by Trey Pruitt
For businesses with substantial customer revenue and costs beyond the first transaction, customer lifetime value can be a useful analytical tool to normalize and evaluate all potential investments. CLV supports and helps inform strategy but isn't strategy.
CLV is also just a set of predictions about an uncertain future. These estimates can be useful, but they shouldn't be confused as set-in-stone facts.
To be accurate, CLV must incorporate relevant costs to provide products or services and to support a customer over time. If you are in a high gross margin business like software/SaaS, the mistake may not be as obvious. But in lower-margin businesses (like ecommerce), using revenue alone in CLV calculations can be very wrong and lead to bad decisions.
Many people make the mistake of using only revenue to estimate customer value. Even 2 of the top 5 results in Google for the query "customer lifetime value" make this error (both this mis-infographic from KISSMetrics and this video make the revenue-not-profit mistake).
If your teams are using a single measure for all customers, they are doing it wrong. I suppose it's ok so say that an "typical" customer is worth $237, but don't use that average estimate to make business decisions. That single estimate doesn't reflect the variation across customer segments. For example, a recent study suggested that customers acquired from organic search generates ~50% more lifetime value than average, whereas customers from Twitter were ~20% lower than average. So CLV varies considerably depending on important customer attributes such as acquisition channel.
I sometimes see clients spending 100% (or more!) of first-time transaction profit to acquire a customer. If your team is doing this under the rationale that "we'll capture the profit" in the future since CLV is $237, your CLV model should get a lot of analytical attention and monitoring. This is particularly true if your paid marketing programs consume a large portion of near-term profits.
I sometimes hear comments like "we have the CLV nut cracked --- we just need to keep acquiring customers below that customer lifetime value and we'll have a $100MM business in 5 years." However, CLV ---even within the same channels --- can change significantly over time. Not measuring changes in CLV over time can be a disastrous mistake. You should be asking your teams for trends in CLV, by marketing channel and customer cohort. As the mutual fund regulations suggest: "Past Results Do Not Guarantee Future Performance"!
In summary, customer lifetime value models can be a powerful analytical tools. But to be successful, use them properly and keep the points above in mind.
Other resources on customer lifetime value (CLV):