Customer Spending Research Articles

Customer retention is not enough

PurposeEvery company knows that it costs far less to hold on to a customer than to acquire a new one. That is why customer retention has become the Holy Grail in industries from airlines to wireless. Yet defecting customers are far less of a problem than customers who change their buying patterns. Today's typical metrics of customer satisfaction and defection do not tell a company how susceptible its customers are to changing their spending patterns. This article seeks to investigate this problem.Design/methodology/approachThe investigation was carried out through a two‐year study of the attitudes of 1,200 households toward companies in 16 industries.FindingsMcKinsey found that focusing on smaller changes in customer spending can have as much as ten times more value than concentrating on defections alone.Originality/valueThis is important for those companies that wish to retain their customers.

The Effect of Musical Style on Restaurant Customers' Spending

Classical music, pop music, and no music were played in a British restaurant over the course of 18 evenings. The mean spend per head for each table was calculated for starters, main courses, desserts, coffee, bar drinks, wine, overall drink bill, overall food bill, and total spend. Total time spent in the restaurant was also measured. A MANOVA analysis revealed that there was an overall significant difference between the conditions with classical music leading to higher spending than both no music and pop music. Univariate analyses indicated that there were differences between the conditions on mean spend per head on starters, coffee, total spend on food, and overall spend. These findings were consistent with the limited previous research, which indicated that the playing of background classical music led to (a) people reporting that they were prepared to spend more and (b) higher actual spending. The results indicate that restaurant managers can use classical music to increase customer spending, and the results are discussed in terms of three possible explanations for this.

First-Store Loyalty and Retention

Stores prosper when customers give them a high share of their spending (share loyalty) and retain them over long periods (retention). We investigate share loyalty by measuring the percentage of the customer's spending in their first store. Retention is measured as first-store retention over a period of 21 months. We study the association between these two forms of loyalty both directly and, by examining their correlates, indirectly. We find that first-store loyalty and retention are positively related but have only two correlates, brand loyalty and attitude to the store, in common. Several factors are correlated with only one form of loyalty. Store retention is particularly related to the accessibility of the store; approximately half of the reasons given for defection are related to store access. Specific age-income segments may have high first-store loyalty but low retention and practitioners can use this evidence in their choice of target segments. Store loyalty is explained as the outcome of several customer processes: as a 'time-saver' effect – a rational allocation of effort given available time and money, control by the environment, attitude to the store and a propensity for routine. The first was not well supported; there was evidence in favour of the last three.

A Scalable Parallel Algorithm for Self-Organizing Maps with Applicationsto Sparse Data Mining Problems

We describe a scalable parallel implementation of the self organizing map (SOM) suitable for data-mining applications involving clustering or segmentation against large data sets such as those encountered in the analysis of customer spending patterns. The parallel algorithm is based on the batch SOM formulation in which the neural weights are updated at the end of each pass over the training data. The underlying serial algorithm is enhanced to take advantage of the sparseness often encountered in these data sets. Analysis of a realistic test problem shows that the batch SOM algorithm captures key features observed using the conventional on-line algorithm, with comparable convergence rates. Performance measurements on an SP2 parallel computer are given for two retail data sets and a publicly available set of census data.These results demonstrate essentially linear speedup for the parallel batch SOM algorithm, using both a memory-contained sparse formulation as well as a separate implementation in which the mining data is accessed directly from a parallel file system. We also present visualizations of the census data to illustrate the value of the clustering information obtained via the parallel SOM method.

Manager Effect on Customer Behavior: An Unobtrusive Assessment

The present study sought to determine the effect of personalizing customers’ experience in a service organization by having a senior management staff member greet customers in a “warm, friendly manner.” It was hypothesized that manager behavior would significantly influence customer spending and retention time. Male customers were randomly chosen from the client population of a service organization and randomly assigned to either a control or experimental condition. Experimental group members had the opportunity to interact socially with management, whereas the control group did not. The total amount of time and money spent by customers was recorded in an unobtrusive fashion. Statistical analyses revealed that a significant difference existed between the behavior of the two groups of customers.