What happens when brands know too much about you? Customers are happy to share their preferences on Facebook and through the use of loyalty cards, but sometimes it pays to be less clever than is actually possible when it comes to each separate incidence of customer engagement.
If customers are used to using six, seven, or even more channels to contact brands then tying all those interactions together in a logical way that helps to manage customers more effectively does create a lot of data that can be extremely useful. This process of connecting channels is often referred to as ‘omnichannel’ service and if it improves service interactions then surely this could only be a good thing?
The issue is that sometimes companies can forget that there is a fine line between helpfully tying together the various strands of interactions across various channels and just appearing to know too much. A piece of personal data shared in a transaction online, in a web-chat or on a voice call cannot necessarily be assumed to be information that can be referenced arbitrarily at some point in the future.
There are many stories about retailers sending pregnancy related discounts to young women before their family know about the pregnancy. The New York Times published a detailed example of how this works back in 2012, but it’s a recurring theme – just how much information should be used, even if it is available?
Even fairly innocuous examples can be creepy to some consumers. Nobody minds visiting a restaurant and being asked by a waiter you know if you want ‘the usual’. But if a waiter you had never met addressed you by name and asked if you wanted your favourite dish, because your entire purchasing history is available to the restaurant staff, then that could be helpful or scary – depending on your point of view.
Information on customers can be very useful and can unobtrusively improve service. An airline automatically selecting an aisle seat for a customer with this preference is a good example. However, the use of preferences as a result of data mining and propensity profiling is fraught with difficulty in the absence of a filter that judges the ‘appropriateness’ or ‘acceptability’ of the application of that preference in a particular customer scenario – and should be planned very carefully. Moreover, ensuring customer care advisors are well trained and properly managed is fundamental in a non-face to face environment where only voice tone or use of syntax can help judge what is pertinent vs. impertinent use of data.
Photo by Hartwig HKD licensed under Creative Commons