Analysis: Using data to understand customer behaviour

Legend has it that when Dunn-humby first presented results of its Clubcard initiative to the Tesco board, then chairman Lord MacLaurin declared: “What scares me is that you know more about my customers after three months than I know after 30 years.”

That was 20 years ago and ever since retailers have been seeking to harness the full power that data mining has to offer. Where the grocery industry has led, other sectors, including fashion, have followed. Rapid technological advances have enabled retailers to evolve from a position of treating every customer the same to understanding sub pockets of behaviour within their customer base and, most recently, understanding how individual customers behave in the online environment.

By successfully analysing and leveraging this data, online fashion retailers can give themselves a significant competitive advantage over their peers, but achieving this is not without its challenges. So how should they go about it?

Across the data universes

Broadly speaking, there are three main data universes that retailers should concern themselves with, as Steve Borges, managing director of specialist ecommerce agency Biglight explains. “The first is all of that data that sits within your order history file; so if you’ve been trading for six years you can know how many customers you’ve sold to, what they’ve bought, how many you’ve acquired and how many you’ve lost. If you’re lucky enough to do home delivery you’ve got post code and payment card information as well.”

Then there is information about subscribers, especially those people who opt into emails, which when coupled with the buyer data creates a single customer view. The third universe, says Borges, is around web analytics, both in terms of how consumers are behaving on your own website and what they’re saying about you off-site, on social media and in online forums.

John Lewis head of online marketing Emma McLaughlin says the retailer is constantly mining data about how customers use its website to fine-tune and optimise navigation.

“For instance, on our desktop site, we’ve used analytics and testing alternatives to change the product hierarchy for wedding clothes – to make the most popular products easier to find, by moving them up the left hand navigation,” she says.

The growth of social media, meanwhile, has allowed marketers to collect previously unattainable qualitative data about how consumers feel about their brand. By marrying these data universes, online retailers have been able to hammer down into not just what consumers do and how they behave but why they behave as they do, enabling them to offer a unique service to each customer.

Derek Eccleston, head of research at eDigitalResearch, says that in recent years online retail shopping has become increasingly about providing visitors with a personalised experience. “For example, by combining purchase data with user insight, retailers should be able to tell what a customer purchased [or did not purchase] on their last visit, and offer up to them similar products, which, in turn, should help to increase conversion rates,” he says.

Amazon is widely acknowledged as the trailblazer in personalisation, while grocery retailers are also expert at offering discounts and promotions based on their knowledge of what individual consumers buy.

Fashion retailers have, on the whole, been slower to embrace complex personalisation but the signs are that the wheels of change are in motion. Matalan has been working with technology firm Rich Relevance to launch an online personalisation service, which means customers will be presented with suggestions for products based on their previous browsing and purchasing activity, while greater personalisation is set to be a key feature of John Lewis’ forthcoming website relaunch.

Sizing up customers

New technology is also allowing retailers to gather useful information on size. By allowing shoppers to ‘try on’ clothes in a virtual fitting room where consumers are asked to input their dimensions, retailers can not only reduce the chance of the consumer returning the product due to poor fit but also gain valuable insight into why they are making a purchase and, more importantly, why they are not.

“We are not only giving brands the data on what sizes they should carry but what sizes they should carry that they aren’t carrying,” says Heikki Haldre, chief executive and co-founder of virtual fitting room provider Fits.me.

Such data can also be used for justifying investment decisions. Hawes & Curtis head of ecommerce Antony Comyns says the shirt maker was recently considering what new shirts to add to its range. “We had already put into work a very slim-fit shirt that was missing from our range and we did that with a minimal amount of research based on very simple information that people were giving us. We were able to do some proper analysis from our Fits.me fitting room, which showed that we were missing that shirt and it gave us the numbers to back up what we initially thought.”

Of course, access to data can be a thorny issue and, despite the proliferation of free information in the digital age, consumers are still fiercely protective of their privacy. Bombarding consumers with offers or requests for data can be fraught with danger and winning the trust of shoppers is not to be taken for granted. Take email marketing, for example. Most consumers are only signed up for a handful of fashion retailers’ emails so you have to work hard to make the list. The Drapers Customer Insight Report 2012 found that 42% of consumers are signed up for one to three emails, and a further 22% for four to six. Even when you make the shortlist, reaching the customer is not a given. Only 35% open three quarters or more of the emails they receive from fashion retailers and 24% open less than one quarter of them.

Driving decision making

Nevertheless, customer feedback and opinion is key in driving decision making, says McLaughlin. John Lewis recently changed the order of entering name and credit card details on its mobile site as a result of customer feedback – a small change that McLaughlin says really improved conversion rates.

Naturally, there are ways of improving your chances of getting customers to part with information. “What we’ve found is that using online research techniques can help to mask people’s identity and, in turn, encourage them to share and disclose more about sensitive topics,” says Eccleston.

“We can offer people a shield and give them the confidence to talk more honestly and reliably about topics ranging from health issues to sex toys.”

In some situations, people are more willing to take part in research and offer up their opinions if they get some noticeable recognition in return, says Eccleston: “For example, we saw a boost in return rates when we linked survey feedback to a branded loyalty card scheme, giving users the chance to gain extra points for every piece of research they completed,” he says More personally targeted emails are also better received than generic circulars. In the Drapers report 34% said they would like more personalised emails with products most likely to appeal to them, and the same number would welcome personalised special offers and discounts depending on what they buy. The report also uncovered that consumers are willing to share sizing information if it leads to better availability.

What is certain is that the rapid advancement in data exploitation spells good news for online fashion retailers that are just as concerned with what you say about your clothes as they are what your clothes say about you.

Future gazing – big data

“In the future, the consumer will be in control. Every shopper will have instant, trusted and secure access to all of the information and data they need to make a purchase decision, right there and when they need it, and in the format of their choice. This could include anything from relevant purchases they’ve made in the past and how they rated and reviewed them, as well as what their friends, trusted networks and respected others had to say,” writes Derek Eccleston, head of research at eDigitalResearch.

“Expert reviews from trusted sources could also be integrated, alongside supplier collateral and background information. Independent operators will also provide the best price or deal, highest rewards and most attractive purchase terms and instantly allow shoppers to purchase at the click of a button.

“All of this will be secured via a two- or three-factor authentication to help improve trust between shopper and retailer. Finance to support these types of purchases will be close to instant, too, to help make the entire customer experience as smooth and as seamless as possible.”

The growth of social media has allowed marketers to collect previously unattainable qualitative data