Big Data

Big Data: Why Marketing Will Never Be The Same Again

May 14, 2014 

On paper, big data is any marketer’s dream come true – thanks to the internet, if a business wants to know about its customers, more information is available to them than ever before.

Companies have been collecting data on us for decades, but since we all went online - and in particular since we got social online - the amount we give away about ourselves and the number of ways a business can exploit that information, have exploded.

Of course this has been great for forward-thinking companies like Target, Amazon and Wal-Mart who have been developing the infrastructure for years in anticipation, but those slower on the uptake risk going the way of the dinosaur. With big data comes big challenges – collection, storage and analysis all take thought and resources, and must be done more efficiently than the competition is managing.

From the consumer’s point of view there are the ever-growing privacy concerns. People are becoming more accepting of the fact that they can get better prices and better service from companies in exchange for a limited amount of information about themselves. The younger generation in particular seems more receptive to this idea but we all have our own ideas of what is a reasonable amount of information to divulge.

Many may not be concerned about the example I’ve previously used of Wal-Mart bringing together weather, stock and customer data to target a promotion involving a barbecue cleaner to specific customers. But what about when it is insurance companies that want to dig around in the digital detritus we leave as we share, chat and organise our lives online?

Big data is already causing massive change in the marketing industry – last year a poll by market research firm GfK found that 62% of marketers said it had already “fundamentally changed their role” and 86% said it would continue to do so in the future.

Examples are rife – if you shop online you will have seen big data marketing in action.

• Amazon were the pioneers and still considered field leaders. They have recently announced they are investigating “anticipatory shipping” – using what they know about you to predict what you will buy and start sending it to you before you order it.
• Online ticket retailer StubHub also uses big data analysis to offer promotions to customers most likely to be interested. Some markets are particularly ripe for these strategies – sports fans for example, with their lifelong undying loyalties, make perfect targets, in the eyes of data marketers.
• Last year eBay revamped the front page that customers see when they log in, to provide a more big data-driven experience. You can choose categories to follow, allowing them to log your interests and use them to work out what you might be interested in seeing.
• Netflix developed its complex “personalised genres” system based on the way their customers watch just about every film or TV show ever produced, and categorizing them with hundreds of “tags”. This allows it to know that you particularly enjoy watching, say, foreign films with a dramatic twist at the end – and serve them up as suggestions.
• In the bricks-and-mortar world, supermarkets are experimenting with “near field communication” which allows them to send targeted messages to customers based on their previous purchases, as they walk past.

Those examples mainly rely on structured data – companies having access to lists of structured information such as your age, gender, location and purchases.

But by some estimates this accounts for just 20% of the information we share online – far more exists in the unstructured Facebook posts and tweets, and the mountain of blog posts, videos and sound recordings created each day.

Gathering, analysing and using that ethereal, unstructured data is a greater challenge but offers huge rewards to the companies that do it right. For example, retailers can now send their CCTV video images to a cloud based service provider that then uses algorithms to understand who customers are (based on face recognition) as well as how they behave (how they walk through the store, what they pick up, whether and what they purchase, how long they stand in line at the check out, etc.)

Once again we turn to Wal-Mart for the best – or depending on your point of view on the privacy debate – worst – example.

Its Social Genome Project involves what is termed “deep semantic analysis” of everything from Facebook and Twitter to Youtube and blog posts to build up individual profiles of its consumers – and sell to them.

Questions arise about the level of access they will get, and how it will differentiate between publicly-shared data and privately-shared data.

The extent of access to customer data that Facebook allows its biggest corporate customers is not always clear. And paying Facebook a large chuck of its multi-billion advertising budget, Wal-Mart would certainly be considered a valued customer.

Whether you are optimistic or cautious about the way companies will market to us using big data in the future, it will continue to evolve. Governments may see further legislation as necessary if public concern grows over the level of access corporations have to privately-shared data – as I think it is likely to. But marketers will adapt, because that’s what they do