The SaaS Apocalypse
Is it actually so over?
THEMATIC DEEP DIVE: The Death of Software or the Evolution of Software?
Inflection Point: The decoupling of software value from “Seats” (human labor) to “outcomes” (AI labor) and “physics” (infrastructure). Tickers Mentioned: ASAN, CRM, FIVN, SNOW, DDOG, CRWD, IOT, GTLB, CSU.TO
The Great Rotation and the Structural Revaluation of Code
If you look at your screen today, it looks like the market has decided software is uninvestable. For the better part of a decade, SaaS was a synonym for Alpha. You bought a basket of high-growth cloud stocks, you went to the beach, and you compounded at 20%+. That trade is not just broken, it is being actively liquidated.
The software industry in 2026 finds itself navigating a precarious inflection point, a moment of reckoning that many market observers have termed the “SaaS Apocalypse.” This narrative is grounded in a profound structural shift in the economics of technology. Following a turbulent 2025, where the software sector, represented by the IGV ETF, returned a mere 6% against the S&P 500’s 17% gain, a deep-seated anxiety has gripped the investment community. The prevailing fear is not just cyclical but existential: that Artificial Intelligence - specifically the rapid maturation of generative and agentic AI - is transforming from a feature enhancement into a potent deflationary force that will systematically dismantle the unit economics of the traditional Software-as-a-Service (SaaS) model.
The core tension defining this era is the divergence between the 2D World of screens, code, and digital interfaces, which is facing a massive deflationary supply shock, and the 3D World of atoms, energy, and physical infrastructure, which is entering a super-cycle of scarcity and inflation. As Generative AI models collapse the marginal cost of producing code and content to near zero, the premium valuations historically assigned to software intellectual property are being rigorously questioned. Investors are increasingly asking a fundamental question: will enterprises continue to rent expensive seats for pre-packaged software when they can increasingly build bespoke, autonomous agents via natural language prompting or vibe-coding?
However, broadly painting the entire software sector as uninvestable ignores the critical nuance of this technological transition. While the chaff - defined as seat-based, workflow-heavy horizontal applications with shallow moats - faces existential risk from commoditization and agentic displacement, the wheat - data-centric platforms, cybersecurity leaders, and mission-critical infrastructure - remains not only vital but increasingly valuable. Evidence from the enterprise frontline contradicts the total collapse narrative; the RBC 2026 CIO Survey indicates overwhelming momentum, with 95% of IT leaders planning budget increases and 90% prioritizing AI spending, treating it as a new investment layer rather than a reallocation from existing budgets. Furthermore, despite this bearish sentiment, the structural demand for AI workflows is strong.
The Total Addressable Market (TAM) for enterprise SaaS AI workflows is projected to nearly triple, growing from $65 billion in 2025 to $190 billion by 2030, representing a Compound Annual Growth Rate (CAGR) of nearly 24%. The critical question for investors is not if this growth will materialize, but where it will accrue.
The divergence in fortune will be defined by a company’s ability to pivot from selling tools for humans to providing infrastructure for agents. We dig in and provide an analysis of these dynamics, exploring the existential risks facing incumbents like Salesforce, Asana, and Five9, identifying the compelling structural opportunities in Datadog, Snowflake, CrowdStrike, Samsara and GitLab and then evaluating the vertical market software aggregator Constellation Software as it navigates this paradigm shift.
