AI is forcing a reset in SaaS - Here’s where value will be created

The recent SaaS market sell‑off reignited a fundamental debate: will AI agents replace traditional software business models, or reshape them? At Fortino, we believe the answer depends on who owns the workflow - and how fast execution follows strategy.

Last month saw a severe SaaS sell‑off, wiping out $300 billion in market capitalisation as fears mounted that AI agents would fundamentally disrupt traditional software business models. Microsoft, Salesforce, and others took significant hits. The panic accelerated when Anthropic released Claude 4.6 Opus, designed to handle complex professional workflows that many software providers sell as core products. Anthropic’s productivity tool for in‑house lawyers sent legal software stocks tumbling: Thomson Reuters fell 16%, CS Disco 12%, and LegalZoom 20%. Jeffrey Favuzza from Jefferies termed it the “SaaSpocalypse”- indiscriminate “get me out” selling that erased $830 billion in software market value.

The core fear is straightforward: AI agents replacing the per‑user SaaS licensing model. If AI tools can accomplish the same tasks with significantly fewer staff, the traditional per‑seat licensing model collapses. Palantir CEO Alex Karp warned that AI has become so capable at writing and managing enterprise software that many SaaS companies risk obsolescence.

Not everyone agrees. Nvidia CEO Jensen Huang called the notion that software would be replaced by AI “the most illogical thing in the world,” arguing that AI will use and enhance existing software rather than replace it. Markets, however, did not wait for the debate to settle before repricing risk.

The real AI divide in SaaS: features vs. orchestration

The market is repricing undifferentiated feature vendors. Companies that own workflows in their verticals—where AI executes decisions rather than merely assists - will see very different outcomes.

Our focus is B2B legacy SaaS in vertical markets. These are established software companies with proven business models, embedded customer bases, and domain expertise accumulated over years or decades. Most were built before cloud‑native architecture became standard. Many carry technical debt. All face the same strategic question: how to capture the AI upside without getting displaced.

The majority operate as vertical software businesses, with a small number following different models, including horizontal SaaS and IT services.

Generic horizontal SaaS faces direct AI substitution risk. Vertical software integrated into operational workflows, with proprietary domain data and deep system‑of‑record integration, has a fundamentally different risk profile. The data these companies generate - execution logs, decision traces, exception patterns, and domain‑specific parameters - cannot be replicated by foundation models trained on public data.

This is about owning the orchestration layer: the layer where decisions are made and logged, domain expertise is encoded into workflows, and AI agents must integrate because these workflows are deeply embedded in customer operations. By controlling this layer, companies can capture the AI upside, accelerate top‑line growth, improve profitability relative to peers that fail to execute, command significant price premiums, and earn higher market multiples.

The emerging pattern is clear. Frontier models handle exploration and synthesis. Execution happens through domain‑specific agents. Orchestration maintains state and logs outcomes. Companies that own that orchestration layer in their verticals are positioned very differently from those providing AI‑assisted features.

AI unlocks authority‑driven SaaS premiums

Legacy B2B SaaS monetised human effort. Seats, features, and usage reflected the time and attention of humans operating the software. Value scaled with usage because humans retained authority.

Agentic AI fundamentally breaks this logic. When software can act autonomously, value is determined by the authority delegated to the AI - the scope and consequence of the decisions it can make - not how often it is used.

This shift is already visible. ServiceNow prices around workflow ownership. Zendesk prices around resolutions rather than interactions. Stripe prices AI around transaction value. In each case, pricing tracks delegated responsibility, not technical execution. Where pricing lags authority, value is understated and customer tension emerges.

The implication is clear: SaaS leaders must rethink pricing as a reflection of delegated authority and accountability. Products that replace roles or decision loops should be priced in line with the responsibility they assume - not the number of clicks, API calls, or users. AI‑enabled SaaS with clear authority‑based pricing can command significant premiums over legacy models.

The SaaS reset has started - execution will decide the winners

The “SaaSpocalypse” repriced horizontal SaaS on AI substitution risk. Vertical software with workflow ownership faces different dynamics - but execution velocity determines outcomes.

B2B legacy SaaS companies have structural advantages: embedded customers, proprietary domain data, and deep workflow integration. They also face structural constraints: technical debt, legacy architecture, and limited resources.

The transformation challenge is organisational and driven by execution speed. European B2B SaaS lags US peers by two to three years, and legacy platforms face additional friction from accumulated technical debt. The opportunity is real. The execution timeline is now.