Software stocks swung wildly this week, exposing a narrative fault line. On Monday, Italian buyout firm Bending Spoons agreed to acquire Airtable for under $1.3 billion—roughly one-tenth of its $12 billion peak valuation in 2021. The following Tuesday, HubSpot fell 19%, its worst day in a decade, on concerns about AI agent disruption, while Datadog sank 19% the next day as the company disclosed its largest AI client (suspected to be OpenAI) cut usage since June. Meanwhile Atlassian reported its most profitable quarter since 2021 and surged 35% on Friday, its best day since going public in 2015. Twilio also jumped 20% on strong results, and Cloudflare gained 5.6%, suggesting the death-of-SaaS narrative is premature.
The mood in software has been apocalyptic since March, when the iShares Expanded Tech-Software ETF plummeted 24% in Q1—its worst quarter since 2008. Investors had panicked that coding agents like Claude Code and OpenAI Codex would hollow out economics for companies like Atlassian and Salesforce. Salesforce has lost 40% of its value since end-2024 despite accelerating revenue and steady margins, evidence of pure investor sentiment collapse. RBC analyst Matt Hedberg noted sentiment was so bleak in Q1 that investors didn't even want to meet with software companies: "People were pencils down."
The fracture is between high-quality SaaS with pricing power, product-market fit, and direct customer relationships (Atlassian, Twilio, Cloudflare) and generic or exposed categories like marketing automation and monitoring (Salesforce, HubSpot, Datadog). The AI agent threat is real for some workflows, but companies that reported earnings and demonstrated customer strength got reprieve. For investors and operators, the lesson is that "agents will disrupt SaaS" is true but granular—it matters which software category, which customers, and which pricing model. One-size-fits-all death predictions were wrong.