Software stocks fell nearly 13%. Your renewal is a different story

If you have software renewals coming up, did February make you a little nervous about them? For about a week, a lot of the conversation online came down to one word, SaaSpocalypse, which started as a strategist's phrase and turned into a meme within days. I get why it spread, because when a whole sector's share price drops that fast it feels like news about the software you pay for every month. Some of it was, to be fair, but most of it was news about investors, and I think it's worth pulling those two apart before anyone makes a buying decision off the back of it.

What actually happened

The timeline is short, and The Batch from DeepLearning.AI lays it out well. Anthropic introduced Claude Cowork, "an agent designed for professional work," on January 12, and on January 30 it released 11 open-source plugins, each aimed at a white-collar job function. Four days later the S&P Software & Services Index fell 4 percent, erasing more than $285 billion in market capitalization, and Jeffrey Favuzza, a strategist at Jefferies, gave the moment its name.

It kept going from there. Reuters, republished by BNN Bloomberg reported that the S&P 500 software and services index slid nearly 13 per cent over five straight sessions, in the same week the S&P 500 scaled an all-time high. The plugins automate tasks across legal, sales, marketing and data analysis, and J.P. Morgan's Toby Ogg summed up the mood by saying the sector "isn't just guilty until proven innocent but is now being sentenced before trial." Honestly, that line describes the mood online pretty well too. The plugins had been out for less than a week, so hardly anyone declaring the end of SaaS could have used them for long, and the verdict arrived with mostly a launch announcement to go on.

Then the story flipped. On February 24, as The Batch notes, Anthropic announced integrations with some of the companies Cowork was supposed to threaten, including Docusign, FactSet, Google's Gmail, Intuit and Salesforce. Instead of going around those applications, the new plugins connect to them directly. Reuters, via Kitco News, reported that shares of software companies that partnered with Anthropic helped lead a rebound, after a week-long selloff earlier in the month had wiped out about $1 trillion in market value.

So within a month the same lab went from being the thing that would replace your vendors to being the thing your vendors were happy to plug into. The damage was still real for shareholders, since by February 23, the day before that announcement, the Software & Services Index had lost 25 percent of its value since January 12. But between the crash and the rally, the CRM your sales team uses did not get worse and then better again, and the contract tool kept sending contracts the whole time. A share price tells you what investors fear about the next few years, and very little about whether the thing your team opens every morning still does its job.

What it means at renewal time

The calmest read I found came from the Daiwa Institute of Research, a short report by Miho Tanabe asking what "the death of SaaS" actually means. She points out that replacing SaaS with in-house AI agents brings "significant compliance burdens, including security governance and audit logs," and that designing, implementing and maintaining those systems one company at a time is a heavy load, which is why AI built into SaaS stays the practical route for most companies. She also suggests that as AI becomes the main user of software, what sets products apart may shift from the interface to "external integration (APIs, etc.), permission management, and auditability," and she expects traditional pricing based on user numbers to be revisited. I could be wrong about how quickly that plays out, but none of it says software goes away. It does say the reasons you pay for it are moving, from "my people need to click on it" toward "my systems need to connect to it safely, and I need to see what they did."

I'm not going to tell anyone what to do with their portfolio (that's well above my pay grade), but if you buy software for a company, February handed you something useful, and it isn't a reason to rip anything out. It's a better conversation at renewal. If a vendor charges per seat and you expect agents to do more of the clicking, it's fair to ask how they plan to price that, and what their API can do today rather than what's on the roadmap. I'd also ask how permissions work when the user is a piece of software, and whether you can actually get an audit log out of it. Those are the questions the Daiwa report points at, and they're reasonable to raise whatever the stock did this week.

What I'd put on the backburner is treating a ticker as a product review. The selloff was a story about what AI might do to software and the rally was a story about partnerships, and both moved a lot faster than any team could change the way it actually works. Pricing pages, API docs and changelogs are worth a look instead. When a vendor quietly moves away from per-seat pricing, or ships proper audit logs for agent activity, that tells you more than a week of red candles, and for what it's worth, it's also the stuff we think about most when we build Workbricks.