Are you thinking about software companies’ metrics correctly?

Software is eating the world – probably the most overused and least contested phrase in today’s investing world. Software companies have had a great run this year with the likes of Twilio and Datadog returning over 200%.

John Huber of Sabr Capital markets wrote recently that a strange dichotomy exists in the stock market right now. Any stock labelled with “SaaS” has been touted as the next big thing, and investors are certain that these companies are bound for success in the future.

Indeed, the Covid 19 pandemic has catalysed the secular tailwind for technology and software companies. While I do not refute the notion that “Software is eating the world”, I find it difficult to come to terms with the market implied speed of how fast software is going to eat the world. In essence, I believe that the market is right about the direction but not so much in terms of magnitude.

While humans have to be inherently optimistic to have evolved to where we are today, excessive optimism leads to unrealistic projections of the future without considering the unknown unknowns that lie ahead. The history of the auto industry is an example we can draw parallels with, especially in the early 1920s where over optimistic auto executives projected for car sales to top 6 million annually by the end of the decade. That level was only attained in the 1950s.

It is inevitable that some software companies will fail in the future. The key here is to avoid those companies and try to identify high quality software companies that have the highest likelihood of thriving in the long run. Understanding key metrics is not the holy grail but it will guide investors to assess the health and growth prospects of software companies.

I will discuss 5 key metrics that I personally find helpful. Aside from the definition and formula of these metrics, I will also talk about caveats I consider when using a particular metric to evaluate a company. I hope readers will focus on the latter as anyone is capable of punching numbers into a calculator. Valuable insights are often derived from an independent analysis of what the numbers really mean.

1. CAC (Customer Acquisition Cost)

Customer Acquisition Cost, or in short CAC, refers to the sales, marketing, and associated costs to acquire one new paying customer. Ideally, we would want CAC to be as low as possible and decreasing over the time horizon. Increasing CAC might be a sign that the company is finding it harder to acquire new customers due to a saturated market, unsatisfactory product quality etc…

CAC = Sales & Marketing Expense / No. of new paying customers over same period

Blind spots

CAC is a relatively straightforward metric when it comes to calculation. However, there are numerous caveats that one must consider. I will list two for consideration below:

CAC is not reflective of the total costs to acquire a customer.

Dropbox for instance operates on a freemium model, meaning that a new user starts out free, and subsequently chooses to upgrade after running out of storage. This might take months, years or never. Costs involved in supporting free trials and freemium products are not accounted for in the typical CAC calculation. In a similar vein, not all customers are new customers, but some are returning customers.

A metric, known as CPA (Cost per Acquisition), introduced by Andrew Chen attempts to address this. To summarize, he points out that an honest assessment of customer acquisition costs distils between new and returning customer, looks at the length of sales cycle, and considers total costs and resources needed to support marketing efforts to acquire customers. You can read more about it here.

CAC differs across the user acquisition curve.

At the beginning, loyal enthusiasts and fans are the cheapest and easiest to acquire. As a company progresses further out along the user acquisition curve, it becomes increasing difficult and expensive to acquire the late joiners. It is pertinent to consider the user acquisition stage before comparing two companies’ CAC.

2. Churn Rate

A vast majority of software companies operate on a subscription-based business model.

Understanding customer churn rate helps us to determine the percentage of customers who have cancelled their subscription over a specified time period.

No. of Customer lost over a period /
No. of Customers at the beginning of the period

Alternatively, customer churn rate can be also calculated via (1 – Retention Ratio). Customer churn rate gives us an idea of the stickiness of a company’s product. It also offers a glimpse into the average customer lifespan that is given by the following formula:

Average Customer Lifespan = 1 / Customer Churn

A low churn rate is preferrable as companies do not have to spend to acquire new customers to make up for lost customers. Moreover, a long customer lifespan also translates to a predictable revenue stream due to the recurring nature of a subscription business.

Blind spots

Analogous to CAC, there are certain blind spots that customer churn rate fails to account for.

First of all, a high customer churn rate can be interpreted as a positive signal under the right context. Shopify for example, probably has a pretty high customer churn rate. A high customer churn rate probably indicates that Shopify has an easy to use platform that has attracted a massive number of new users, which eventually flame out. By getting more users to the top of the funnel, it increases the chances of Shopify riding on a serendipitous path to success. Compare this with the “intuitive” claim that that Shopify’s high customer churn rate is an indication that customers are growing increasingly dissatisfied with their products. Ben Thompson, author of the blog Stratechery summed it up perfectly:

“To that end, I would argue that for Shopify a high churn rate is just as much a positive signal as it is a negative one: the easier it is to start an e-commerce business on the platform, the more failures there will be. And, at the same time, the greater likelihood there will be of capturing and supporting successes

Moving on, another glaring issue with customer churn is that it is not linearly related to revenue loss. Therefore, it is important to distinguish between Customer churn vs Revenue churn.

Revenue churn – measures percentage of revenues lost over a specified time period.

