The Value of Customer-Lifetime Value in the Hospitality Industry
Putting the Value in CLV
By Matt Lindsay President, Mather Economics LLC | August 28, 2016
As simple as it sounds, knowing which customers are profitable is a challenging task for many businesses. Predicting which customers have the most profit potential is even more challenging since it requires estimation of future business. Applying these same modeling techniques and profit calculations to potential customers adds yet another degree of difficulty.
Customer lifetime value (CLV) is a (relatively) old concept. Businesses have been calculating the expected operating margins received from a customer from long before the birth of the internet and modern data tools. Even though technology has changed significantly, the challenges largely remain the same. Let's start with the data. For a hotel manager, measuring CLV requires knowledge of the variable revenue and costs from an individual customer, which are determined by the nature of their activity with the hotel. Collecting data from these activities and combining them at a customer level is challenging. Does restaurant activity not charged to their room get associated with that customer? Do the incremental costs associated with their use of the business center, gym and executive lounge get captured and assigned to them? If the revenue and costs are assigned, are they treated correctly? Is there an allocation that occurs? If so, what is that calculation? We have seen many businesses draw faulty conclusions about the profitability of certain customers due to an internal cost allocation decision that does not reflect the true behavior of costs in response to changes in customer activity.
Having a loyalty program that provides an incentive for customers to share data on their activities within the property and the brand at large is one way to tackle the data challenge. Using statistical modeling to measure the variation in costs due to customer activities instead of static cost-sharing rules is a solution to the bias from internal cost allocation.
The second challenge for measuring digital CLV is forecasting the expected purchases of a customer. In many cases, businesses use historical purchase data as a guide to future behavior, which is a reasonable approximation in most cases. A best practice is to develop a forecast algorithm using econometric techniques that can adjust purchase forecasts for customers in response to changes in factors that affect behavior such as price changes, loyalty campaigns and product enhancements.
If a CLV score is calculated using a forecast algorithm instead of a static historical retention curve, we call it a "dynamic CLV" score. Developing these algorithms using data from the entire customer base and then applying them to individuals with similar characteristics enables predictions for customers that have minimal or no historical purchase data. Validation of the predictive powers of the algorithms can be done with A/B testing, where a representative sample of customers is selected for use as a control group. The combination of econometric modeling and testing of the predictions is a best practice not only for CLV scoring but other activities as well.
Once a business has an effective CLV metric in place, it can monetize that investment through customer acquisition efforts. An effective customer acquisition tactic is to offer discount offers to prospective customers. This can be done through a number of marketing channels with differing degrees of customer targeting specificity. Dynamic CLV can guide businesses as they decide what offers should be presented to individual customers and what channels should be used for the communication of those offers. Some customers are more receptive to the price of the room, some value the amenities of the property, some want proximity to other destinations. Understanding which type of incentive is most important to a customer can dramatically improve the ROI on marketing investments.