s geographic scope included 1,500+ stores (supermarkets, convenience stores, petrol stations, cafés, etc.) across the UK.
The Smart Cube’s team conducted exploratory workshops with key stakeholders to determine the test and control stores. While the test stores were those where the initiative was implemented, control stores were identified by mapping various trends and proximity attributes for each test store. The project team simultaneously defined specific KPIs (such as sales, volumes and number of transactions) and store attributes (such as size and competitor density) to be considered for the analysis after client approval.
As initiatives were introduced in test stores, the team started collating data from both test and control stores to analyse it across levels—such as type of campaign, store cluster, and product category. Some of the key models that helped determine the impact of the initiatives were around the algorithms that determined “lift” and “net rise in sales volumes”.
Based on the first analysis, further iterations across store initiatives were implemented, to develop a robust methodology and an automated solution that could be utilised for future initiatives with minimal effort and faster results. It was developed using various techniques including co-integration, slope test, Euclidian distance, and correlation.
Multiple client teams – marketing and promotions, space and formats, merchandisers and buyers, and strategy and innovation – have used this solution to trial new initiatives.
Some key initiatives where the Merchandising Analytics solution delivered impactful recommendations include:
- Store layout transformation: The client was trialling an alternative checkout plan to encourage customers to form better queues near checkout tills, and wanted to assess the impact of these changes on store performance. The project team analysed multiple KPIs – sales, volumes, number of transactions and percentage of transactions at self-service tills – and store attributes (size, competitor density, number of tills and ratio of self-service tills) across test and control stores. As a result, we provided the business with estimates on net increase in store performance (in terms of sales and other attributes) and recommendations on improving the trial before full roll-out.
- Store strategies to mitigate competition: The client wanted to assess the effectiveness of various types of initiatives in countering the impact of a competitor’s stores. To determine the effectiveness (increase in sales, impact on store and store cluster, etc.) of these initiatives, the project team analysed data for test and control stores across levels; for example, weekly and daily summaries of KPIs, and aggregates across time periods – such as pre- and post-, as well as during, the initiative. Our solution identified store types best suited for different initiatives, and provided recommendations on an optimal roll-out plan for the full portfolio of initiatives.
The Smart Cube provided the client with a solution that enables:
- Evidence-based planning: A consistent solution framework to measure impact of different types of trial initiatives, currently in use to support a range of use cases in store design, space allocation, and new campaigns
- Cost-effective innovation: A cost-effective way to experiment and validate new initiatives, before commitment to large changes in operations or campaigns
- Strategic driver of profit: Recommendations on improving trial features through analysis of short-term tests, resulting in a more profitable long-term strategy