Formula: Revenue lost over a period / Revenue at the beginning of the period

When taken in context with customer churn rate, revenue churn gives you a sense of the customer segment that has churned. Effectively, it tells you if your big customers or small customers have churned. This is especially important if a significant portion of a company’s revenue comes from a concentrated number of customers. A prime example is Fastly where revenue projections were slashed amidst fear of lower than expected revenues from a single, but also their biggest customer, Tiktok. Tiktok accounted for 12% of revenues in June 2020.

3. LTV (Lifetime Value)

LTV can be measured in terms of revenue or gross profits. Simply put, it is how much total revenue or gross profits you can expect to earn from the average customer. For the purposes of this post, we shall assume that LTV refers to revenue. In order to calculate LTV, you will need the average customer lifespan and ARPU (Average Revenue Per User)

 Average Customer Lifespan = 1 / Customer Churn

ARPU (Average Revenue per User) = Revenue / No. of Users in the same time period

With this, you can now calculate LTV.

LTV = Average Customer Lifespan * ARPU

It goes without saying that a high LTV is desirable.

Blind spots

However, take note that ARPU may not be representative of what a true customer spends because it assigns equal importance to all customers in terms of revenue contribution. In reality, Pareto’s 80-20 rule is a more pragmatic way to look at revenue contribution. It is more likely that 20% of the customers account for 80% of revenues. Just like my discussion on Atlassian in my previous post (read here) – they reported 174,000 paying customers but a mere 3.38% of customers spend over $50k annually.

4. Unit Economics:

Unit Economics refers to the ratio of LTV to CAC. In simple terms, it helps us to determine how much more money is a company is able to earn from a customer compared the cost to acquire the customer. It can be calculated via:

Unit Economics = LTV / CAC

Generally, a rule of thumb is that a LTV value of 3 signifies a healthy business. That means that the company is able to make at least three times the revenue from a customer relative to the cost it incurred to acquire that customer. A low Unit Economics number could be a warning sign that the company is spending too much to acquire new customers, executing with poor pricing strategies, failing to monetize existing users or simply has an inferior product.

Since Unit Economics is derived from LTV & CAC, all pitfalls relating to LTV and CAC have to be considered and understood in the context of the particular business before making a thoughtful evaluation of this metric.

5. DBNER (Dollar Based Net Expansion Rate)

Fastly, a CDN (Content Delivery Network) company described DBNER succinctly in their 2QFY19 financials:

Dollar-Based Net Expansion Rate (DBNER) measures the change in existing customers’ revenue from usage of our platform over a twelve-month period, excluding the effect of new and churned customers.”

 DBNER therefore, can be calculated using the formula below

Revenue from customer base this current 12M period /
Revenue from same customer base in previous 12M period

Customer base includes customers who remained customers as of the last day of the current 12M period but excludes new customers that have signed up in the current 12M period

In simple terms, DBNER shows how much more a customer that stayed with the company is spending, compared to the previous year. For instance, Okta, a software company posted a DBNER of ~ 120% over the past 2 fiscal years. This indicates that customers who remained with Okta have been spending 20% more than what they did in the previous 12m period.

A high DBNER is desirable as it demonstrates the ability of the company to expand use cases within their existing customer base. Through development or acquisition of new products/features, a company can expand use cases laterally and vertically within their customer’s organization. Okta for instance, started with workforce identity management before expanding into neighbouring vertical of customer identity management.

DBNER can also be used to assess how a company’s “Land and Expand” strategy is faring. If you are unfamiliar with the “Land and Expand” strategy, you can read more about Atlassian’s “Land and Expand” strategy here

Conclusion

The biggest issue that I have with such metrics is that it relies heavily on simple averages when the world simply does not work in that way!

Averages are easy to understand and simple to calculate but we must be very careful when interpreting the numbers. Averages hide variation and are easy to distort especially when there are outliers or that the underlying data is skewed. Sometimes, studying the outliers might reveal valuable insight.

However, without access to an inside view of a company, the metrics above are the best benchmarks that investors can rely on to evaluate the growth of a software company. The imperative lies in understanding the context behind each metric and in relation to other metrics. Shopify is a great example – a high churn rate as a positive signal seems counter intuitive but is exactly what we want. (Read the above section about churn rate if you’ve missed that).

Consumer style metrics such as DAU (Daily Active Users), session length, active user retention rates will be a great supplement to the metrics discussed above. However, these are information that companies are not obliged to disclose. On a similar note, companies are also not obliged to report CAC, Churn rate, DBNER metrics in their financials. A tip is to refer to investor relations materials or earnings calls to scrape for such information.

Each company is unique and therefore, it is essential to think through what a metric is conveying instead of using a one-size-fit-all approach. During my time in the military, there was a common saying “Go through motion only”, which translates to doing something without putting in much thought.

As an investor, you want to avoid “going through motion”. You want to think, and act with purpose.


If you have enjoyed my writings and wish to be updated on future posts, do enter your email address on the “Get Updates” box on the top right hand corner!

You can also reach me at investingcurator@gmail.com.
See you around!

Related Posts

Leave a Reply

Your email address will not be published. Required fields are marked *