AI is no longer a forward-looking theme in SaaS. It is a live operating shift reshaping product design, economics, and competitive positioning. Yet despite the intensity of discussion, there is a clear divergence between conversation and capability.
Across the market, many organisations remain anchored in conceptual debates about AI – its promise, its disruption, its theoretical impact on software. A smaller but increasingly important group has moved beyond this phase entirely. They are embedding AI into operating models, restructuring workflows, and rethinking how value is created, delivered, and captured.
This gap is not cosmetic. It is becoming commercially decisive.
Fredrik vom Hofe describes this transition as a shift into “AI+SaaS execution mode”, where the question is no longer whether AI matters, but whether companies can translate it into measurable customer value and sustainable unit economics.
The execution gap is now the real constraint.
The most immediate challenge facing SaaS businesses is not access to technology. It is execution readiness.
While AI capabilities are widely available through external models and platforms, most organisations have not yet built the internal structures required to deploy them effectively. Fewer than one in four companies, according to recent industry analysis referenced in the discussion, have established robust foundations for AI adoption.
This creates a widening gap between Group 1. Organisations experimenting at the edge of AI capability and Group 2. Organisations redesigning core workflows around AI-native execution.
The latter are moving faster, not always because they have better tools, but because they have aligned product, engineering, and commercial functions around a shared execution framework.
AI is forcing a redefinition of what “good” looks like.
Historically, SaaS performance has been measured through relatively stable benchmarks: ARR growth, retention, expansion revenue, and gross margins supported by scalable delivery models.
AI disrupts each of these assumptions simultaneously.
In Fredrik’s framing, AI is not simply improving efficiency. It is changing the nature of software from a system of record into a dual construct:
This shift, then, fundamentally alters what “good” looks like. It is no longer sufficient to build scalable software. The expectation is now the ability to embed intelligence into workflows that materially change customer outcomes.
Customer value becomes the defining principle.
As execution pressure increases, customer value shifts from a strategic objective to an operational imperative.
Companies that succeed in this environment are those that can demonstrate a clear path to expanding the complexity of customer problems they solve. This is no longer about feature depth alone. It is a about whether AI enables step-change in value creation across the customer lifecycle.
Fredrik highlights three interlinked requirements:
Where this is absent, AI becomes additive rather than transformative – and the commercial impact remains limited.
Execution, not experimentation, is now the differentiator.
The most important shift underway is the move from AI experimentation to embedded execution. It requires more than technology adoption. It demands organisational redesign.
AI is now being deployed across functions to drive operational leverage – engineering, customer support, sales operations, and product development. But the real differentiator is not isolated efficiency gains. It is whether these gains are structurally integrated into the operating model.
Companies that get this right are starting to scale non-linearly. Those that don’t are still tied to linear cost and headcount models, even if they are labelled as AI-enabled.
The implication for SaaS leadership
The conclusion is increasingly clear. AI readiness is no longer a technology question. It is an operating model question.
The gap between hype and execution is widening because execution requires alignment across product architecture, commercial strategy, and organisational design. Most companies are still early in that journey.
Those that close this gap will not simply be more efficient SaaS businesses. They will redefine what SaaS performance looks like in the AI era.
The rest risk being left in a category defined not by innovation, but by execution delay.
