I am a frustrated fan of customer segmentation - and I really do have to share my frustration with you. Let's start simply however with a childhood memory.
One of my best childhood memories are the many, many hours I spent sitting in my grandmother's living room window sill on the fourth floor watching the cars as they drove around Esbjerg's largest roundabout directly below, while German TV ads played in the background. "Ford. Die tun was," they would say. I loved cars. I made detailed observations about what kind of people drove which type of cars. A rich fishing skipper in a silver Mercedes 190 turned out towards the villas along the coast road. A family in a red Toyota Carina turned towards the bungalows in the suburbs.
Fast forward to today and in many ways, I am still doing the same thing: putting people and things into boxes. Imaginary boxes, that is. And you do it too!
You, me and everyone else do it - it's just the way our brains are programmed. Just look, for example, at Daniel Levitin's fascinating and well-researched book. We categorize, group, classify and segment. It is - to put it rather simply - a way in which we can navigate a path through all the complexity, information and wealth of choices we all meet in life.
Daniel's book went directly to the nerve of my frustration. Because customer segmentation is a necessary tool for how we (companies and external consultants such as myself) can create customer experiences, which not only provide value for the customer but also improve the bottom line. Without them it is difficult to understand how we are to use all the information and knowledge we have about the customers. Without them, we do not work efficiently. All too often, however, we also fail with them.
Time and time again I get frustrated over customer segmentation that never really manages to get high enough off the ground to make any kind of change. To put it bluntly.
This is a latent problem, when 50% of the companies, according to Gartner, expect to design their business around customer experience by 2018. At the same time, the bar for when we regard good as being good enough is constantly being raised. In Sweden and the UK, for example, which both have national benchmarks for customer satisfaction, scores have been noticeably rising over recent times. This means that in order to be just a reasonably good bank, grocery shop or real estate agent, one must constantly improve one's performance!
If customer segmentation is to be a useful tool, however, as in the future predicted by Gartner, it must really be tightened up. Not only in the quality of the segmentation itself but also in relation to how we work with it to create better customer experiences. Which is why I have been carefully considering what I believe are the biggest hurdles one must overcome along this road.
I have chosen to categorize my hurdles into three areas:
#1: The customer's needs are central. But can they see this themselves?
Customer segmentation should be based on the customer and the kind of value one delivers to the customer. This is not easy to do correctly however, even if you put the work in.
One example is the British supermarket chain Morrisons. A few blogs posts ago, I praised them for a having produced a new, data-driven customer loyalty program called Match & More, which offered the customers a certain level of value. They had put in some skilled analytical preliminary work in order to see the shopping process through the customers' eyes. The program was targeted towards the customer segment that shops at different places over the course of a week in order to buy things cheapest. The solution was an advanced loyalty program, where everything that the customers bought was price-matched against the current lowest price of competitors via an advanced algorithm. The price difference was paid out in £5 vouchers, which could be redeemed in Morrisons. It was a solution that offered both clear value for the customer (you need only shop one place, thus saving a lot of time) and the bottom line for Morrisons (increased share-of-wallet). I thought it was a stroke of genius. It turned out however that Morrisons (and myself) were wrong.
Less than a year later, it was dropped. The feedback from Morrisons was that the customers did not understand it. It has now returned to a simple program much like those offered by the other supermarket chains. What a shame. The lesson, perhaps, is the word reciprocal: We need to categorize and segment customers. By necessity, however, they must have the same need! The customers find it hard to understand why Match & More has to be categorized and compared with the loyalty programs of its competitors, for example Tesco’s Club Card or Sainsbury’s Nectar Card.
#2: Amygdala, Hippocampus & Partners
No, Amygdala, Hippocampus & Partners is not a law firm. Rather, it is a new approach to getting to know customers and their needs better. It is about the importance, looking forward, of being able to attach neuroscientific knowledge to traditional customer insight methods. Because customer segmentation should be supported by facts and customer insights, and with the help of neuroscience, we can raise the quality and precision of these customer insights to far greater heights.
We now know, for example, how memories (including customer experiences) are stored in the brain's hippocampus, which is a part of its emotional center, the amygdala. It is estimated that 95% of our decisions are influenced by the amygdala. This is important information when it comes to designing customer segments: which emotionally-based reaction patterns can we observe? Go further with this train of thought and consider what this means to how we design a customer survey - what are the right questions to ask? Or which data, which is especially interesting in a Big Data project, with neuroscience providing prior knowledge of the results.
And this is why I call it Amygdala, Hippocampus & Partners. Neuroscience as an integrated partner with traditional customer insight methods, where this interaction increases the quality significantly. I am in no doubt that this is the future. Most people however still need to learn to crawl before they can walk. And there are many hurdles in wait during the learning process, where the chances of us not using the insights gained from neuroscience correctly are great.
#3: All for one and one for all. Or what?
In my opinion, customer segmentation must be used actively across the entire organization. All for one and one for all. If just one link in the chain works according to its own rules, the whole thing starts to fall apart. This is perhaps the hardest of all the three hurdles: How do we ensure that the entire organization actually uses it and works with the customer in focus? Below are three examples of where things often go wrong.
- Avoid a closed loop in sales and marketing. I have experienced many times how customer segmentation doesn't really have a life outside of sales and marketing. Segmentation must not simply be a sales tool, however. The customer journey does not stop once the product has been sold. It is pointless for marketing to use an ad to promise its Platinum customers in a hotel chain loyalty program, for example, an extra-special service at check-in, if the hotels' receptions are not able to deliver this extraordinary service or don't even know about it.
- Avoid #CMO = #Segmentation? A new CMO often means a fresh start. A new customer strategy as well as new customer segments. A great deal of analysis and new development is made - or perhaps the CMO simply arrives with his own preferred approach to segmentation under his arm. It is only natural for a new captain to set a new course. Just think however how often a CMO changes job. And then think too how long it takes for the entire organization to get used to the new segmentation. There is a good chance that it won't manage to prove its worth before a new CMO arrives.
- No one is happy if a rollout fails. It can be incredibly difficult to create a segmentation which everyone thinks is great. The different personality profiles (yes, I'm categorizing) who work across an organization's classic functions do not think alike. The finance department may feel that the segmentation lacks precision, the sales department that it is too complex and technical, while the R&D department (God help me) report back that it lacks innovation. This places huge demands, not only on the quality of the segmentation but equally on how skilled we are at implementation, communication and rolling out new strategies in the organization. More often than not those people, who sit and develop a new segmentation, are not implementation experts. This is what I believe many organizations overlook.
The nature of a hurdle is their unpredictability. Once we can predict them and tackle them - well, then they are no longer hurdles. I look forward then to becoming a customer segmentation fan free from frustration, because we - all of us, who work with it in the companies and as consultants - have learned our lesson and know the solution.
Back on my grandmother's window sill I had a trick, if I was especially curious about a car and its passengers on its way around the roundabout: Her old opera glasses. It was a poor trick really, because the car would have been long gone by the time I managed to have the opera glasses in place. When I think back to those opera glasses today however, they remind me of a general ethical problem in relation to hurdle #2: All this new research and knowledge that has been gathered in the crossroads between big data, small data, brain scanners and neuroscience makes for a pretty potent mix. And where do we draw the line for how close we can and should get to the customers? Is there even a line? I will be looking closer at this mix in my next blog post. Completely free from frustration.