Fresh Data Blog
Fresh Data Archive
Case Study: Using Big Data for Customer Acquisition
Date: February, 2015 --
Case Study: Using Big Data for Customer Acquisition
Much has been discussed about leveraging the contents of your database in order to acquire customers and in this Case Study we’ll detail how one online retailer did just that in order to reverse downward trending customer acquisition figures. We’ll highlight the type of one to one communications that were created using data-driven segmentation methodology, detail how intelligence was gathered from acquisition contacts for whom there was no transactional or affinity data to begin with, and outline the response metrics to the associated campaigns used to measure success. This Case Study was assembled by Data Services, Inc. in cooperation with our friends at the Prescott Report.
With more than £1.7bn ($2.6 million) in annual sales, Shop Direct is the UK and Ireland’s leading multi-brand digital retailer, delivering 48 million products every year to four million customers. 86% of its sales are online. And 51% of that 86%, being 44% of sales, are from mobile devices.
Traffic in its many on-line shops is robust at over one million visits per day. Its shops’ brands are household names and range from the relatively new Very.co.uk and isme.com, to the previously offline-only Woolworths.co.uk. Also in the collection are the venerable Littlewoods.com and KandCo.com, which brings together Kays, Universal, and Empire Stores.
In short, this is not a “mere digital upstart” and its work was a Silver Award winner in the UK DMA’ Awards competition.
The Problem: The customer is Tired of Her Wardrobe
Shop Direct was seeing weak response rates to its digital advertising for women’s clothing, while at the same time research showed women were becoming tired of their wardrobes. Customers were suffering from the “a wardrobe full of clothes but nothing to wear” syndrome. It was concluded that the company needed to reinvigorate its relationship with its fashion conscious customers with something new.
Find the Old Spark with “Love Your Wardrobe”
The chosen strategy was to engage a carefully selected audience with a campaign urging the recipient to “Love Your Wardrobe”. This would be done by spotlighting a selection of attractive styles for seven “occasions” of the Spring and Summer seasons: ‘a day at the office’, ‘wedding’, ‘summer holiday’, ‘festival’, ‘gossip with the girls’, and ‘barbecue’. (We marvel that there is no “day at the ballgame” or “day at the beach”, and are brought quickly back to the realization that England unfortunately has not yet discovered baseball, and as for beaches, they’re too cold to spend time on them looking nicely dressed.)
Each set of wardrobe recommendations was carefully constructed to be tasteful and relevant to the occasion and be inspirational within that mood/context. Of course all the offerings would be available for purchase at the Very website.
In addition, care would be taken to segment prospects to talk to them about credit offers from the company. This would reinvigorate credit customer acquisition, often treated in the UK as a separate business line.
Segmentation is Multi-Stage & Multi-Media
The segmentation of potential targets was step one. The ‘Love Your Wardrobe’ campaign was carefully targeted, multichannel, and designed to help new customers solve their fashion dilemmas and fall in love with their wardrobe again. At the same time it was driving browsing and response behavior that enabled revision and retargeting as the campaign progressed, a nice “feedback and modify” cycle.
The campaign employed timely use of both digital communications (email and social media) and direct mail. This was accomplished by the use of emails to showcase the new Spring/Summer offerings in the ‘occasions’ listed above. The initial emails were then followed by more targeted and segmented messages based on each prospect’s click-through pattern and browsing behavior. Direct mail was used for a unique objective described below.
Put another way, the strategy was to first determine customers’ “fashion confidence”, and to then use dynamic content to serve inspiring looks.
And for the new customers…fashion-awareness confidence profiling.
With no transactional data for cold prospects, a solution was developed that would not only reveal how fashion confident each prospect really was, but would help deliver an engaging brand experience. These prospects were sent a launch email inviting them to take an online style personality quiz.
Their answers gave insights into their shopping behavior, favorite shops, favorite purchases, attitude to fashion and purchasing influences. The prospect pool was then segmented into three wardrobe attitudes based on these insights: ‘style safe', ‘chic style hunter’ or ‘fashion forward’. Each “attitude” was presumed to reflect a particular level of fashion confidence. Based on this segmentation, Shop Direct sent a dynamic, personalized follow-up direct mail pack revealing their “style personality”.
Now we know you and can better advise!
With all this behavioral data on hand, the balance of the email campaign would continue with email messaging linked to constantly rotating and shifting collection of outfit suggestions. Landing pages were used as style advice “centers” for each of the seven segments and served to spotlight the outfits’ suitability for numerous other occasions.
Love Your Wardrobe produced 12,667 new customers and exceeded targets by 22%, at a time when response rates were falling. In addition, the targeting delivered a 20% reduction in the cost of acquiring a new credit customer.
With an average first-order value per customer of £165 ($248), this was an initial return of over £2.5m ($3.76 million) for an ROI of £4.76 ($7.16). This seems satisfactory, of course. However, taking the average lifetime value of a Very.co.uk customer into account, total ROI is expected to be more than four times that amount.
Note that the campaign involved the use of substantial amounts of data. The plan to generate “new-to-file” names and contact details was complex. Address details had to be entered and confirmed quickly, email addresses and consents and opt-outs recorded. Preferences coded and entered. There were many steps that could have gone wrong.
Remember to discuss your marketing plans with the database management experts at Data Services, Inc. to ensure your critical customer and prospect data is fit before running your next data-driven marketing initiative.