Let’s first look the generic model of the customer lifecycle in the image above, from left to right. In general, a company is trying to get the interest of a group of prospect with a specifi marketing offer. Only a portion of these prospects becomes interested to our proposal (response rate);  out of these, a fraction finally becomes a customer of our business (conversion rate).  We can define as average cost of customer acquisition (Subscriber Acquisition Cost)  all that we spent to: a) identify our potential customers (create the lists), b) get their awareness of our offer and c) convince them to buy, divided by the total number of acquired customer.  Once these people become active customers, they buy from us, generating recurring transactions with an average revenue per transaction (v) and an average transaction frequency (f). As we saw before, each transaction has both a revenut but also a cost to serve, thus providing a ‘transaction profit’ for the service consumed or product bought. If we observe the customer after a certain period, a fraction of all customers will stop buying from us (churn rate), while those who remain loyal (return rate) will continue to buy in the period afterwards. This means that the value of that customer is not only one purchase, but the sum of all the purchases he will do as long as he reamin with us. The real value obviusly will be the profit made out of these purchases, ie. the revenues minus the cost to serve the customers. Hence, the CLTV can be defined as the net present value of all the profit from from purchased services along the customer lifetime, less the initial cost of customer acquisition and the present value of all costs incurred to retain these customers (Subscriber Retention Cost, SRC), and to keep them loyal to our company. As you can see below, the CLTV is composed by:

  • the NPV of all cash flows generated from profits on sold services (revenues per customer multiplied by the total number of customers in that period and by the profit margin);
  • the total acquisition costs incurred at the beginning  (average cost per customer multiplied by number of customers);
  • the NPV of allthe costs incurred along the customer life for his retention.

We could not simply sum the cash flow from each period, but we need to use their net present value because flows happening in the future must be discounted to account for their value loss over time, which is why we need a discount rate (t). .

2012-09-01-Groupon-CLTV-Customer-Lifetime-Formula

Obviously the number of active customers in a given period is equal to the number of customers in the previous period multiplied by their retention rate. With few simple mathematical steps, assuming that the average revenue per customer and transaction frequency are constant across all period, we obtain the final formula shown above in gray. This formula tell us exactly what are the main value drivers of the CLTV, i.e. the factors that can increase the average value each customer will bring us along his lifetime with our business. The main parameter improving the CLTV are:

  • the average revenue from each sale (v);
  • the buying frequency (f)
  • the profit margin the company has on each transaction (PM%);
  • the customer retention rate (rr)
  • the average life of our relationship with each custmer (here represented by ‘n’  , ie the number of years of the relationship);

The higher these factors are, the higher the Customer Lifetime Value is;  conversely, the CLTV decreases when the cost of acquisition and retention costs increases (SAC and SRC).  According to this formula, it is clear that to optimize the overall value of  a Groupon offer, we need improve the CLTV of customer buying the coupoun, i.e. we must take all the actions influencing the specific value drivers affecting the CLTV, such as, for example,  the retention rate or  the profit margin per single transaction. This is not always done: even experienced merchants do not always takes proactive actions to stimulate recurring purchases across deal buyers. Focusing on the CLTV formula helps devise and prioritize best strategies for value maximization. The formula is also paramount because it allows to answer our initial question: how do we find out the maximum value we should invest into acquiring customer before this destroys value? If we assume for simplicity that SRC = 0 (no retention costs), we still extract  value out of a long term relationship unitl our CLTV is greater than zero. For this to happen,  the SAC must be smaller than the net present value of the profits from purchases. Hence,  the lower is our profit margin on transactions, the lower will be the cost of customer acquisition we can sustain. In other words, the less we get from a single purchase, the more risky is for us to incur large upfront cost to convince a customer to buy  the first time before recovering the investment thoughout the promise of future repurchases. This explains why merchants using Groupon do need to work on this formula to get experienced with the key performance indicators:  profit margin, return rate and  transaction frequency. It also helps to achieve a clear understanding while dealing with Groupon salesforce for a successful negotiation of the deal term and conditions: improving negotiated fees of a few percentage points means changing the cost of acquisition, potentially turning a Groupon offer from an economic failure to a profitable marketing strategy.

The forumula behind the CLTV is enough for the preliminary evaluation ; however, in reality its application is possible if all the cash flows are due only to transactions from newly acquired customers. We will see in Valore Groupon per commercianti: modello NPV di offerta that a realistic Groupon offer is too complicated to be modeled wiht the simplified, albeit effective, CLTV formula shown here.  Still many of the suggestion for merchant to improve Groupon value discussed in Groupon: decalogo pratico di valore per commercianti can be inferred simply by reading the formula above.

Addendum: the average customer lifetime duration can be calculated as the reciprocal of the churn rate (n = 1/cr)

Published by Carlo Arioli