Software Is Fine
Everyone is freaking out about software. Like in the old days when AWS would announce something and investors would collectively gasp, so too is happening when anything in AI comes out.
Let’s not forget that the same human problems exist. For all the investors freaking out about Claude Code, have you tried it? I have and it’s incredible. I may actually use it to build a custom CRM. But for me to build something that everyone else in the company would use trusting my vibecoding and my prompts & maintenance is unlikely. We would still rely on someone else for that critical data.
Now the obvious question would be, won’t a startup rise to build custom software that they are able to maintain because they are AI native and deep enough in agent capability to do so? I think the answer to this is 100% yes. This is how disruption happens. They will target the low end first and then first slowly and then rapidly tackle the high end and larger customers.
As a venture investor, this is what excites me. Finding founders rethinking what is possible with AI and delivering such a great user experience that customers are willing to switch. The fact that pricing could also initially be lower given the leaner cost structure is an added tailwind as fast iterating teams will land at customers, drive value, and expand quickly.
A true AI Native custom software company should be able to price lower as they need less humans to build and manage the agents which are building & managing the software. They hopefully will need less humans to manage the humans using the software. On the customer support side, it definitely seems so. On the sales side, enablement, procurement, etc the tail might be longer but should trend down.
So...does all of this matter to the current freakout in software stocks amongst investors.
Absolutely not!
I see so many tweets talking about the End of Software or trying to justify how it won’t end. None of that matters.
What’s driving software right now is flows.
It’s fairly simple.
Semiconductor, datacenter exposed, etc are going up. When those go up in the tech sector, funds need to hedge their exposure. The natural place to hedge is software. This causes software names to fall and break key technical uptrends which begets more selling from quantitative strategies like the ones SIG, Jane Street, Citadel run.
When correlations go to 1, nothing matters, everything falls.
That being said the highest quality software stocks are still not down that much in terms of valuation. Cloudflare, Snowflake, Shopify, Crowdstrike, Palantir. Only a few like a Samsara or Datadog have fallen to 10x FWD which still isn’t cheap by historical standards but for their quality one could argue is.
The rest of the crowd being punished are companies that are already seeing slowing or choppier growth, still have large amounts of stock based comp with uncertain payoffs due to slowing growth, and have made large acquisitions recently adding to that SBC and reinforcing the narrative of if the core is losing in the age of AI.
Atlassian is one that typically fits in the exceptional high quality area. That being said they have spent $1.7B on 2 acquisitions recently, not a small amount for a sub $50M market cap company. ServiceNow has been on a tear spending $12B on multiple acquisitions even while past ones are yet to be integrated. Adobe paid $1.9B for Semrush. It’s not like any of these companies are buying AI Native companies to build the future. They’re buying pre AI companies that maybe have a new AI narrative to them but are not fundamentally changing anything.
And so this is why the stocks are trading where they are. Do I think they’re too cheap here, absolutely. Can they go lower, absolutely.
I don’t see what will change the narrative until semis start to peter out and given TSM’s earnings, Intel coming back into the folk, and megacap tech about to rip new Capex numbers it likely won’t happen till we go higher and have a bit of a blow off top.
At that point, funds will start degrossing. They will sell semis and datacenter plays and buy back software shorts and things will start to normalize.
Until the next narrative takes off.
Don’t get fooled by flows and positioning. Observe them, understand how they impact markets, and be aware of how that will impact you as well. Investors in classic software will need more time to see multiples come back potentially. Founders in pre AI software need to be trying to figure out that new AI module that can show growth in order to get a premium multiple.
But software will be fine. Human behavior doesn’t change. Disruption will continuously happen. Some will get disrupted and some will figure out a way through. Don’t confuse flows for fundamentals.