Customer experience 101: Navigating today's data minefieldby
Data, data, data. It’s hard to perform any kind of business process without being expected to have reams of the stuff to justify your suggestions and decisions.
Understandably so. It’s been reported that 90% of the world’s data has been created in the last two years. The amount of information at our fingertips is staggering, and tools for analysis are becoming faster, leaner, cheaper and more innovative. Everything can be backed up with analytics. The processing power behind some of the technology is remarkable.
Big questions remain over what type of data is most important, however – especially in the marketing world where confusion is rife about the differences between ‘big’ data, ‘small’ data, ‘dark’ data and…well, the list goes on. The jargonists have been flummoxed by their own doings. Clarity as to what’s real and what’s spiel becomes a necessity, and has been from the moment ‘big data’ was unleashed upon an unsuspecting public.
Teradata’s longstanding CTO, Stephen Brobst, an expert on the topic of data having spent over 25 years in the data warehousing industry as well as once being an influential member of Barack Obama’s first term technology advisory committee, suggests the obsession with pigeonholing datasets with different terms is damaging many businesses’ attempts to successfully understand what types of data are useful to them:
“When I talk about Big Data, it really isn’t about the size, it’s more about interaction vs transaction data. Even small data isn’t necessarily small,” he says. “It’s just small in relation to the interaction data. We shouldn’t get confused by these marketing terms because they hamper our understanding.
“Whether you’re a start-up or a large business, the value density of the small data is much higher. It’s not an issue and you don’t have to choose – generally people start with small data because you can extract the value much more quickly and in some respects it’s easier, but that big data, which we call the interaction data, enhances the capabilities of what you can do with the small data.”
As well as the justification of size, the term ‘dark data’ also throws up questions about our understanding of what is structured and unstructured. However, Brobst suggests the term ‘unstructured’, in a sense an alternative term to dark data, is one that shouldn’t be used in the first place, and is often thrown around by “database bigots who only understand relational data and believe anything that isn’t relational is unstructured”.
The truth is, most organisations won’t need to delve into the world of dark data, nor will they have the capabilities to do so. Gartner refers to this type of data as "information assets that organisations collect, process and store in the course of their regular business activity, but generally fail to use for other purposes". There are plenty of reasons for said ineptitudes, however:
“Firstly, the term ‘dark’ makes more sense than ‘unstructured’ because all data has structure – if it’s interesting it has structure,” adds Brobst.
“But if you look at data that isn’t being fully exploited like, say, rich media data, video data, the reality is there is a ton of content in there if you have the algorithms to get and extract it; it requires different kinds of processing models and technologies over the traditional relational depositories. That’s why we have the concept of the unified data architecture instead; with this it allows you to combine traditional structured with the non-traditional data. So, the idea is to shine the light on dark data and extract the value from it. You will have to use different technologies and techniques, but there is structure in that data.
“The reality is though, the people who are sophisticated in this are not what I’d call ‘businesses’; they’re the people in dark trenchcoats looking for bad guys. Special agencies and such. Those types of organisation are probably the most aggressive at mining that dark data. What happens in those agencies with no names, well eventually those technologies trickle out and get used and become more mainstream; but at the moment most businesses aren’t at this stage. There’s perhaps a small minority of leading edge businesses investing in this stuff, but it’s early for them at the moment.”
Beyond our knowledge of what different types of data are and how they are useful to us is perhaps the Holy Grail for marketers: personalisation.
Many businesses are currently in a state of flux about the best approach to making their brand offer more personalised experiences. Some of the world’s biggest organisations are convinced it is central to the future of brand loyalty, marketing and customer service, yet our understanding of how to actually deliver on this promise is still in its most nascent form.
Of course, data is at the heart of personalisation, yet Brobst suggests many businesses are missing the point, stating that, “a lot of the data we originally collected about customers and still collect about customers tells you the customer value – what did I buy, what’s in the basket and all that stuff,” and that it, “doesn’t give you a full picture of the customer experience.”
He adds: “If you want to get one-to-one targeting and so forth, you need to understand the customer experience, not just the value. The value tells you how much you should be willing to invest to get the customer, retain the customer, maintain the relationship. But how to invest is very much based on the customer experience.
“If I’m an online retailer, my traditional data warehouse will have everything a customer bought on my site; the coupons and all that stuff. However, a more sophisticated online retailer – someone who has made the transition into data for personalisation – is not just looking at what was bought, but every click and every search that led up to that buy. If you can understand those things, you can better understand what the customer experience is and how to make it better.”
Brobst argues that too many businesses, especially in ecommerce, get this point wrong, and that it is starting to irritate consumers:
“Let’s say you bought some stuff for your brother for Christmas, and your brother’s tastes are very different to your tastes. You want to be able to say “that wasn’t for me, don’t show me those books, that’s what my brother reads, it isn’t what I read, show me something different”.”
He recommends testing as a solution to this problem, regardless of how laborious it may seem: “Test everything – from the colour of upper right corner of an advert to do customers prefer red or green in the first place. Everything. Experiment.
“Organisations that build big data warehouses say they have huge amounts of data, but that doesn’t mean analytics sophistication. It’s about running experiments and letting them inform how you run your business. The big difference is using data to justify decisions; the data itself isn’t actually helping you.”
Location, location, location
One area of data that Brobst implores businesses both with or without a physical space to pay attention to is location marketing. Using mobile signals to track customers has been seen as hazardous by some, but the benefits of being able to collect data about people both in and out of the digital space means many businesses are likely to forge ahead with their own location marketing plans in the near future. How the data is used is likely to be central to success:
“If you have a mobile app and it’s enabled to understand where you are, it can dramatically enhance the customer experience and give a lot of data to the enterprise. In terms of cross-selling, in terms of informing customers where they need to go to find a particular product… there’s a lot of movement happening in this space. Marketing people used to use postal codes, and that was their idea of geospatial analytics, but that’s no good anymore. Now, our more sophisticated customers understand longitude, latitude, 3D geospatial tracking, that kind of thing.
“Retailers got excited because online you could track a customer’s path through the digital store, but now with the mobile app, that’s part of the shopping experience and now they can track through the physical store, depending on the return value – helping customers find products, giving them discounts. This is when the customer experience innovation lies. Not in how much data you actually have and determining how structured you think it is.”
Chris is Editor of MyCustomer. He is a practiced editor, having worked as a copywriter for creative agency, Stranger Collective from 2009 to 2011 and subsequently as a journalist covering technology, marketing and customer service from 2011-2014 as editor of Business Cloud News. He joined MyCustomer in 2014.