Every day, we hear people using the term Lifetime Value (LTV) and meaning different things – so we have written this guide to clearly define what LTV is, why it’s important, and how you can use it.

In today’s world, generalization is the opposite of what customers desire and expect. You cannot assume that all customers want to be treated equally with the same level of service and product offerings. This would be unrealistic. Different customers generate different levels of revenue for a business. That’s a simple fact.

Understanding and measuring LTV is what can really set you apart from the crowd. And when you truly understand it and how it can benefit your business – that’s when you’ll see opportunities for your existing and new customers in a whole new light.

Defining LTV

LTV in its simplest form is a measure of a customer’s actual or likely value to your business. The three most common versions of lifetime value are:

  • Total spend (total spend to-date)
  • Predicted value (future potential spend)
  • Past and future value (spend in the past and predicted future)

For each of these measures, they can also be time-bound and expressed as total sales or sales less acquisition and product cost.


So, why is LTV important?

LTV is a whole of business measure and provides utility to a broad range of focus areas, including:

Insights

Researching and identifying high-value customers will provide critical insight into who you should be serving. And in turn, understanding low-value customers also provides insight into business focus and resource allocation.

Product

Understanding the product preferences of your best/highest value customers provides direction to future design and service offerings – this will yield benefit in the long term, rather than time-consuming, short-term profit strategies.

Acquisition

Targeting prospects that look like your best customer will yield lower CPA’s. Once you recognize the characteristics of this segment of customers, it will be much easier and faster to identify and attract them. And the longer a customer stays with you, the lower the CPA.

Growth

It is more costly to gain a new customer than to retain an existing one. By focusing on growing business from existing customers will ensure repeat purchase and reactivation of customers already in your reach.

Cost

Recognizing your most profitable customers based on LTV minimizes wasted resource and spend. You can then flex marketing and service costs to your most profitable segments, building a group of highly engaged, brand evangelists.

A long-term customer is of more value than a single-deal customer

Larry Myler at Forbes

What you need to know

When looking at the ways to measure LTV, you can break it down in one of three ways. With all options, it’s critical to understand how they are calculated, and what you should be considering when using either metric.

1. Total spend (looking to the past)

How is it calculated?

Total orders – returns

By summing the value of all a customer’s orders minus their refunds.

Benefits
  • Total spend will give you a measure of fact – it is the most real.
  • It is simple and fast to calculate.
  • It is easy to compare against other information like time since last purchase to understand inactive high-value customers.
Limitations
  • Total spend is backward – looking at the past, so it doesn’t tell you about future value.

Read how a luxury skiwear brand was able to dramatically increase their revenue per email by focusing on acquiring customers that “looked like” their high-value customers.

2. Predicted value (looking to the future for potential spend)

How is it calculated?

Predicting future spend can be performed in two ways

1. Aggregate historical prediction

Average spend per year multiplied by the average customer lifetime.

This highly simplified method is implemented by identifying the average tenure and spend per year for all customers. Armed with these two data points you can then predict how much each existing customer is likely to spend. For example: If customers spend $40 per year and stay with you for 4 years you can predict a two-year-old customer is likely to spend another $80 with you.

2. Individual customer prediction

Apply a machine learning model to predict each individual customers probability of re-purchase. Multiply this probability of re-purchase by their average re-purchase price and adjust for seasonality & growth.

This highly analytical approach is capable of producing very accurate predictions at an individual customer level. The production of the model is most commonly done on more than three years of historical data so you can hold out the most recent 12 months of data to assess the accuracy of the prediction.

Benefits
  • Predicted LTV considers the future and doesn’t assume that historical high-value customers will continue to contribute.
  • Knowing a customers likelihood to spend helps you focus on those customers who are most likely to yield a positive ROI.
  • Predicted LTV can be recalculated as and when more data becomes available – more data means improved accuracy.
Limitations
  • Requires more sophisticated data infrastructure and IT resource.
  • Can be inaccurate in a business undergoing change – past behavior may not be a reliable indicator for future behavior.
  • Accuracy is dependent on the available data – limited data can result in lower accuracy.

3. Lifetime value (referring to the past, and looking at the future)

How is it calculated?

Past + predicted spend (using two methods above).

This is where we calculate everything they have spent in the past, along with predicting future spend (using an aggregated model).

Benefits
  • This will give you a very complete view as you’re calculating everything from the past, along with the prediction for the future.
  • Greater knowledge and understanding of future expectations means greater and more intuitive planning can be done.
  • This holistic view allows for focus on the long-term health of your customers.
Limitations
  • Predictive modeling for the “future” component requires sophisticated infrastructure and resource.
  • Again, the accuracy of future predictions is based on the data available – so limited data can have limited accuracy.

To successfully implement LTV, we recommend that you:

  1. Assess your company’s technical capability to generate the measures.
  2. Identify the teams and use-cases for the measures.
  3. Implement an agile test-and-learn approach.
  4. Where successful automate and make it accessible in your day to day workflow.

To find out more about how Lexer software solutions can help you deliver incredible customer experiences get in touch for a demonstration or a chat!

If you would like to read more articles like this one we also wrote an interesting new piece explaining 1st, 2nd and 3rd party data, and published this playbook on how to build a data-loving culture.