Happy Business Starts Here

Zuora Product Team

Part one of a two-part series on the important metric: Customer Lifetime Value (LTV)


It seems like everyday Amazon is adding some new “free” benefit into its popular subscription membership service Amazon Prime -- free videos, free music, free book rentals. The list goes on and on. The same goes for lots of other vanguards of the subscription economy. Google launches a few free apps that make living your life and work just a bit easier; Netflix ships a brand new batch of kids movies that would’ve cost $1000s in Blockbuster rentals back in the 90s; Tesla pushes a few wireless updates and suddenly your car just got better at driving itself. These subscription economy companies are the envy of entrepreneurs everywhere who want to surprise and delight their customers with new and unexpected features and benefits seemingly with a spirit of unbounded generosity.


Such “free” customer benefits are not limited to the most innovative subscription economy companies either. Look no further than your cell phone providers like AT&T and Verizon and their inducements of the newest and most hi-tech smartphones at free or near-free prices when you agree to sign-up for their service for two-years.


What is driving this seemingly endless string of giveaways by subscription companies?


The answer can be summed up in a single, very important metric: Customer Lifetime Value (LTV). Sure it goes without saying that all of these companies care about the experience they provide their customers, but they offer these wonderful benefits as part of the experience through a lens grounded in a set of financial facts around how much value a given customer can create for their shareholders.


This paper provides an in-depth explanation of LTV that will help you execute your subscription business with optimal customer value creation in mind. It is organized into four parts.

  • Part one will look specifically at Amazon Prime and show you that behind all the unlimited access there is an underlying business logic grounded in LTV.

  • Part two will then start to explain what goes into LTV by guiding you through the key concepts that go into a reliable LTV calculation.

  • Part three provides the one precise LTV formula that Zuora recommends above all others to its customers for calculating LTV.

  • And finally, part four will share how Zuora’s technology helps its customers calculate and maintain LTV stats for their subscribers and key subscriber segments.

I. Why Does Amazon Give Things Away? The business logic of LTV


A search through Amazon’s Investor Relations website or published interviews with its CEO Jeff Bezos turns up few detailed numbers about the Amazon Prime program. Amazon keeps secret the real numbers driving Amazon Prime, but a few analysts have, however, taken a close look at Amazon and uncovered some very interesting information about the customer lifetime value of a Prime Member.


At the time of this writing, Amazon Prime is an annual subscription priced at $99/year and a 30-day free trial option.




Brendan Mathews a research analyst at Motley Fool has published his research on Amazon concluding that the average Prime subscriber has a LTV of $2283, more than twice the LTV of a non-subscriber, which he estimated at $916.


LTV Amazon


So while from a customer viewpoint it seems like Amazon is giving an unlimited number of benefits away, getting someone to convert and become a subscriber to the service drives an incredible amount of value for the company. It is nearly certain that whenever Amazon makes investment decisions about sending money to acquire new Prime members, that spend relative to the roughly $2000 lifetime value of a member looms large in the calculus.


II. Conceptual Elements of LTV


Let’s now examine what goes into a comprehensive and reliable LTV measure. Remember, LTV is a measure of how much profit you expect to make from any given customer in the future. It is comprised of four key factors.


LTV Conceptual


Element 1. Life Expectancy


The first element that goes into an LTV calculation is the life expectancy of your customer. Each of your customers will stay with your service a different amount of time. Some, who are dissatisfied and exhibit account health risk factors, may churn this renewal period; others, who are healthy and happy, will go on to renew for many renewal periods to come. In order for customer lifetime value to have real meaning, Zuora recommends you estimate churn rate individually for each of your customers based on their unique behaviors, demographics, pricing, packaging, billing, and payment histories. At Zuora, some of our most sophisticated customers have moved beyond simple predictions based on a few intuitive factors to sophisticated statistical models that predict individual account churn rates by modeling the top factors responsible for past churns. Once you have established a churn rate for your customers, then you can estimate life expectancy according to the following formula:


Life Expectancy = 1 / Churn Rate


For example, a customer whose renewal period is annual with an estimated churn rate of 32% will have a life expectancy of 1 / 0.32 or approximately 3.1 years.


