Jefferies’ “SaaSpocalypse” slashes SaaS valuations as AI reshapes demand

A sector-wide selloff, dubbed "SaaSpocalypse" by Jefferies analysts, has repriced the entire software-as-a-service category, driving the iShares Expanded Tech-Software ETF down nearly 25% in 2026 and knocking major stocks lower: Salesforce off about 28%, Adobe down roughly 23%, and ServiceNow tumbling 10% on the day it disclosed US$3.5 billion in subscription revenue, up 21 percent year on year.

The market move is not, executives and analysts say, about isolated misses. Instead it reflects a deeper reckoning with the economics that have underpinned SaaS valuations for a decade. "SaaS valuations are built on two assumptions—fast growth and high net revenue retention (NRR). Both are breaking," a leading sector analysis states. The argument centers on NRR: existing customers have historically driven margin by expanding usage and adding seats; when they begin replacing vendor modules with internal tools, retention falls below the critical 100 percent threshold and the valuation model implodes. "The fatal one is NRR erosion. Existing customers are the profit engine of SaaS. They expand into higher tiers, add more seats, drive margin. When they start replacing parts of your platform with internal tools, your NRR drops below 100%. When NRR drops, your valuation model collapses."

That math is changing because modern AI and low-code tooling make bespoke alternatives far cheaper and faster to assemble. "When renewal comes and the price increased 15 percent, and you only use 40 percent of the features, and they don't quite match your workflow anyway, and your engineering team could build what you need in two weeks—the conversation is different now," the analysis says. It goes further: "If your product is 'a SQL wrapper with a billing system,' you now have thousands of competitors: every engineer at every customer with an AI agent and a Friday afternoon."

Technology is acting as an amplifier. AI systems are testing the assumptions behind the rent model by automating large portions of human workflow. "AI is now testing every part of that logic at once," a report explains, adding that modern AI "can replace large portions of human workflow outright. Research, analysis, drafting, reconciliation, and coordination no longer need to live inside a single application. They can be executed autonomously across systems." That dynamic has pushed some firms to shift strategy away from being monolithic applications and toward platforms that let customers build and extend or attach AI agents that orchestrate workflows across tools.

Vendors are already responding. Platform plays and low-code products are being emphasized as survival strategies: "ServiceNow built Creator Workflows so customers can build custom apps on their platform. Salesforce opened Agentforce to let partners build custom agents." Salesforce's Agentforce and Data Cloud have reached nearly US$1.4 billion in annual recurring revenue, evidence that some customers are buying into platform approaches even as the market frets.

The panic has a sentiment component. Analysts note that many software companies had not yet reported results when the selloff accelerated, suggesting fear has outpaced observable deterioration. "When the stock market punishes good results, it is telling you that sentiment has overtaken fundamentals," observers say.

Voices are divided on the magnitude of change ahead. One high-profile investor warned, "The Great SaaS Meltdown has started and there's no going back… A new AI-oriented workflow is coming…" while an industry leader countered that the idea AI will wholly replace software is "the most illogical thing in the world." As one systems forecaster cautioned, "we tend to overestimate the effect of a technology in the short run and underestimate it in the long run." For now, the market is forcing a rapid reassessment of which SaaS businesses must evolve into platforms and which may see their economics permanently altered.