According to Bain & Company and Harvard Business School, where both report that increasing customer retention by 5% increases profits in a range of 25% to 95%. The same study reports that existing customers are 50% more likely to try new brand products and spend 31% more, compared to new customers. Will the above be entirely accurate? In my opinion, it does not necessarily apply in all cases and for all circumstances. To look at the case further, you have to look at customer longevity and organizational profitability. This also according to a report written by the Harvard Business Review. The correlation between these two variables must be strong and close to 1, otherwise the further from 1 means a weak relationship between profitability and customer longevity.
It also has to do with the metrics that are selected to determine the performance of the loyalty strategy and the way in which they are used. There are several myths that have become solid over time and today there are few scholars who debate and question them. It should be done with greater suspicion and discipline to support and validate certain hypotheses and thoughts that have existed for a long time. In particular, the RFM method, which is the most used but has several defects or details that are not normally taken into account to carry out the analysis.
The RFM Method for Determining Customer Value
It is the most common method used to determine customer profitability; it is based on the frequency as well as the amount of consumer spending with the brand. The RFM methodology can be useful in some cases, but it is not entirely accurate. According to HBR, the study carried out by Reinartz and Kumar establishes that relying on frequency and recency can lead brands to make incorrect decisions about which customer segments they should invest in. The logic used is very simple. Taking recency, a marker is assigned by segmenting customers using their purchase activity, for example, in the last six months, between six months and a year, and more than a year. Following the same hypothetical case, the frequency is then used to determine how many times the customers bought in those same 3 periods: twice, once or never, also assigning a marker. The two markers are added. In short, the more times and more recently the customer has purchased, the higher the score. So far, everything looks normal. But one detail is missing to add: the rhythm or cadence. Let’s use the case of Juan and Pedro, who started buying products in month one. In the course of the first year, they both buy at different periods: Juan buys in short intervals of time, making transactions in the second, sixth and eighth month, while Pedro waits a longer period of time, transacting after seven months, that is. say in month eight.
The traditional RFM analysis would suggest that Juan is more loyal than Pedro, since his purchases are more frequent and recent, therefore, the brand should invest more in him. But the RFM method will fail to take into account that Juan normally buys every 2.3 months, but by month twelve he hasn’t bought anything in the last four months. Pedro has not bought anything in the last four months either, but his purchasing pattern is like this, that is, he buys in longer periods of time, it is in his normal time range. On these bases, it is statistically more likely that Pedro will buy again in the future, therefore, the brand should invest more in him than in Juan. This model of buying behavior is called historical events. Using this model, there is a 20% probability that Juan will buy again, while there is a 45% probability that Pedro will do so. (The details of the calculation and the rest of the text can be consulted in the book “Increasing Customer Loyalty” published by the Harvard Business Review.).
This model is very useful for predicting how quickly a customer’s activity will decrease as the probability of remaining active decreases, thus preventing brands from over-investing in profitable but disloyal customers.
