Thread by @patio11

I often ask myself "If I got a do-over on my entrepreneurial career, what would I have hoped to have known when starting and so not had to learn the hard way?" Today's installment, on the business model behind SaaS companies. stripe.com/atlas/guides/b… Some thoughts:

One of the reasons why it is hard to advise SaaS companies is that they are all similar enough to each other to be a cognizable group but there are some within-SaaS cleavages that make the advice very, very different.

When I was starting, I thought the big one was B2B vs B2C.

This turns out to be very, very wrong. The single biggest difference isn't B2B vs B2C, it isn't the market, and it isn't whether a company is funded or bootstrapped.

Every other fact about a SaaS company gets bent by gravational waves emanating from the sales model.

The sales model is classically either low-touch (the product mostly sells itself, probably on a free trial, with assistance of the website, email, and judicious support / customer success work) or high-touch (a heavily human-driven process).

This drives pricing decisions, marketing channels, the relative importance of departments (design, marketing, engineering, sales), top priorities for the management team, etc. To a degree that most people don't appreciate until they've worked in SaaS a while.

Here's a fun game: "What's your ACV [annual contract value]?" *you name a number* "I will freehand your org chart."

This game is easier to play than tic tac toe.

Regardless of the sales model, SaaS companies are driven under the hood by the best economic model anyone has come up with for software. You only need four numbers to judge the health of a SaaS company: acquisition, conversion, price, and churn.

It works like this:

That equation is sort of stylized and has some simplifying assumptions, because all models are wrong but some models are useful. One which is certainly wrong: it assumes that churn is a continuous product.

In low-touch SaaS as it is actually practiced, churn happens early.

A lot of customers will essentially extend their trial, sometimes with the permission of the company and sometimes not. (A surprising-to-younger-me portion are happy to pay for the privilege! "Oh didn't get around to using it yet but don't want to cancel; I still might...")

So you don't settle into your long-term churn rate until ~3 months into use of the software, frequently. Savvier firms track the seasoned rate and ones which have extreme amounts of savvy (and time on their hands, and data) do ongoing cohort analyses.

But the simple model works!

Speaking of the simple model working: one of the reasons I'm so jazzed to write about SaaS at Stripe (and we will be writing a lot more in the future) is that there is a wide body of practices which are in the uncomfortable position of being known by everyone and done by few.

I sometimes got asked during my consulting career what the secret black magic was. The secret black magic is 98% boring execution on the same set of things that everyone else knows to do and few successfully organize their firms into doing consistently.

"What's the other 2%?"

I'll tell you if I ever figured it out. My one original thought, in ~10 years in the business, was altering the first-run experience and flow of the free trial in response to exact marketing channel a user came in through.

(Please steal that. It worked.)

(Conceptually this is "If they Google [$PRODUCT api docs] and end up on your API docs and then sign up for the free trial, and your free trial comes in Developer, Finance Professional, and Operations Professional flavors, you don't have to ask them which trial they want.")

Since the various sub-disciplines in a SaaS company get arbitrarily deep, you get interesting returns to specialization. Effectively none of them are taught to any useful degree in school.

This causes networks to be very, very powerful in SaaS.

Rather little of the technology of running SaaS businesses is written down. (Working on it!) A huge portion is bringing together a team of people who've done similar things before.

This is one reason why Silicon Valley tends to race ahead with regards to funded SaaS.

I live in Tokyo and love it here, but you could scour Tokyo high-and-low and probably not find someone who would give the 47th best SaaS lifecycle email writer in SF a run for their money. (That is a reasonable professional speciality to have given how useful it is.)

And since the rate of skill growth is generally proportionate to the rate that the company hits new challenges at, the best thing you can do for yourself in SaaS is to work at a team where the functions supporting you work reasonably well.

(That might be true of most work?)