On the flight back from the London Big Data Week, tasked with writing an article about the rise of identity theft, I had a rare moment of insight where the two topics suddenly presented themselves as two aspects of the same phenomenon.
One of the primary values of working with Big Data, after all, is that one may use insights drawn from one aspect of consumer behaviour to drive innovation and insight into another. At a pretty basic level, for instance, the purchasing history of your customers can be paired with data on the way in which they navigate your website, and the insights gained can be used to drive improvement in each field.
The thing is, this process is almost identical to that performed by identity thieves: comparing apparently discrete data sets to undermine the anonymity of both sets. This process can be usefully understood through the DIKW (Data, Information, Knowledge, Wisdom) Pyramid: at the lower levels of the pyramid, your customers might not care that you know their social security number or email address. However, if you are using analytic techniques to generate data on the behaviour of individuals, those same individuals are going to start to feel creeped out.
While I’m not suggesting that companies working with Big Data are interested in anything other than improving their sales, the apparent similarity between the two processes, at least in the eyes of the consumer, raises some problematic questions.
If you are using the voluntarily-given information to generate insights into individual customers’ behaviour, at what point will your knowledge of individual customers stop being seen as useful, and instead be regarded as a breach of trust?
There is, of course, no simple answer to this question, but below I will outline a few strategies that I have found to be useful in avoiding precisely this problem.
Primarily, I think, the way to ensure that your customers don’t resent you using their data is, to be honest with them, and to build a corporate culture that is as honest and transparent as possible without giving away trade secrets.
Be Transparent With Your Customers
First and foremost, companies should realize that customers are significantly more informed about privacy issues than even a few years ago. Indeed, today it sometimes feels as though there is something of an arms race’ going on between consumers and companies about this specific issue. Individuals consistently and increasingly tell pollsters that they value their privacy. Services such as VPNs and encrypted cloud storage that promise to protect this privacy, and that used to be the exclusive domain of hackers and security professionals, are now mainstream.
On the other side, companies are collecting more data on their customers’ than ever before. Sometimes this is done in an open and transparent way, and driven by a genuine desire to improve services or product delivery. Sometimes it is not. I have seen instances, for instance, in which companies have collected and stored sensitive (and potentially identifiable) data because, basically, they can. Facebook’s data collection is a prime example. Big data marketers have told companies for years to collect everything they can on their customers, just in case it becomes useful for a future marketing strategy, and after all this data is (or was, until recently) relatively easy to collect.
Some would point out, of course, that consumer behaviour belies the idea that customers are actually interested in privacy. While claiming to value the security of their personal data, many people willingly give it to companies who promise no such thing. This, however, doesn’t justify such practices. Just because you can collect something, it doesn’t mean you should. Further, there is actually a good business case for not collecting certain kinds of data. If (or when) your customers find out you know their address, DoB, and every country they visited in the last year, they are going to feel either betrayed or angry.
The solution here, as it often is in such cases, is, to be honest with your customers up front. Tell them what you know about them, and they’ll likely accept such data gathering as the price of an improved service.
Be (almost) Transparent With Your Rivals
This transparency, I would also argue, should also extend far beyond your relationship with your customers. On the most immanent level, this means building a corporate culture that values transparency, and allowing the public not just your customers to see the way that you work with data. This, in fact, is the advice given by many business experts: Cathy O’Neil, former Wall Street guru and author of Weapons of Math Destruction, has argued that the potentially negative effects of big data can only be combated by being open about your analytic capabilities.
This will be uncomfortable advice more many companies. Every company is in a competitive environment, after all, and in sharing the way you work with data you run the risk of giving away important (and potentially lucrative) trade secrets. The trick, therefore, is to strike a balance between building a culture of transparency without giving away the keys to the farm.
Striking this balance can be difficult, but there are a few simple steps that can help. Customers are generally most concerned about systems like neural networks that appear to work in a very opaque way: Black Boxes‘ that generate insights without showing their workings. While it is possible to use such systems in a responsible and transparent way, doing so entails being open about the kind of data you have, and the types of insight you can generate.
In fact, if done well, this kind of explanation can actually be turned into a marketing opportunity, where you explain to your customers (and make your competitors jealous) your advanced systems. It will also mean that when you send personalised marketing to your customers, they will feel valued, rather than accusing you of stealing their identity!