Recently a client asked us to propose on building a customer lifetime value model. This wasn’t unusual, we’ve built this kind of model dozens of times, but in the last couple of years we haven’t done as many stand-alone CLV projects and I stopped to think about why that was.
First of all, despite the glowing articles of a couple of years back saying “it’s really quite easy” and “it can drive your whole strategy” we’ve learned the opposite is really true. Figuring out the revenue side of the customer value equation is usually pretty straightforward, it’s the cost side that can really be a stinker. Marketers who aren’t traditional direct marketers don’t naturally calculate the cost to acquire a new customer, and figuring out the “cost of goods” and the “cost to serve” requires input and buy-in from parts of the organization far, far from the marketing team. Finance usually must get involved and cans of worms can often be opened!
Another critical component of the value of a customer is the length of the customer relationship. We are always surprised when we do retention analytics at how difficult it is for many marketers to really understand customer retention and attrition in a detailed way. Given the old saw that “it’s easier to keep a customer than find a new one” you might think that monitoring and understanding retention rates would be a top KPI – but without a good customer database, it’s pretty hard to track with precision.
So Customer Lifetime Value isn’t so easy without the right data, tools and know how. But it is still pretty important to understand who your best customers might be – or those who could become top customers with a little love. And deciding how much to spend to acquire new customers is difficult if you don’t have a long term view of your customer’s potential profitability.
If you have all the data, and buy in from Finance to help you figure out the cost side – great! If not, we’ll look at the relative performance of different customer segments and create some algorithms that will allow us to rank customers in terms of their likelihood to be high or low value. Of course, more data over time will get us closer to a more precise estimation.
But why don’t we do too many standalone CLV efforts? I think because we now bake customer value into so many other analytical efforts to fine tune messaging and sales efforts. A customer value estimate might become one component in an acquisition or retention model. Or, we’ll include customer value in a segmentation scheme along with demographics and lifestyle components.
Today, predicted customer value is less often the end game, and more often one more piece of “smart data” that can work with a wide number of other elements to help us deliver the right messages and offers to prospects and customers – in direct mail, email, digital ad targeting and in the sales process. CLV is one tool in a whole tool box and if it is too difficult to create the perfect equation because of lack of data or internal politics, we’ll create a rough estimate and use it with other “smart data” to help make the most of the customer’s experience with our clients’ brands.