Everything you need to know about using data to accurately recognise your customer

Data is fast becoming the key currency in the brand-consumer relationships, according to Archant’s chief marketing officer Will Hattam. It is now exchanged at all but a few engagement points, with significant benefits to both parties he says.

The ability to know who an individual is, whether online, offline, mobile, in-store or wherever else, is every marketer’s holy grail, “whether they realise it or not,” adds Jed Mole, Acxiom’s European marketing director.

Without tackling customer recognition, nothing else happens, Mole says, because the better you can do that the better you can do every other aspect of data-driven marketing. Being able to identify your customers means you can connect the right data to the right people and develop a more accurate customer view. And that means that marketers can generate smarter insights and, in doing so, deliver a better customer experience through people based marketing.

There are many life stage triggers and intent data that marketers can pick up on – and these are of high value for today’s marketer. Just look at what changes daily in the UK alone: 9,590 households move; 1,496 people marry; 810 people divorce and 2,011 people retire.

However, how do you recognise those selfsame customers and prospects in the online and offline worlds? Jane Smith from Manchester might be identified as ‘JS Smith buying a car’ or ‘J Smith buying a holiday’. If she moves, marries, or both, that’s yet more complexity – how can a marketer know that she’s the same person if she now has a different surname and address?

“Fuzzy matching” – the traditional way that marketers marry an identity – is limited says Mole. “With fuzzy matching, you’re asking computers to deal with things that they don’t like doing. They like 1s and 0s, black and white. And when they’re asked to deal in probabilities, they are limited to the quality of data they’re dealing with.”

Hattam explains further: “Data anaylsts often use fuzzy matching when comparing disparate datasets that have no persistent key.”

Questions are often raised regarding the quality and accuracy of the resulting matches, due to possible human error. In fact, some industry experts believe it can be a brand risk.

Says eBay’s UK and EU multinational advertising head Rob Bassett: “We know that when it comes to customer recognition, it’s paramount for brands to get it right – and socalled ‘fuzzy matching’ poses huge risks both for publishers and brands.”

He says that by supplying user data which isn’t fully formed or wholly accurate, publishers will misinform brands, who will – in turn – mis-target users.

It’s why many in the industry are looking towards knowledge-based matching. It is a technique which uses a trusted knowledge-base of data to understand different presentations of the same individual over time, and determine if they are the same person (or not).

Simply put, a marketer can ’see’ Jane Smith of Manchester and JS Brown of Liverpool and be confident that she is the same person.

Multiple digital identities are even harder to determine. Most people will have both a work email and at least one personal email address, if not more.

According to GlobalWebIndex (GWI), the average consumer has 3.64 connected devices – be that a PC, laptop, mobile and/or tablet. But are they truly ‘connected’?

Consider cookies, which most websites use today to ‘remember’ a consumer. They can help deliver a much-improved customer experience but some consumers are suspicious of their misuse and will delete at will, making it more difficult to deliver a person-centric experience. They are also limited in their use on mobile environments. It means that today’s marketer must look beyond yesterday’s technology to truly connect with the consumer.

By using a range of online identifiers (in a privacy compliant way) it is possible for marketers to stitch together a rapidly growing online mass of fragments for data and add it to the offline clues they might have about a consumer. Knowledge-based identity resolution allows a marketer to recognise the same customer online and off, across channels and devices – and to build more accurate insights.

Adds Bassett: “By better understanding their audience [in this way] brands can make informed predictions on a customer’s next logical purchase, based on previous activity and shopper trends.”

This article was originally published in The Drum magazine.