How to use Big Data - In marketing, Customer Lifetime Value is future projection of the net profit over the entire future relationship with the client. Some define it as the dollar value of customer relationship based on the present value of the projected cash flows from that customer. Even though the term “lifetime” suggest that the calculated value encompasses entire future relationship with the customer, practical uses are limited to a year and maximum to 3-5 years span. Longer term calculations are just too inaccurate for any practical uses. This is a very useful tool as it quantifies value of the relationship with the customer within a period of time and provides a yardstick for selecting appropriate customer acquisition strategies as well as for choosing optimal services for specific customer value-based segment.
Until you identify exactly how much combined profit a customer represents for the life of that relationship, you can’t begin to know how much time, effort and expense you can afford to invest in your marketing activities and business operations either overall or toward the specific customer or customer segment that he belongs to. Another major benefit is it encourages marketers to focus on long-term value of customers instead of investing in “cheap” customers with the low value.
When you hear words from business owners that “there is no marketing budget for that” - it means that they are in a dark in terms of how many customers it will take to pay for specific business investment. Conversely, once you know Customer Lifetime Value in your business you take guess-work out of the equation and you no longer have to worry and wonder about what you’re spending on marketing, so you can budget with complete peace of mind.
Development and retention of the most profitable customers should be vitally important to any commercial organization, and long-term revenue goals should not be replaced by the activities that are aimed on to generate revenue and profits on short term basis. Value and profitability are dynamic constructs.
Often the most current profitable customers are in last stages of their careers and may soon retire and not even be around 5 to 10 years in a future, so business needs to replace them by the customers who have high value potential but they are not there yet. Another similar concept is customer profitability which is defined as the difference between the revenues and the costs associated with the customer relationship during a specified period. Main difference is that customer profitability measures the past profitability by calculating and reporting customer actions within some period of time, and as such it is relatively easy to calculate in comparison to more useful but also far more difficult to quantify CLV as the forecast of future customer value.
Formula to quantify CLV is quite simple:
Total profit of an average client over the relationship’s lifetime including all sales LESS advertising, marketing and incremental product or service fulfillment expenses EQUALS Lifetime Value of the Customer.
Let’s work with a real-life example to better help you see how to apply this formula:
Some landscaping company charges their customers an average installation fee for their standard irrigation system R120 000. They then charge R250 per month landscaping services for the security monitoring service. The industry average says that people will stay with the same landscaping company for the company for 5 years.
R250 per month X 60 months (5 years) + R12 000 installation fee = R27000, which represents the gross revenue per the lifetime of each customer. Now, this company spends R6000 overhead expenses for marketing, advertising over the same 5 year period (averaged out per customer, of course. Lifetime value for each customer will be R27 000 – R6000 = R21 000.
The company can now see whether or not the revenue they receive from each client is enough to cover marketing and fulfillment expenses at a profit, and they can determine a marketing budget. There are many variations of this basic formula that takes the present value and applies some probability on future cash flows and calculates the discounted sum of future cash flows. Because the present value of any stream of future cash flows is designed to measure the single lump sum value today of the future stream of cash flows, CLV will represent the single lump sum value today of the client relationship. So, if margins and retention rates are constant, this formula can be used to calculate the lifetime value of a customer relationship:
Customer lifetime value = Ave Monthly Spend * Margin ÷ Monthly Churn Rate
(eg: R100 average monthly spend * 25% margin ÷ 5% monthly churn = R500 LTV)
Advanced Analytics / Data Mining Method of Calculating CLV
One of the biggest criticisms of current methods of calculating CLV is that they are fairly inaccurate and therefore should not be used for making important business decisions. Inaccuracy originates from discounting component which is often averaged across all customers and ignoring relationship with specific customers.
Data mining approach starts with Future Cost Component (FCV) equals Current Customer Value (CCV) (Current profit minus costs) to which future component is added. Future component consists of two parts: Risk Adjusted Revenue (RAR) and Affinity Adjusted Revenue components (AAR) minus Future Cost Component (FCC). Risk Adjusted Revenue (RAR) is the sum of revenue for all customer products adjusted for the probability that he may cancel all or some of them. Affinity Adjusted Revenue (AAR) is the sum of all revenue for the products that he does not have yet within his portfolio adjusted for the probability that he may acquire them.
FCV = CCV + (RAR + AAR – FCC)
Advantage of this approach is that uses a variety of other variables to calculate the probability of his canceling specific product as well as acquiring new product therefore factoring in his overall relationship with the company reduced to specific product that he has and the product that he may acquire. Disadvantage is complexity of discounting process. Mathematical models need to be produced for every product (or product grouping) within the full portfolio and each model produces probability for product either purchased or cancelled. As mentioned before - data mining relies on the data that is available.
Early in the lifecycle not that much data is available, especially on soft customer characteristics, such as satisfaction and loyalty. For new consumers, companies should start to build product-related data mining models. With growing knowledge about the customer (established by growing data on the customer from other units such as customer service), models can be combined to start building a single customer model framework that encapsulates all customer dimensions that are relevant for the customer life time value and that includes larger number of customer behavior dimensions (revenue generators, cost generators, lifetime generators). And finally, due to its complexity CLV using data mining needs to be approached stepwise with the goal of building a comprehensive framework of customer behavior models.