SaaS – The Fears, The Future, The Opportunities. A broad look at Saas and its roll in our AI future.

I've been a software engineer for 14 years now and have been through a lot of iterations of software over my time and AI has been one of the most fun and exciting revolutions in recent history for me. I have been reflecting on it as of late due to the market fears, comments about AI being useless, and others saying its revolutionary.

The Fears:

Seat Compression: Less licenses, less money.

Commoditization of Code: Code is cheap, "Lowering the MOAT".

Economic Displacement: White collar jobs are replaced by software.

Types of SaaS:

Systems of Engagement – DOCU, ZM

Systems of Record – CRM, NOW WKDY

Cybersecurity – PANW, CRWD, NET, ZS

Infrastructure & Data – SNOW, DDOG, MGDB

Specialty – SNPS, CDNS, SAP, CSU, VEEV, INTU

DevOps – TEAM, GTLB

Hyper Scalars: Google, Meta, Amazon, Microsoft (All down like 20% +)

My thoughts:

I spent some time trying to understand why INTU in particular has been hit so hard. My wife has a small business and for payroll there aren't a lot of great options outside of intuit. Commoditization of code: was what I found to be the most prevalent "disruptor" to Intuit. For example, if Google or a startup could link into your bank accounts, and agentically pull your banking records and automatically compute payroll or expenses it would remove the need for Intuit. There is also some fear of an AI advisor, but an AI advisor that is bound by any real responsibility is questionable at best.

While I was working through this case study with Gemini, it pointed out Google could do this at some point but it would incur fiduciary responsibility and lawsuit risk. It dawned on me and its been true for the last 10 years. Google, Amazon, Microsoft and these other tech giants have hoards of the best software engineers for the last 20 years and aren't going to start entering an accounting space for a 100B dollar market cap company.

If giants won't enter this space what about the small guys? A small scrappy start up? Does AI make VC money so much more efficient that they can bust into highly established business processes or highly regulated markets? Are they so much more efficient now, that the business's weren't worth pursuing before? What is that multiple? A team of developers needs to be 10-15x more efficient so 100M stretches to 500M in salary output of the past? Are you going to outcompete the established teams with moats of money?

You are always going to have overhead with these startups, cloud hosting, security, snow, ddog, you name it. Half of these "seat compressions" will be filled with the competitors trying to disrupt the kings.

AI hasn't "changed software", it still does the same thing. Startups don't have some implied advantage because the game has changed. They can just work faster now like everyone else. There are some fun inferences you can make with AI and the logic you can build around it but functionally the results are the same.

One last thing. The earnings reports have all been pretty incredible. Google's CEO during his earnings call even reiterated bullishness on SaaS companies. If this were structural I would be more concerned, but the capex spend is reinforcing the narrative software is growing. The chip companies are still falling despite record spend.

I have DCA'd half of my assets into software names over the last 3 days. I will continue to add the rest over the next week or two as things shake out. I am spread across most of the names. The safest bets for those who are risk adverse are SNPS, CDNS (duopoly in chip software/analysis/QA) and Cyber Security as vibe coders deploy exploits i guess? Do your own research and be patient.