Become a better investor
Lesson in Course: Medes Newsletter (advanced, 11min )
At Archimedes, we are building a Software as a Service (SaaS) business, and below I will touch on why SaaS companies are more valuable per dollar earned compared to non-SaaS companies.
Software as a Service broadly categorizes a swath of businesses with a unique delivery model that’s emerged from the 2000’s to today. Some big ones most people have heard of are Salesforce, Adobe, and Slack. What sets these companies apart from the other tech giants like Google, Apple, or Microsoft is that these SaaS companies do not produce any physical products and the software is sold as service rather than singular licenses. The products are strictly written in code and the value delivered to customers is a discrete problem solved through a software interface. Salesforce enables sales teams to collaborate and share information on deals, Adobe allows creative teams to collaborate on projects, and Slack lets a team communicate and collaborate at work.
Outside of a software interface, the economics behind the business greatly set a SaaS business apart from a non-SaaS business — a true SaaS company can be very profitable and thus very valuable.
Four key differences that set SaaS companies apart are predictable revenue, low scalable production costs, the scalability of distribution, and the ability to charge more for additional products.
SaaS companies often have a subscription-based revenue model. For example, Netflix charges $12.99 a month for streaming services. As long as we remain customers of Netflix, the company knows they will receive $12.99 from each of us every month. In contrast, Apple is not a SaaS company. We may buy a new iPhone for Christmas; however, Apple can’t accurately predict when we will buy a new phone again. Having predictable or recurring revenue helps the business predict cash flows and budget better.
Another reason why recurring revenue is better is that it’s much easier to keep an existing customer paying than to acquire a repeat purchase from a customer many times over. The amount of marketing and sales spend decreases when companies don’t have to chase the repeat purchase.
The production cost for software companies is quite low. There aren’t any raw materials that need to be assembled to create software. In large, the cost for the creation of software is summed by the payroll expenses for the product team developing the software, which includes product managers, designers, and software engineers.
Unlike an iPhone, which requires parts and assembling for each unit sold, once the software is created the additional cost to serve the first customer to the millionth customer is very low because the software does not need to be recreated for each customer. Every new customer will contribute to the profitability of the company in a very meaningful way.
The way software is distributed to additional customers also scales very well. Software does not require shipping or retail costs to get the product into the customer’s hands. Anyone with internet access can have access to the product and any updates or replacements can be deployed instantaneously without requiring the customer to return anything.
An additional benefit of software companies is the ability to add features or services and charge customers more for them. As Netflix produces more content, the company increases the subscription cost. AirBnB created another revenue stream by adding Experiences to the core platform. These value-adding improvements to the product expand the current revenue collected from a customer.
Now that we have looked at certain advantages of SaaS businesses, how can we identify the difference between a good SaaS business and a bad one?
There are many metrics looked at to compare SaaS businesses. Below are four simple ones that I think are helpful to gauge the health of a SaaS business.
Annual Recurring Revenue or ARR is the lifeblood of a SaaS company. How much recurring revenue the company makes in a year shows how much the current market values the product. Acquiring new customers or increasing services to an existing customer will increase the ARR. Healthy SaaS businesses will have a steady growth of ARR and will limit the amount of non-recurring revenue collected.
Note, recurring revenue can be reported on a monthly basis instead of annually. Monthly recurring revenue is known as MRR and generally speaking 12*MRR = ARR.
Churn is the measurement of customers lost and is one of two reasons why a company sees a reduction in ARR. When a customer leaves for a competitor or decides that they are no longer willing to pay for the service, they are considered to have churned. Many investors and startup operators have done the arithmetic around acquisition and lifetime value of customer cohorts and the consensus is for a good SaaS business, customers churned should be 10% or below the number of customers signed up.
Expansion, mentioned in the section above, is the ability of the company to increase the revenue collected for additional services provided. When a company has successfully solved a problem for a customer, it is far easier to solve a separate problem for the same customer than to go acquire a new customer. A company with strong expansion revenue has a loyal and strong customer base.
Contraction is the opposite of expansion. Contraction happens when customers downgrade their service level and elect to pay less for a decrease in services. The customer has not yet churned but has decided that the product offerings at a certain level aren’t worth paying for anymore. Contraction can happen for reasons not pertaining to the business, such as the customer had to slash their purchasing budget for the year, or in other cases, a direct competitor to the business is starting to offer services at a discount or for free to the customer that downgraded.
Net Dollar Retention (NDR) is the revenue of a group of customers retained after a period of time once expansion, contraction, and churn is factored in. NDR is calculated by taking the revenue collected at the end of a period divided by the revenue at the start of the period of the cohort (the same set of customers). Equations below for the mathematicians:
NDR = End of Period ARR / Beginning ARR
— or —
NDR = (Beginning ARR + Expansion ARR - Contraction ARR - Churned ARR) / Beginning ARR
Let’s step through a working example. Let’s say we signed up 10 customers paying us $10 a month. That is $100 a month total or $1200 a year in recurring revenue for the company. Over the course of the year, we convince 4 customers to upgrade to the $15 a month plan ($5 of expansion a month or $60 of expansion a year) and unfortunately, we lose 1 customer (churned recurring revenue of $10 a month or $120 a year). Our end of period ARR will now be ($1200 + $60*4 - $0 - $120) = $1320.
