Optimizing Your Customer Lifetime Value Calculation: SaaS Companies

By Eric Mersch

3 Methods for Calculating Customer Lifetime Value and Choosing the Right One for your SaaS Business

Executive Summary

The Customer Lifetime Value to Customer Acquisition Cost (CLTV / CAC) ratio is one of the most important SaaS metrics, but it is not designed to fit all companies. Calculating CLTV is particularly difficult because the most basic formula can be too simplistic for your business.  In this article, we will look at three separate formulas for Customer Lifetime Value (CLTV) providing flexibility in calculating CLTV for your subscription company. By understanding each method, CFOs will be able to produce the most appropriate unit economics for their financial reporting.

Overview: Three Methods of Calculating CLTV

The most basic CLTV formula is simply a function of the average customer’s gross profit and the expected lifetime of that customer. You would use this calculation for customer cohorts with no expectation of expansion and constant churn rates.

The second method applies a more complex formula that incorporates customer expansion over a given period of time. We express Expansion using a dollar-based, linear expansion amount per period, while churn decays per a constant exponential rate.

The third method shares the components of the first two but uses two new variables that account for non-linear expansion and variable churn rates. Additionally, we apply a discount rate to account for the future uncertainty of any forward-looking calculation.

CLTV Formula Variables and Definitions

Before we get into the formulas, I want to take you through the individual components that we will encounter in this article. I find large variations in component definitions from company to company with differences largely due to complexities inherent in business models. Sometimes the formula is not well understood, and variance arises from individual interpretation.

Average Revenue per Unit (ARPU), in $’s – We define ARPU as the average revenue per unit, with unit defined as a set of customers, over some period of time, which is typically one month. Additionally, we define the set of customers as those that we acquire in one month. This is known as a customer cohort. Therefore, the ARPU is the average revenue per customer cohort per month. You will also see this metric termed Average Revenue per Account, or ARPA, which Account defined as a customer cohort.

Subscription Revenue is the recurring revenue derived from an individual customer. This definition is specifically limited to recurring revenue earned from software subscriptions. As such we do not include Professional Services revenue because the margin profiles are so different and professional services revenue comes from one-time, project-based work. SaaS revenue generates high gross margins, typically percentages in the high 70’s and low 80’s. Professional services revenue typically generates gross margins in the 40% range. Often, this revenue is heavily discounted to drive new customer acquisition and this can drive margins negative. So, to include gross profit from both revenue streams would skew this metric and prevent proper benchmarking.

Gross Margin, in Percentage – Gross Margin is percentage of revenue remaining after subtracting Cost of Revenue or COR (often mislabeled COGS, which should be used for eCommerce and other sectors). We define Cost of Revenue expenses as those associated with the hosted software service delivery and categorize these into three segments as follows:

• Hosting and Infrastructure– This is the term I use in my chart of accounts for expenses associated with cloud computing and datacenter operations as well as telecommunications connectivity. SaaS companies relying solely on third-party cloud provides, such as AWS or Azure, pay for cloud computing, storage and bandwidth. Whether the cloud is hosted internally or externally, companies will likely have built proprietary software required to run the company’s platform. Software developed internally should be recorded as prescribed by IAS 38 and amortized when this software is placed into production.  Subsequent updates will add to amortization schedules when placed into production. Amortization expense should be included in “Hosting and Infrastructure.”
• Customer Service / Support– Customer Support is the technical support function for customers’ questions on system access and functionality. In early-stage companies, the Customer Success and Engineering teams may be involved in directly supporting the customer. These teams engage when the customer has specific technical needs above the skillset and/or capacity of the Customer Support teams. For example, a software engineer may need to help a customer may experiencing a software error if that customer has a customized version of the product. If this becomes a recurring need, then a portion of that engineer’s time should be allocated to Cost of Revenue. As another example, you may have particularly advanced users who require Customer Success team members who have more experience with the software than those in Customer Support. Again, if this is a recurring need, then you should allocate some of the Customer Success expense to COR.
• Third-Party Services – Companies also employ third-party software for monitoring performance, which is typically referred to as Application Performance Monitoring (APM) and includes providers such as AppDynamics (Cisco); Datadog APM & Distributed Tracking; Microsoft Application Insights; New Relic APM; Splunk Enterprise and IT Service Intelligence; and Stackify Retrace, among many others. These types of software directly support your customer solution and should be classified as Cost of Revenue. Security is always paramount and every single SaaS company will use cybersecurity security software that supports the real time service delivery. Kenna Security is an enterprise SaaS company that provides real-time penetration testing; its associated expense should be in COR. Companies engaged in commerce will use third-party payment processing companies such as PayPal, Chargebee Clover and a host of others. Some companies incorporate external data sources into their product. For example, Salesforce purchases news content to enhance its CRM platform.

