Why Lifetime value (LTV) calculations need Data Science 

Lifetime Value (LTV), sometimes referred to as Customer Lifetime Value (CLTV), is a technique used by businesses to predict the net profit of the entire future relationship with a customer. LTV is best thought of at a high-level as simply Total Customer Revenue – Total Customer Costs. Two key components to recognize and understand regarding LTV are the fact some customers hold more value than others and a customer is not just a single transaction but rather a relationship far more valuable than just a one-time deal.

While the ability to predict value at the customer level is important on its own, LTV predictions become even more powerful when used to analyze and obtain insights on your entire customer base as a whole. Assessing the financial value of each customer and understanding how a customer base operates can provide benefits in every imaginable aspect. Improving an LTV model with powerful data analytics enable you to estimate marketing costs better, analyze customer acquisition/retention/expansion strategies, and improve products and business operations. An accurate view of your business’s LTV and customer base plays an integral role in both calculating the payback of advertising spent and determining the optimal mix of marketing and sales investments. While Marketing learns how much should be spent to acquire a customer, Sales gains an understanding of what types of customers their reps should prioritize, Product teams learn what features can be improved or developed to increase LTV, and Customer Support knows how much time should be spent to retain a customer and the associated return on doing so.

Improving an LTV model with powerful data analytics enable you to estimate marketing costs better, analyze customer acquisition/retention/expansion strategies, and improve products and business operations.

The concept of LTV has been around since the late 1980’s but thanks to the recent surge in data science and analytics, LTV models have been elevated to higher levels of accuracy than ever before. Netflix is an excellent example of a company that has utilized LTV models and data analytics to not only forecast better, but also improve every other aspect of their business. Even before Netflix was the video streaming behemoth that it is today, the company had an insane amount of data on its customers who at the time were renting DVDs. Through careful analysis, Netflix saw that if a customer did not continually rent movies, they were likely to cancel sooner or later. To combat this, they added features such as the queue so that as soon as you sent one DVD back, you would receive the next one on your list.

Taking this analysis on customer impatience a step further is what brought Netflix to where it is today as it led the company to begin streaming movies online and eventually on mobile devices. Online shoe retailer Zappos is another great example of how LTV analysis can empower a business to make successful critical decisions. Zappos found that the customers buying the most expensive shoes were their best and most profitable ones. Analysis also showed that these customers had a return rate of 50%. Realizing this encouraged Zappos to focus intensely on customer service and is the reason why they offer no charge 365-day returns and free two-way shipping. It is clear in both of these examples that LTV analysis enabled the companies to make decisions that led to happier, more engaged customers that would continue paying for products and services for much longer.

As seen in the above examples, the approach taken to LTV and data analytics will be unique to each business depending on the data available and what they are looking to learn from it.

  • Which sources are the most valuable customers coming from?
  • How can we get more value out of the good customers and reduce how much we’re spending on the bad ones?
  • What features of my products can be improved to entice more people to use them and keep the current ones wanting to use them for as long as possible?
  • Which areas of my business can benefit from a more optimized and efficient strategy according to LTV analytics?

Though LTV is an output from the model, the answers to all these questions also begin to materialize when data science is used effectively. These answers then evolve into business decisions which in turn makes them inputs for the LTV model once again. When used in this way, LTV and data analytics feed into one another and create a cycle of continuous improvement that benefits the entire business immensely.

If you have questions about LTV’s contact us and we will be more than happy to talk to you about LTV calculations.

Greg is a Manager of Data Science for Five Element Analytics, an analytics firm based in NY. He graduated from Hofstra University in 2015 with an MBA is Business Analytics.