Customer lifetime value and pricing

STEP 1 Create insight into CLV of current customer base

Start by visualizing the lifetime value of current customers. Start with sales figures, then refine based on product mix, margins and cost to serve.

Be careful with averages! An example from our consulting practice: We analysed new customer sales over the first 5 years. On average, sales from year 1 doubled during years 2 through 5. But a closer analysis showed that a large group of customers (>>25%) hardly generated any additional sales in years 2 through 5. However, the top 20% customers more than tripled sales from year 1. The key question: what are the characteristics of these groups and what explains the variation in CLTV?

We offer this analysis to every company with a sales force. Because we see with regularity, especially in B2B, that salespeople give high discounts to customers because of the customer’s supposed future potential … but no one checks whether that potential ever comes true.

STEP 2 Build model to predict CLV

In order to apply CLV, it is necessary to translate past and present insights (step1) into the future. This requires understanding the drivers of variation in CLV. Those who understand that can build a model that predicts what a customer’s CLV will be. That is necessary to make the concept of lifetime value actionable.

Companies that have large customer bases and data sets can use artifical intelligence to build a CLV-forecast model. But again, if you want to get started with it, start with the most obvious characteristics of the different customer segments based on CLV and then start refining. Example: we found for one of our clients that business customers were 3 times more profitable over their lifetime than residential customers. Within the B2B segment, we were able to further refine the prediction of CLV based on variables such as:

– expected growth rate of the customer organization (explanation: a customer who is growing rapidly is likely to buy more than a customer who is not growing);
– cultural fit (explanation: it was found that companies with a cultural fit stayed customers longer and purchased more)
– cost of serve (background: customers who had their own technical service came with fewer questions and thus had lower service costs).

Similar mechanisms can, of course, be found in the B2C market as well. Consider the customer who purchases his insurance online, or who calls to do so and wants to be helped by an employee. It seems likely that the difference in behavior during the buying process also says something about the customer’s preferences in the service process, and thus about the cost of service. After all, past behavior does not say everything but it does say something about behavior in the future. This is what the concept of “claim-free years” is based on (those who make claims are more likely to make claims again).

STEP 3 Integrate CLV into marketing and pricing strategy

Last step is to integrate the CLV forecasting model into marketing and pricing strategy. The applications are numerous. Well-known example: insurers differentiate their premium based on CLV. Drivers include: age, zip code, claim history (think claim-free years), behavior (smoking).

Some other examples:

Offer scarce products selectively: ‍An e-warehouse preferably serves out the last available products to customers who bought on installment (because they make a much higher profit contribution). Marketing automation based on CLV.

Price differentiation based on prediction return rate: ‍Zalando and peers are struggling with return rates of up to 50%. Extremely costly. Several companies are currently working on predicting return rates. A link to price is obvious.

Marketing in specific zip code areas: ‍An energy company managed to use smart algorithms to predict which customers would stay longer once they became a customer, so they knew exactly where to go by the doors.

Price differentiation based on channel usage: ‍Clothing store offers part of its assortment and deals only online.

Acquisition discount based on probability of repeat purchases: ‍A company in the events industry provides deep discounts only when the CLTV forecasting model indicates a high score, with the prediction of repeat purchases being a particularly strong determinant.