Measuring customer value is now something old. If we dated things, perhaps we could set in 1987, more than twenty, a formalization of the most successful the proposal submitted Schmittlein, Morrison and Colombo (1). Our readers offer probably other references, and earlier dates. In the 90 years that this idea of measuring and modeling customer value has taken off, from 2002 to literature reviews are available (2). More recently it will see Gupta, Hanssens et al (2006)The sophistication and refinement of models has been the extent of their bloom. Today they are like the flowers of a meadow in spring, occupying all the ecological niches of the market, even if lines of force structure in the clear variety. Between models of retention (lost for good), migration patterns (always a share), the models purely probabilistic, deterministic, autoregressive, Markovian, in those cohorts, the parametric and nonparametric, the survivals and the semi, blends the full range of statistical tools is used.
And yet there is still some confusion, customer value, the life time value, customer equity, are still varied terminology, addressing the same idea, has difficulties to account for the nature of concepts that it means. It is partly the fault of this abundance, which in a competitive spirit, has neglected the economic reasoning and took advantage of the munificence technique.
Let us return to the concept. The basic idea is borrowed to finance on the basis of this discovery 80 years, is managing the relationship is the real purpose of marketing that most purchasing decisions, and considers the idea that the customer is an asset whose value is measured by the method of discounting. In short, the value of a customer is equal to the sum over time of value which he hopes to produce each period. In this reasoning, and this definition, every point is important.
The first is the idea of cash flow. The value comes from the difference between what it calls, rather than costs, but investment and liquidity that obtains. From this point of view there are many models that have lost sight of the definition, confusing cash flow with profit, which is easily calculable, but worse with income. Number of models Life Time Value, what we translate customer value, are now in force models of Customer Life Time returned. An accountant will be very difficult to confuse the two concepts, although we understand that in the minds of marketers things can be confusing.
The second point is that the unit of analysis. Similarly the value of a security not to be confused with the value of the portfolio, the CLTV is a concept unique to each asset portfolio of clients, each client, each client relationship. And each asset is characterized by its expected value, but also risk, usually measured by the variance of this expectation value.
The third point is precisely this concept of hope value. The future gain is not assured. It makes sense that under a policy implemented today which looks at the extensions in time for back to the date hereof. This is the meaning of the update, not a forecast but an equity. What is the value today of cash flows in the future produced by the current policy? The financier to resolve this problem corrects the future value by a coefficient that represents either an arbitrage opportunity, the current performance of the assets, either by correcting the deteriorating value of the currency. It also includes expectations of growth. In the case of marketing, things are a little different since we do not expect that assets are maintained over time. Marketers are pessimistic, they know that there is little chance that a customer today is still in 7 years. Simply put, if the rate of customer attrition is around 30% after 5 years the probability of a client is still below 20%. In other words, if finance the expected gain is a constant average time for marketing is strictly decreasing.
Continuing on this point by examining the concept of hope of winning. From a mathematical point of view the expected gain is the sum of various contingencies, and each is the product of a possible value and the probability that it occurs. In the case of the expectation of customer value, it implies how the larger the probability that decreases over time, on average, and the gain that hopes to achieve, also varies paths that the modeller must determined. Often the assumption is that the expected (average) gain is constant, but the likelihood is that it occurs strictly and strongly decreasing. But one can imagine that the hope of gain follows a parabolic trajectory resulting in the life of the relationship that learning leads to develop the client's expected income, then that like everything else in this world, wear and fatigue erodes income. And regardless, we can say that the risk of being more customer grow in the initiation and establishment of the relationship, and then remain stable, may be weakened later when boredom takes over . Naturally these trajectories are peculiar to each individual, each story.
Fourthly, we emphasized that these changes were consistent with the hypothesis that the marketing policy, ie the effort to acquire the customer first, then to maintain or develop, was the same across time. The update does not measure future, but now it is what it is today and that it continues stubbornly. The acquisition led effort to select a certain type of client. We know that it is important, it is likely that customer is unfaithful and opportunistic, we know that if we chose freely, without being force it is likely to be faithful and generous . Likewise if one does nothing to nourish the relationship, he will be leaving us soon. One can imagine the optimal policies for managing these customers. The dynamic optimization techniques are at our service. The central idea is that much depends on CLTV effort acquisition, retention and development. It is therefore at least in principle the idea of an optimal CLTV which we must compare the current CLTV.
The value of a customer is not a given thing. It is the result of what he is, the circumstances, but especially what he did. It is without doubt one of the most misunderstood. There are no customers with high value or low value, there are customers that are successfully exploited the potential of assigning resources, and those that are neglected. This is probably the key point.
This reasoning also allows a third idea, that of the intrinsic CLTV, which is defined simply as the value generated by a client for whom it would be no investment. This is the value it produces, regardless of the marketing effort. It is obvious that the issue of lifetime value is a comparison of these three indicators: the intrinsic value that requires no investment, the present value measures the reward of the investment policy adopted on the day calculation and that of a hypothetical optimal value which corresponds to the best investment.
At this stage of the analysis we have done almost around what models should be able to report. And we should look now at this other notion is that the portfolio's value, otherwise the Customer Equity. This will keep in mind that if that is the sum of individual values, it is not a new concept, but is confused with what we have analyzed. We will keep for another post this discussion.
We'll just have to take into account an idea that has been developing for some years. Customers are not alone and talk to their neighbors, they themselves recruit new customers, which has led some researchers to want to include these secondary gains in the calculation of CLTV. Our view is that this bias is wrong, less by principle than by its consequences, but we shall return later, we simply reiterate here that the strength of the concept of Customer Life Time Value, is to consider today the future consequences of current marketing policy, taking into account as reasonably possible, the time paths of the relationship with each consumer participating customer base
by : i-marketing
Aucun commentaire:
Enregistrer un commentaire