Unit or Account – Defining the Unit or Account for your analysis is more complicated and will vary by pricing strategy. Most SaaS companies sell an enterprise app to customers with pricing based either on a flat fee or variable basis. For flat fees, the atomic unit is number of customers paying that monthly fee. Variable pricing methods are often based on user count with rate cards specifying pricing tiers for users. In such cases, the atomic unit could be users or enterprises. For such cases, we use ARPU for individual users and ARPA for the enterprise purchasing the seats for the users. However, you should pick one of the other for reporting purposes. In deciding, consider your go-to-market strategy. If you are selling to the enterprise, then use ARPA; if to users, go with ARPU.

One common use case to consider is one in which a SaaS company sells to enterprises with large number of businesses. The company sells to a parent company or corporate for implementation in one or more business units. And, in this case the unit can be the parent company or the locations. Examples of such customers include retail companies with multiple locations and large corporations with lots of business units. There’s a tradeoff here: you can show a low customer count with actual or forecasted expansion; or, a high customer count with no expansion. Again, it depends upon your customer definition. If you sell to the parent company or corporate, then count these as your customers. Otherwise use count locations or business units. In my experience, SaaS companies in this situation count the parent/corporate as customers because they want to show the powerful expansion lever in driving revenue.

Another example comes from a work management and collaboration company, Smartsheet, which is a B2C SaaS company that serves small teams of workers. The team’s decisionmaker is responsible for the purchase process so Smartsheet developed an ARPU definition to match its customer behavior. It defines customers as either a domain-based account, i.e., a unique email domain name such as @flgpartners or an Internet Service Provider account, a Gmail, Outlook or Yahoo account in use by a team. When Smartsheet went public in Q2 of 2018, its APRU, or average number of users per domain, was only 1.2. Over time, Smartsheet was able to show growth in this metric demonstrating its success in growing ARPU. Most recently, the company moved to a set of three metrics defined by ACV price points of $5,000, $50,000 and $100,000. This change was designed to help investors see the company’s product adoption by enterprises.

Customer Lifetime, in months or years– Customer Lifetime is the average customer term which, in turn, is defined as the actual plus projected experience.

Discount Rate – We will incorporate discount rates into the more complex LTV calculations when we need to account for the present value for future cash flow. In corporate finance, we would normally start with the Weighted Average Cost of Capital, or WACC, formula. However, startups typically use non-liquid debt, i.e., debt that can’t be traded, WACC equals the cost of equity, which we calculate using the Capital Asset Pricing Model, or CAPM,  as follows:

For CAPM, we need the Risk-Free Rate, beta, and the expected market return. The U.S. 5- & 10-Year Treasury Note rates serve as proxies for the Risk-Free Rate. Currently, the rates are 0.42% and 0.97% respectively. We will use the midpoint of 0.70%.

For beta, I like to use the Bessemer Venture Partners Cloud Index, which tracks the performance of 56 publicly traded companies, and NASDAQ performance for my regression analysis. Currently, this approach gives a Beta of 1.328.

The best source for the expected market return is the Cambridge Associates research on US Venture Capital returns. The most relevant report is the “US Venture Capital – Early Stage Index” which, as of June 20, 2020, shows a 25-year return of 36.22%.[1]

Putting this all together gives us a discount rate of 47.87%.

Three Methodologies for Calculating Customer Lifetime Value

With knowledge of our input variables, we can move onto the formulas. As discussed, we will start with the most basic CLTV formula and add complexity for expansion and churn rates as well as applying a discount rate to adjust for the time value of money.

The Basic Method

The most basic CLTV formula is simply a function of the average customer’s gross profit and the expected lifetime of that customer. Use this calculation for customer cohorts with no expectation of expansion and constant churn rates.

The Expansion Method

The second method applies a more complex formula that incorporates customer expansion over a given period of time. We express Expansion using a dollar-based, linear expansion amount per period, while churn decays per a constant exponential rate.

It is important to note that the expansion variable is in dollars and not percent. Also, make sure that the expansion and ARPU variables use the same period.

Non-Linear Expansion & Churn Method

The third method shares the components of the first two but uses two new variables that account for non-linear expansion and variable churn rates. Additionally, we apply a discount rate to account for the future uncertainty of any forward-looking calculation.

The k-factor takes into account the discount rate and is calculated as follows,

Summary

I’ve worked with several startup companies that experience difficulty in calculating Customer Lifetime Value for use in their unit economic analysis and reporting. Differences in business models, go-to-market strategy, competitive environment and company stage makes calculating CLTV difficult. With these more detailed models, CFOs can achieve more precise CLTV/CAC metrics that best illustrate the company’s performance. And more precise metrics allow companies to make better use of benchmarking leading to better decision making and, ultimately, company valuation.

[1] Cambridge Associates LLC US Venture Capital Index® as of June 30, 2020, https://www.cambridgeassociates.com/wp-content/uploads/2020/11/WEB-2020-Q2-USVC-Benchmark-Book.pdf