Hugo Boss develops personalized consumer strategies from mountains of data
According to Hugo Boss’ Anthony Milano, “We have all these assumptions about the customer. Retail thinks they’re doing one thing, marketing thinks they’re doing another — but what is the customer doing, really?”
As vice president of e-commerce for the premium high-fashion retailer, Milano has a significant stake in understanding Hugo Boss’ customers. The company is global, with more than 6,000 points of sale in 124 countries; Hugo Boss AG, the parent company, owns some 900 retail stores and has more than 1,000 franchised outlets. In 2015, its total sales topped $3 billion.
Part of Milano’s responsibility involves sifting the customer data acquired by the company’s online operation, merging it with data from store transactions, flowing the combined data together with third-party information (Google searches, Facebook likes, etc.) and extracting a coherent picture of the behavior and intentions of various customer segments.
Then, based on that information, Milano and his team develop personalized — or at least highly segmented — promotional campaigns that would have a positive effect on sales.
That was the idea, anyway. “We were using a heavily customized legacy customer relationship management system,” Milano says, “and it was difficult to get a dynamic look at our customer base and understand what was really going on. Projections and analysis took longer than we needed, sometimes weeks and months.”
The search for a more effective solution led Milano to a company called AgilOne.
AgilOne was founded by Omer Artun, who holds a doctorate in computational neuroscience and physics from Brown University, where he studied under Nobel Prize winner Leon Cooper. Before starting the company a decade ago, Artun worked as a retail consultant at McKinsey & Company, as vice president of strategic marketing at CDW/Micro Warehouse and as senior director of B2B marketing at Best Buy.
He also served as an adjunct professor of marketing at the New York University Stern School of Business, where he taught graduate-level relationship marketing and analytical marketing courses.
Old idea, new tools
“The basic concept of what we’re doing is not new,” Artun says. “People have been talking about customer relationship management and one-to-one marketing for many, many years. But a decade ago, when we talked about CRM, it was relative to transactional data — what have I bought before, am I a valuable customer, and so forth.
“What’s changed is that we now have access to data on when customers are browsing, what they’re searching for online, what their behavioral patterns are in their mobile device interactions and so on. This kind of data, subjected to predictive analytics, can give you usable information about that customer’s preferences and intent to buy things.”
As Artun points out, however, there are a couple of difficulties involved. One is that this kind of data is extremely time-sensitive: If something isn’t done about it within a few hours, or at most a few days, its usefulness expires.
The other is that the data stream from which customer intent data is extracted is huge and at least partly unstructured. Making use of it requires the assistance of people with backgrounds in things like machine learning and artificial intelligence.
It also requires a great deal of computational power. “If you want to process the data in real time and if you want to make it available in ways your mobile application provider can use, this thing needs to live in a live, superfast environment in the cloud,” Artun says. “There’s no way to bring all this data into your own data center and process it overnight.”
A step at a time
What Artun describes is precisely the way AgilOne is set up, which the team at Hugo Boss has so far found very effective.
“We wanted to start by getting a basic idea of how our customer groups react and what their behaviors are,” Milano says. “It was something we couldn’t immediately get from our previous provider, and AgilOne was able to do right out of the box.”
“Our higher-lifetime-value customer joins the database during a full-price period. Rather than focusing our acquisition efforts on capturing the most clients, we can focus on capturing more higher-quality clients.”Anthony Milano
This beginning phase, Milano says, has provided Hugo Boss with some interesting and at times surprising information. For example, it has challenged some ideas about customer acquisition.
“We get the most new entrants to the database — people who make the first purchase and become a customer — during highly trafficked sale periods, whether online or in the stores,” Milano says. “But we learned that our higher-lifetime-value customer joins the database during a full-price period. Rather than focusing our acquisition efforts on capturing the most clients, we can focus on capturing more higher-quality clients.”
Phase two involves analyzing the trends that emerge by customer segment and adjusting to them, often by offering narrowly targeted promotions. Rather than driving traffic with an across-the-board discount, for example, Hugo Boss tried offering an incentive specifically to customers with a lower propensity to buy, as measured by AgilOne’s predictive analytics.
Rather than driving traffic with an across-the-board discount, Hugo Boss tried offering an incentive specifically to customers with a lower propensity to buy.
“As a premium luxury brand, we have to be very conservative with discount incentives,” says Milano. “When looking at a campaign, we are able to say ‘OK, here’s group A. We know that these people have a higher propensity to buy, so let’s [not offer them an incentive for the campaign]. But this other group are people we want to get back into the store, and they could use a bit of a motivator.’ So we tried it, and we definitely saw a lift.”
A culture shift
AgilOne provides its clients with a playbook of strategies that can be executed very quickly, several of which — such as cart abandonment and more targeted email campaigns — Hugo Boss has already put in place. In both cases, Milano says the retailer experienced an almost immediate lift in revenue. He and his colleagues have a number of other new strategies in mind, many of which are in the process of being reviewed by the larger organization.
“It’s a bit of a cultural shift, to be honest,” he says. “We all have these campaigns we want to produce, and now that we have better data I would say — to anyone — prove that what you’re doing is important, and that it actually provides some kind of lift. Once you’re really looking at the data, you’d be surprised to see how many resources are put toward things that don’t have the lift they are assumed to have.”
But the data is available and, Milano says, its value is making itself clear. Both he and Artun emphasize that this technology really does constitute a new day — real, rather than hoped-for, customer centricity — and that the key is to put it to work immediately to solve problems.
“Advice I give to a lot of my customers is, don’t let the perfect get in the way of the good,” Artun says. “Don’t over-analyze. We challenge them to think backwards from a customer’s perspective. I’d rather have basic analytics that the customers can experience than more sophisticated analytics that live on a PowerPoint slide somewhere.”