As you estimate life expectancy, you will want to express your life expectancy using a consistent time period that corresponds with your subscription renewal cycles. For example, if your business has contracts that renew annually, then your LTV calculation will estimate how many years you expect to keep each customer.


Element 2. Revenue Expectancy


It goes without saying that how much value you think you will earn from your customer in future periods is largely dependent on the revenue you expect from your customers. To estimate this, you not only need to look at current bookings (or revenue under contract), but you also need to estimate how the revenue you get from your customers is expected to change over time. Do you expect your customer to be upsold, to continue to pay the same amount, or downsell? A good way to estimate future changes in current subscription revenue is to use your Net MRR Retention rate. Net MRR Retention is defined as the dollar amount of your renewals divided by the dollar amount up for renewal. It gives you an aggregate measure of upsells relative to downsells and churns across your entire customer base.


In an attempt to find better alternatives to using Net MRR Retention in the revenue expectancy element of LTV, Zuora is currently experimenting with a few new methods aimed at improving the state-of-the-art by attempting to estimate revenue expectancy individually, by customer. These methods attempt to goes beyond using just standard Net MRR Retention to a more personalized estimation. The idea holds promise. For example, picture the Apple Music service. Apple Music offers three plans.



Apple Music


Inherent in this packaging structure are two predictable upgrade paths (Student to Individual, and Individual to Family). Modeling those upgrade paths into individualized estimates of upgrade or downgrade likelihood could prove quite effective at improving LTV accuracy beyond a wholesale average net retention factor.


Element 3. Cost Expectancy


The third element of lifetime value is estimating how much it costs you to deliver your product or service. Each of your subscription products have a particular contribution margin associated with it that needs to be estimated. Contribution margin represents the variable costs that go into providing your product and is a measure of the profitability of your subscription products.


In the case of Amazon Prime, Matthews estimated that the variable costs of operating the Amazon Prime service at 12%. Some businesses leave out cost expectancy from LTV, but doing so leaves out an important element if you plan to make spending decisions based on LTV.


Element 4. Risk Expectancy


The final element of LTV is estimating how much risk your future revenue streams face. Risk expectancy is critically important because LTV is an estimate of what will happen in the future. Your estimates of the future can be conservative or aggressive. A common mistake that many make in calculating LTV is that they do not sufficiently account for future risks. Zuora recommends you strive to underestimate LTV for customers -- err toward a conservative LTV. Why? Let’s say you overestimate the average customer LTV at $2000 and it winds up being $1500. Let’s say based on the $2000 number, you decided to spend $1750 to acquire a customer. Your overestimate would have cost you $250 per customer instead of earning you $250. When working with LTV, you always want to estimate for a worse case.


How should you estimate for a worst case? First, you first want to discount LTV with prevailing interest rates since a dollar in today’s money will be less than a dollar in the future. Also, in addition to interest rate discounting, you also need factor in risk discounting. Wall Street investors offer a compelling model to follow when thinking about appropriate levels of risk discounting. On Wall Street investments are always discounted by higher rates than just the interest rate. The additional percent discount is a risk adjustment commonly called the “spread”. The further out in the future cash is expected from a customer, the more that could go wrong. Additional risk could include changes in economic conditions, competitive landscape, and regulatory changes. In essence, all reasons that future cash might not be paid by a customer gets rolled into the spread. Spread is roughly the percent change of catastrophic loss per payment period. For those with finance backgrounds this is the very same concept as corporate bond spreads -- high grade corporate bonds have spreads of 1-5% and low grade corporate bonds (junk status) have spreads between 5% and 25%.


Factoring all of this risk together, Zuora recommends you use a conservative discount rate and we often work with our customers to determine the right number for their business.



Tomorrow, in Part Two of this series, I'll discuss the LTV forumula and show how to keep LTV up-to-date and reliable.