Thoma Bravo Declares SaaSpocalypse Over, Cites AI Tailwind

Orlando Bravo, founder and managing partner of private equity firm Thoma Bravo, told CNBC at the SuperReturn International conference that "The SaaSpocalypse is over. It's finished, no more." CNBC and Yahoo report the firm manages almost $200 billion in assets and that its portfolio companies generate about $35 billion in combined revenue. Bravo said roughly 50% of new revenue across his portfolio is "AI revenue, agentic revenue," according to CNBC, and he described AI as an "enormous tailwind" for software in coverage by CNBC and The Next Web.

Editorial analysis - technical context: Reporting links the market shock in February to Anthropic's rollout of Claude agent tools, which prompted investor fears that agentic AI could disrupt seat-based SaaS pricing. Industry coverage in The Next Web and CNBC highlights that agentic features could increase per-user compute consumption, which may raise operational and cloud-inference costs for vendors and customers alike.

Reporting in The Next Web documents a clear market bifurcation: companies owning AI "picks-and-shovels" and consumption-priced offerings have rallied, The Next Web cites DigitalOcean up more than 220%, Datadog around 76%, and CrowdStrike over 50% year-to-date, while many seat-based application vendors remain materially below their earlier highs, with The Next Web noting HubSpot down roughly 46% and Monday.com about 45% down despite positive revenue growth. CNBC and Yahoo additionally note investor and management concerns remain around governance, cybersecurity, and returns from agentic deployments.

For practitioners: track three observable indicators in coming quarters rather than infer company intent. First, vendor disclosures of AI-derived revenue and how firms define "agentic" line items, as Bravo cited 50% AI-driven new revenue for his portfolio. Second, unit-economics data for agentic features, per-seat pricing changes, consumption-metering, and cloud inference spend. Third, governance and security reporting tied to agentic workflows, which CNBC flagged as open questions for investors. Observers will also watch whether product positioning shifts from seat-based licensing to consumption or outcome-based models and how that affects integration, support, and monitoring tooling needs.

Editorial analysis: This set of public statements and market moves underscores an industry transition where headline optimism about AI adoption coexists with uneven real-world economics. Companies that capture the underlying consumption layer or provide efficient inference plumbing appear to be reaping outsized market gains, while traditional seat-based vendors face increased scrutiny about replacement risk and margin pressure. For data science and ML engineering teams, that implies more emphasis on cost-aware model architecture, inference optimization, and observability for agentic behavior rather than just raw capability.