1 month ago
5
ARTICLE AD BOX
New Look needed a new way to engage with its customers. Its chief data officer quickly set to work. (Photo: Shutterstock)
Times are tough for UK high street fashion brands. River Island, once a mainstay of the British shopping centre, looks likely to fall into administration. Claire’s Accessories, meanwhile, is poised to restructure its business, and other household names such as Debenhams, Cath Kidston and Select Fashion have either failed or experienced fitful, uncertain resurrections after being carved up by their creditors.
New Look is in the same boat. Employing approximately 10,000 people across around 340 stores, the brand remains the UK’s second-largest womenswear retailer. Last year, the firm made a statutory loss before tax of £21.7m – an improvement on the previous year’s £88m losses, but still worrying enough for New Look to shut down 38 stores, commission a strategic review and invest £30m in its back-end digital transformation.
But therein lay the seeds of new growth, explains its chief data officer, Daniel Chasle. Despite 2024’s falling revenues, e-commerce sales remained strong. But there was also much room for improvement. Customer sentiment online was fickle, and sales at third-party e-commerce partners had begun to falter. A strong brand with a good product portfolio, New Look clearly needed to address its consumer engagement strategy.
Chasle began by canvassing customers on what they really wanted from a fashion retailer. The answer? Options that fitted their unique style – and, it had to be stressed, lots of them. “It is shocking how much our customers expect us to know them and personalise for them,” says Chasle. Like Spotify’s auto-generated playlists, Amazon’s item recommendation and Netflix’s ‘Watch Next’ lists, the New Look customer expected their brand to know them so well that they wanted the thought taken out of clothes shopping altogether.
Facilitating that kind of online telepathy might seem like a tall order – but, as Chasle well knew, anything is possible if you have the data. What New Look’s chief data officer needed to do, however, was find a way to rationalise it.
New Look needed to find a way to unify millions of customer profiles. By doing so, it would be able to see if customers had shopped with the brand in-store, get a view of how many profiles they had online, understand their frequency of contact, see what they buy, and analyse how they respond to marketing emails or other points of contact.
Chasle describes a highly fragmented technology infrastructure that was getting in the way of New Look seeing its customers clearly. “We were working purely on an operational profile, so individual registrations and failures to complete were all treated separately,” he recalls. What’s more, adds Chasle, “we used a traditional data warehouse solution with limited data storage and compute capacity, so there was no ability to do data science.”
A quick change that makes a big statement
The company had tried to bring together all of its customer data before, with mixed results. Stitching profiles together had left black holes and distortions that were causing more problems than were being solved. So, New Look changed tack, beginning by chucking its IBM data warehouse and unpicking the patchwork of systems where its customer data had been siloed. After a brief but stunningly effective pilot programme, it moved to the Databricks Data Intelligence Platform through its engagement with Ameprity, the AI-powered customer data cloud.
The pilot, conducted in January, quickly convinced Chasle he was on the right track. “We signed the contract at 2.30pm and by the next morning millions of transaction history records had been transferred and semantically tagged,” he says. “Ten days later, we had enough results to take our findings to the C-suite. We signed up fully in February and, by May, we were using the first version of the Customer Data Platform (CDP) solution, which we are now refining.”
Very quickly, it emerged that 3.4m customer profiles had been fragmented across multiple records, and more have come to light since then. The customer database is vast and, so far, 33m profiles with a four-year history have been integrated. There has been a 10% consolidation of its active customer base and, crucially, New Look now knows who its most valuable customers are and how to sell to them.
“We knew our highest value customers were omnichannel, and that they tend to be messiest in terms of profiles,” Chasle explains. “They have one profile to do returns, multiple profiles to get discounts, and maybe they play the system a little – but we understood that.”
Stitching all of the interactions together unlocks the treasure chest of personalisation. For example, New Look’s data analysts can now see on what days a high-value customer is likely to open an email, so it can time its offers just right. It understands if you have a preferred price range, if you buy at full price or discount, and everything else it needs to know about purchasing habits. One would hope that the new system will, in time, arrest the decline in sales at New Look and help it avoid the fate of so many of its rivals on the UK high street.
“We needed a new kind of capability to compete with the top quadrant of retailers like H&M, Zara and Shein,” Chasle notes. “Being really data-enabled and using data to change products and contact customers shows that we want to match that level of capability.”
