Greetings from Vertical SaaS land! Crazy to think, we’ve already put out 10 editions of Vertical SaaS, Today. I have so enjoyed evaluating the markets alongside the SaaS community — and hope that some of our analyses and takeaways have helped you as an entrepreneur or investor.
If you’ve ever been a statistics or data science student, you know that you can go down a rabbit hole doing lots of correlations and some of them are dependent on how, and when you cut the data. Which is why, its important to limit the data set and be upfront about limitations and strengths of correlation. For those who have been following vSaaS Today for many months — you know that we are using a public data set from CapitalIQ — based on 19 companies we deem as vertical SaaS leaders with at least 50%+ of their revenue coming from recurring revenue. For this particular analysis, we also used customer counts and estimated ACV — as most of you know, these are often not publicly reported numbers so we had to do a bit of internet sleuthing and estimations. For example, some companies don’t report on how many absolute customers they have — but instead report on locations served and vice versa.
I’ve had lots of entrepreneurs ask me how to scale companies in vertical industries. How do you size the TAM? What counts as a big enough market size? Should you go down market or up market? What constitutes a strong ACV? Does it really matter how many customers you have access to?
As a quantitative thinker, I wanted to go right back to the data to help answer some of these questions. In this month’s analysis, we found a very interesting correlation emerge.
To be a meaningful vertical software company, you need scale.
As you can see in the data above, there is a strong correlation between Market Capitalization and number of customers served. In fact, compared to the correlations we’ve done in the past (read: 7th edition) it is the highest correlation we’ve found at R = 0.718 which is a “Strong Correlation” statistically. Conversely, ACV (or annual contract value) did not have a strong correlation with market cap.
The obvious explanation is: if your revenue is high, you will have a high market cap— but the correlation between revenue and market cap is not as high as the correlation between number of customers served and market cap. In fact, the correlation between revenue and market cap is a “Moderate Correlation” statistically, at best (R = 0.54).
We believe this is because, to be a meaningful vertical software company, you need scale. Scale can often be determined by the vast number of customers that can use your product. In the past, if your company had access to millions of customers it usually meant selling to SMBs, but there are certain companies that prove it doesn’t have to be. Autodesk, as an example, serves 186,533 customers and has a Market Cap of $45.2B, Constellation Software similarly serves 125,000 customers and has a Market Cap of $44B. The major anomaly you see is Veeva Systems, which serves approximately 1000 customers but has a market cap of $34B.
When analyzing vertical SaaS businesses, the ability to get massive scale in your product is more important than large ACVs if the addressable market is not clear. Veeva serves pharma companies which are unlike many other markets in the sheer amount of spend and budget they have, even if it is incredibly concentrated. It is difficult to find an analogous market segment where you can be a $30B plus company while only serving 1k customers.
TAM is not just a number
Not all TAMs are created equal. The easiest way to size an opportunity in venture is through a market sizing — and there are many ways to do this. Some prefer a bottoms up analysis, others prefer top down. You usually want to get to a $1B+ number, without having to incorporate payments to get there. Then you build up from there.
The problem is that most entrepreneurs (and investors!) stop there. If it looks big enough, it probably is. But again…not all TAMs are created equal. That’s why its necessary to see the other players that existed prior to the company today, what the acquisitions have been like, where others have struggled.
Let’s use a tactical example: there are some enormous TAMs in number alone, like oil and gas, which just don’t buy cloud software at the same velocity as other segments — and often opt to build in house which leads to dried up budgets. There are also markets that are highly correlated to interest rates that can weaken immediately after fundamental changes in the market (we saw this in healthcare after the covid tailwind, and are now seeing it more closely in real estate which has faced a number of challenges due to rising interest rates.)
Another example is Blend Labs, which sells to large banks. At a high level, the financial sector has a massive TAM with lots of budgets and buyers that are familiar with purchasing software — but its very highly concentrated with a few customers (346), and very correlated to economic changes.
Churn can be a problem, even when serving enterprise.
One of our assumptions on why larger customer counts correlate with high market capitalizations is that there is more room to make mistakes. Typically, as SaaS investors, we typicall only see churn as problematic in the SMB or mid market segment. However, in vertical SaaS, high customer concentration — despite large ACVs — can really hurt a business because so few customers are driving high percentages of the revenue. And if you lose even one or two customers (which is very typical in horizontal SaaS), it can throw off a forecast for the street — which a public stock can be severely punished for.
Vertical SaaS’ strength and unfortunate achilles heel is that products are purpose built for a specific industry. That can lead to lower customer acquisition costs, and ease of becoming a market leader in a short period of time — but it can also mean you are painfully elastic to market downturns 2U, Q2 Holdings, nCino, and Five9 are good examples here: despite having high NTM revenue — their market capitalization is on the lower end given the total customer count they serve.
As always, if you are an entrepreneur building in software — horizontal or vertical — ping us with any questions or thoughts. We would love to hear from you!
~Thanks to the amazing Rohan Sharma, who has been an avid reader of my blog for several months and helped with this analysis!