SaaS and the Impending Doom?
What Ho? AI.
I am an avid reader of Wodehouse, he published 97 books, I think I own about half of them. He is the greatest comic writer, ever. If I pick up any of his work, within a paragraph I’ll be chuckling. And I’m not usually of a merry disposition. This is one of my favourite lines.
“I’m not absolutely certain of my facts, but I rather fancy it’s Shakespeare—or, if not, it’s some equally brainy lad—who says that it’s always just when a chappie is feeling particularly top-hole, and more than usually braced with things in general, that Fate sneaks up behind him with a bit of lead piping.”
— Bertie Wooster, in “Jeeves and the Impending Doom” Written by PG Wodehouse.
Shareholder of SaaS companies understand precisely what Bertie was on about.
Over the past few weeks and days I’ve read many articles by successful venture capitalists and AI-native (whatever that is) founders telling us emphatically that the cost of software is going to zero, and that most of the moats have dried up. I initially find myself mainly agreeing with them. I nod sagely.
Then I read a response, perhaps by somebody that has heard REM perform live before they were famous, and also wrote the first compiler that still runs 80% of the Internet explain why the founder is wrong. I then find myself agreeing with most of what they say. I nod sagely. This cycle repeats itself several times a day.
Typically the people saying that SaaS is dead have a vested interest in saying SaaS is dead, as they run a business that would benefit from the death of SaaS. Those saying “oh no it isn’t” have their own priors and vested interests.
The latest cycle involved this post from Nicolas. Lots of smart people thought this was brilliant.
I then read this rebuttal. Some other smart people thought this was brilliant too.
I’ve stopped and started writing about the SaaS-apocalypse several times. My draft folder is deliriously discombobulated. What I’ve now realised is that I’m actually not sure what my position is. Clearly something has changed. Perhaps we have been viewing SaaS through rose coloured massive terminal value spectacles, and a market correction is overdue. I can’t help feeling though that those condemning the incumbents as has been are assuming that the will be is as good as already here.
I’ll make several obvious points, and then I’ll go back to being unsure for a little while yet.
Not all incumbent SaaS vendors are the same.
Not all AI first companies are the same. I’m still figuring out what an AI-Native company is, bear with me.
Correlation is not causality: Telling the markets, politicians and employees that AI is taking jobs is not the same as AI actually taking jobs.
This change is bigger than the on-premise to cloud change, so the disruption is likely to be larger.
If you have been saying every year for the last 15 years that HR tech is undergoing a revolution, you can say it again this year, but I’m not sure you are going to be right just yet.
Understand the motivation of the folks writing the stuff you read.
Most AI today is sold as SaaS.
The most innovative and effective department at the large incumbent vendors is their licensing / pricing department. If Startup x thinks they can sit a 3rd party agent on top of SAP or Oracle, switch off the seat licences, and gobble up that datarevenue, they are in for a nasty shock. I suspect those departments are building agentic licence auditors as we speak.
Most significant enterprise AI deployments resemble 1990s ERP projects more than they do PLG SaaS.
Bruised and battered is not dead.
As I almost always do I’ll leave you with a song. I’d hinted so here it is.
And a bonus. Hugh Laurie, who amongst many things, played Bertie Wooster in a brilliant BBC series, is pretty nifty on the blues piano.
By the way, thank you so much for posting both the origina post and the rebuttal. I had the same experience as you. I agree with guy 1. And I agree with guy 2. Confused.
But FWIW I don't think it's impossible to largely agree with both. On some of the details, yes, they're in opposition (e.g., how much of FactSet's value prop was search), but broadly
Yes, a lot of things got easier to build
But I still think people want to buy apps built by people with domain expertise who package them.
Think: we're a bunch of marketing demandgen experts and we're going to use these amazing power tools to build something you really need. That's what I think is going to happen.
It always comes down to two things imho
1. Who wins as the standard underlying platform (e.g., OS, DBMS, now SaaS SoRs, foundation models)
2. Who can capture expertise (horizontal/functional (e.g., HR) or vertical) best and sell it to a buyer to improve their work life.
A thoughtful and honest take. We explored the SaaSpocalypse on a recent episode of The Metrics Brothers, but we didn't land hard on one side or the other. My primary points: the stock market is not the economy (i.e., the stock market is not the business). SaaS stocks that are getting hammered while doing beat-and-raise and growing 20% (e.g., $NOW). This is not, yet at least, the businesses falling apart. This is the stock market thinking the party might be over and realizing well that 15x revenue is perhaps a bit steep for a SaaS stock to begin with.
AI will take jobs, the question is does every CEO cut half their company overnight, crash the economy, and cause massive strife and disruption. I know there's a tragedy of the commons angle here, but hopefully they're smarter and more conservative than that.
Some of the jobs won't be missed. Some will be pushed up into higher value-add. New jobs will get created. The question is simply how fast all this goes down and too fast is definitely a big risk and possibility.
On the dark side, we could see 1960s-level social unrest combined with 1930s-era unemployment. What fun.