NDR =$1320/$1200 = 1.1
The resulting NDR is often expressed as a percent (e.g. 110%). An easy way to think about it is if customers on average paid $100 to the company last year, an NDR of 110% means they are now paying the company $110. Note, this is a useful simplification because in real case scenarios the math may not have worked out as cleanly as the working example we have above.
I consider NDR a good reflection on the current health of the business because it tells us how well the business is serving customers signed up. As long as NDR remains over 100% this means that the customers are becoming more valuable overtime for the business and that expansion is offsetting churn and contraction. If NDR starts dropping below 100% it means that the business has to acquire new customers just to maintain existing revenue numbers.
From Tomasz Tunguz
Paying attention to how NDR changes over time gives us an interesting look at the competitive landscape. If NDR continues to rise over time, this often means that the SaaS business continues to face little competition and continues to create more value for its customers. On the contrary, if NDR decreases over time, this can often be an indication that profits from upsells are being competed away and competitive product offerings are also inviting churn.
An interesting example is in the case of Box Inc. and the increased competition from Microsoft, Google, and Dropbox. From Box Inc’s recent 10Q report: “With the introduction of new technologies and market entrants, we expect competition to continue to intensify in the future.”
Box Inc’s NDR in 2013 = 144%
Box Inc’s NDR as of Sept 2019 = 106%
Let’s take a look at three recent IPOs, Zoom, and DataDog, both SaaS companies with recurring revenue, and let’s compared them to Pinterest, a non-SaaS company that primarily brings in nonrecurring advertising revenue. We’ll compare the revenue multiples, which represent how investors value the companies compared to the companies’ ability to generate revenue — the higher the multiple, the more valuable the company is and the more expensive the stock is.
Recurring revenue: $60.8M in Jan 2017, $151.6M in Jan 2018, $330.5M in Jan 2019
Net Income / Profit: $7.6M
Market Capitalization: $19.42B
Revenue Multiples LTM: 58.7X
An NDR of 140% is fantastic. This means that on average customers are paying 40% more now than when they signed up. The revenue growth is quite strong with the figures more than doubling year over year. Interestingly, Zoom is one of the few recent IPOs to be profitable. While $7.6M isn’t necessarily anything to write home about, it shows that the growth Zoom was able to achieve doesn’t require excessive spending. The backward-looking multiple, calculated by taking the market capitalization dividing by the reported ARR in Jan 2019, is quite high and suggests a lot of investors are excited for the future for Zoom. For context, Lyft and Uber were at 10.9X and 7.3X at IPO.
Recurring revenue: $100.8M in Dec 2017, $198.1M in Dec 2018
Net Income / Loss: -$10.2M
Market Capitalization: $10.45B
Revenue Multiples LTM: 52.8X
An NDR of 146% is slightly higher than that of Zoom and it shows the possible growth in the business. The revenue growth is also on track to double year over year. However, unlike Zoom, DataDog is not profitable. A loss of $10.2M compared to other companies looking to go public is really not that bad (WeWork posted losses of $1.6B in 2018). The difference in profitability here may be a reason why the multiples or valuation being assigned to DataDog is slightly lower compared to Zoom.
NDR: N/A (Not tracked because Pinterest doesn’t collect revenue from their users)
Revenue (non recurring): $472.9M in Dec 2017, $755.9M in Dec 2018
Net Income / Loss: -$63M
Market Capitalization: $14.00B
Revenue Multiples LTM: 18.5X
Pinterest makes money largely based on advertisements shown to its users. While advertising revenue isn’t recurring, Pinterest has shown strong year-over-year growth in the ability to make money. With a net loss of $63M, investors are still optimistic about the future of Pinterest and have rewarded the company with a backward multiple of 18.5X. Whereas Facebook, another social media company with similar a monetization model, is only at 8.4X.
It’s interesting to see that Zoom and DataDog have revenue multiples three times higher than that of Pinterest. While we can’t say this is solely based on the former being SaaS companies, a magnitude of three seems to strongly suggest that the public investors are liking what they see in the business fundamentals.
We’ve seen above that both SaaS companies have strong revenue growth as well as great NDR, which will lead to even further growth in ARR in the future. As we know by now, the economies of scale are favorable for software companies and as these companies continue to serve more customers, profitability should also improve.
Both Zoom and DataDog are also two data points that dispel the fear of the IPO market weakening. Rather, it seems like public investors are demonstrating rational decisions by being hesitant about companies that lose $1.6B a year, and being selective in their investments for companies that create more value for every single dollar earned. There is still a strong appetite for businesses with strong economics and sustainable growth as shown by the IPO successes covered in this post. I think all will be fine in the startup ecosystem as well.