I have been thinking about how we measure (and manage) the lifetime value of customers in the networked age we find ourselves in.
Only a few years ago, most of the work on customer lifetime value assumed that the customer was an ‘island’ of consumption, and thus that his value was only influenced by his actions. Simple lifetime value models just looked at the customer’s revenues, costs and how long they would remain a customer, and used discounted cashflow to calculate value.
CLV = (revenues – costs) * tenure
More advanced models recognised that the future is very hard to predict and introduced scenarios and real-options to refine the customerlifetime value calculation.
CLV = ((revenues – costs) * e.g. 5 years tenure) + ROV (6+ years tenure)
Hardly anyone got around to recognising the risks that eat away at customer lifetime value and to incorporating it into the value calculation. Or to recognising that customers sit within a portfolio of customer segments of varying value and risk that need to be managed as a whole rather than individually. All of these approaches assume that the customer is an island.
Reality is that the customer is part of any number of ‘communities of influence’, each of which – as the name suggests – influence the customer’s behaviour. Studies by Cap Gemini, Double Click and many others show that other customers’ recommendations are often one of the most often used and the most influential sources of information used by customers making purchase decisions. And that applies across the board from automobiles, to mobile phone providers, to banking services.
If customers are highly connected rather than islands, it goes without saying that other customers are key drivers of a customer’s lifetime value. This leads to a couple of questions. Question One: How can we incorporate this into the customer lifetime value calculation? One way to do this is to add a ‘recommender value’, i.e. the value of the customer’s recommendations to other customers, to the customer lifetime value calculation. If the customer was an evangelist of the company, the recommender value would be positive. If the customer was not so keen on the company, the value would be negative. And Question Two: How do we calculate the recommender value? Early work is focussing on social network analysis and other data mining techniques to identify recommender value. Sunil Gupta of Harvard Business School, one of the founding fathers of customer lifetime value is also looking at answering these questions. But it is still very early days.
As I find more about this interesting new aspect of customer lifetime value, I will post the findings on the CRMGuru blog.
What do you think?
Post a comment and let’s get CRMGuru2.0 buzzing.