Circle Case study

Using machine learning techniques to increase revenue, conversions and engagement for DFS

An actionable scoring algorithm that updates continually

About DFS

DFS, the UK’s largest sofa retailer and manufacturer, aims to lead furniture retailing in the digital age. Most famous for sofas, DFS also partners with leading lifestyle brands, such as Dwell, French Connection and Joules, to provide a wider range of furniture products.

CACI has worked with DFS for over 20 years and hosts its customer database, providing insight from CACI’s proprietary datasets to support customer understanding and location strategy. Outputs include segmentation, machine learning models, communications strategy, store catchments and digital targeting algorithms.

The challenge: targeting relevant engagement between major purchases

DFS sells to a market where customers traditionally make infrequent, high value purchases, as Mike Aspinall, Data Activation Manager at DFS, explains: “The main thing we sell is sofas, with a repeat purchase average of seven years. This can be challenging for CRM, which is all about nurturing ongoing relationships.

Our challenge was to deliver a CRM strategy that would enable us to maintain relevant engagement in a targeted way. We wanted to understand the opportunity to encourage repeat purchase through a data lens – which customers might be open to further purchases, when they might be likely to make them, and what kind of products they might want.

The solution: an actionable scoring algorithm that updates continually

Mike asked CACI to work with a large sample of DFS’ previous purchase data. “We have a rich customer database. CACI analysed two years’ data to find people who had bought two items from us consecutively within that period, looking at their purchase patterns and pathways.

CACI mapped out the attributes of people who had made the two qualifying purchases using Ocean demographic and lifestyle data blended with DFS behavioural data. The analysis looked at the identified customers’ over and under-indexing attributes, comparing them to people who bought the first item but not the second.

CACI trained a machine learning algorithm which is updated daily, incorporating the latest transaction and customer information. It is applied to the data in DFS’ customer experience platform (CEP), appending data points to customer profiles.

The benefits: conversion and revenue uplifts through highly relevant, low-waste, accurately targeted multi-channel campaigns

Mike tells us, “we had fairly low expectations of the first email in a multi-month journey. But against the control group, we saw an 866% uplift in revenue from the email campaign alone, within 14 days. That’s a four-times conversion increase, measured against a control group of people in the same segment who didn’t receive the communication.

This project was a perfect fit for CACI as an expert data science partner. With our DFS mission to lead furniture retailing in the digital age, machine learning is crucial to engaging our customers with truly relevant, timely communications. We have been working with CACI for decades – their team understands our business and data extremely well and we have a strong relationship.

Read the full case study here. For more information on how CACI can support you in delivering targeted, relevant, customer experiences get in touch and one of our data experts will happily arrange a time to talk.