> 67% were in sales, marketing and customer success, 20% in engineering, product development and customer operations and 13% in general and administrative
Ratios like this are fairly common.
I think the confusion comes from engineers who look at a SaaS and assume it's mostly engineers like themselves. Engineering is a small portion of the headcount.
A lot of customer-related needs scale with the number of customers. Any company with a huge number of customers is going to have a lot of headcount related to dealing with customers.
For some, the confusion comes from "wow, that's about half the headcount that Microsoft had when I started in the mid-90s, and MSFT was building, selling, and supporting Windows, Visual Studio, Office,..." Perhaps DocuSign's products are, in fact, about half of Microsoft's big three products, but of what I know of DocuSign it's still a big number to me.
Microsoft in the mid-90s outsourced the majority of sales and customer support to retailers and VARs. If you add up all the employees working to sell and support Microsoft products in external companies the total staffing was enormous. The same situation applied to their competitors like Novell.
I wasn't around then but is it possible that we have learned more about B2B SaaS since the mid-90s through trial and error? And the conclusion the industry has come to is Microsoft would have been better off at that time with more sellers and customer ops hires?
How many programmers did Microsoft need for 1 million OS installs vs 1 billion? The reason software can be such a money maker is that the marginal unit costs are close to zero.
I don't know the answer to either of those questions. But I don't think the economics of desktop OS distribution have changed all that much in the last 30 years. I'd imagine the marginal unit costs have gone from very cheap to basically nothing now that physical media is a rarity. If modern OSes take more development effort (which seems likely), I don't think that changes the economics.
I bet there's some market analysis of SaaS companies showing how differently they can be run. I wonder if it's bimodal - SaaS with <10% sales headcount, and SaaS with >50% sales headcount.
> Any company with a huge number of customers is going to have a lot of headcount related to dealing with customers.
Unless you are Google, which is then again where a lot of people in engineering can get confused as most companies don't hate their users quite as much as Google ;P.
This is also where companies can save a ton of money if they develop tools that allow customer facing employees to do their jobs much more efficiently.
Ratios like this are fairly common.
I think the confusion comes from engineers who look at a SaaS and assume it's mostly engineers like themselves. Engineering is a small portion of the headcount.
A lot of customer-related needs scale with the number of customers. Any company with a huge number of customers is going to have a lot of headcount related to dealing with customers.