The structural reset in software has begun. As AI reshapes the unit of work, seat-based SaaS pricing is no longer a certainty but a question of economic survival.
Software equities have experienced a broad and sustained drawdown. Sector performance is down approximately 20% from October highs, with several large enterprise software vendors trading at 52-week lows. Historically, this environment would have been interpreted as a familiar buying opportunity. Over the past five years, market pullbacks in software were consistently followed by rapid recoveries, reinforcing a widely accepted dip-buying playbook.
That assumption now deserves closer examination.
The current decline is not driven solely by sentiment or macro volatility. It reflects a deeper reassessment of SaaS economics as advances in generative and agentic AI begin to undermine the foundations of seat-based pricing.
Structural Drivers of Persistent Dip-Buying Behaviour
Retail investors now account for a materially higher share of daily equity market volume than they did a decade ago. This shift has coincided with the rise of passive investment vehicles, which introduce mechanical buying behaviour largely independent of price sensitivity.
Together, these forces have reinforced a powerful mean-reversion dynamic. Drawdowns trigger retail inflows, passive allocations stabilize prices, and markets recover. Since 2020, this pattern has repeated often enough to shape investor expectations and behaviour.
However, this dynamic relies on an implicit assumption: that market declines are cyclical rather than structural. When valuation compression reflects a reassessment of long-term cash-flow durability, historical recovery patterns become less reliable.
Why the Current Software Drawdown Is Structural
The pressure facing software companies today is not primarily demand destruction. It is erosion of the unit of value itself.
Seat-based SaaS pricing assumes that productivity scales with human headcount. Generative AI disrupts this assumption. As AI systems take on a growing share of operational work, organizations can maintain or increase output with fewer human users. Fewer users directly translate into fewer seats.
Even significant price increases struggle to offset this effect. A reduction in seat count can easily overwhelm price expansion, resulting in net revenue contraction. As AI capabilities improve, this dynamic intensifies rather than stabilizes.
This shift is already visible in enterprise behaviour. Large organizations are actively rationalizing SaaS portfolios, reducing license counts, and replacing portions of traditional software stacks with AI-driven tools. These decisions are operational and strategic, not experimental.
Industry forecasts increasingly reflect this transition. Pure seat-based pricing is expected to decline in relevance, replaced by consumption-based, outcome-based, or capability-driven monetization models.
Underlying Economic Pressures on SaaS Revenue Models
The current software selloff reflects a growing recognition that traditional SaaS monetization is misaligned with how value is now created.
AI agents do not require seats. They do not scale linearly with headcount. As AI performs more work, the linkage between revenue and human usage weakens. This places sustained pressure on pricing power across large portions of the software ecosystem.
For many vendors, the challenge is not competition but arithmetic. As productivity increases, the denominator in seat-based pricing shrinks. Over time, this compresses revenue unless pricing models fundamentally change.
Strategic Constraints Facing Incumbent SaaS Vendors
Most established SaaS vendors recognize the risk and have responded cautiously. AI features are introduced as extensions or usage-based add-ons, while the core seat-based structure remains intact.
This approach preserves near-term revenue and minimizes disruption to existing customers. However, it does not resolve the underlying mismatch between value creation and monetization.
As AI agents increasingly function as digital labour, vendors face a strategic inflection point. They must determine whether they are selling access for humans or completed work for organizations. Maintaining pricing anchored to human usage while value shifts toward autonomous systems creates long-term fragility.
Incremental adaptation delays the problem rather than solving it.
Emerging Differentiation Within the Software Sector
The market response suggests growing differentiation across software categories. Outcomes increasingly depend on business model characteristics rather than brand strength or historical growth rates.
- Data Infrastructure as a Structural Beneficiary of AI Adoption
Data platforms, databases, and observability tools form the foundation of AI deployment. These systems are deterministic, mission-critical, and scale directly with AI usage. As AI workloads expand, demand for reliable data infrastructure increases accordingly.
These businesses face minimal substitution risk and are widely viewed as long-term beneficiaries of AI adoption.
- Systems of Record and Their Evolving Role in an AI-Driven Stack
Systems of record occupy a more nuanced position. While they face pressure from declining seat counts, they retain strategic importance as execution layers.
AI agents do not bypass systems of record. They operate within them. Customer data, employee records, tickets, approvals, and financial workflows remain anchored to platforms with high switching costs and deep organizational integration.
Over time, these platforms may evolve from workflow tools into transactional and decision-execution infrastructure. Market pricing often treats them as commodity SaaS despite their structural similarities to infrastructure assets.
- Heightened Substitution Risk for Horizontal Productivity Software
Horizontal productivity and creative tools face the most direct exposure to AI substitution. These systems are probabilistic, easier to replicate, and generally exhibit lower switching costs.
In many cases, AI does not enhance these products. It replaces them. As a result, pricing pressure and demand displacement are likely to persist in this category.
Valuation Divergence and Growth Quality Reassessment
Valuation dispersion across the software sector increasingly reflects differences in growth durability rather than absolute revenue scale.
Businesses tied to AI infrastructure command premium multiples. Traditional SaaS vendors dependent on seat expansion face sustained multiple compression. The distinction is not growth versus stagnation, but growth driven by AI-native demand versus growth tied to human usage.
Markets are beginning to price this difference, though inconsistently. That inconsistency creates both risk and selective opportunity.
Capital Reallocation Toward Physical and Industrial Assets
At the same time, capital is rotating away from software toward physical and industrial assets. Manufacturing, energy, logistics, and infrastructure investment have accelerated, supported by reshoring initiatives, large-scale capital expenditure programs, and policy incentives favoring domestic production.
The underlying rationale is straightforward. AI compresses the marginal value of digital output, while physical capacity remains constrained. Scarcity shifts from bits to atoms.
This rotation carries execution risk. Labor shortages, supply-chain complexity, and macro uncertainty remain meaningful constraints. However, physical assets derive value from difficulty and capital intensity, characteristics that software increasingly lacks.
Potential Market Outcomes and Repricing Scenarios
Several outcomes remain plausible.
One scenario involves a return to historical patterns, with broad software recovery driven by retail inflows and improved risk appetite.
Another involves continued dispersion, with infrastructure maintaining premium valuations, systems of record stabilizing at defensible multiples, and productivity software facing ongoing pressure.
A third scenario reflects a deeper structural repricing, in which even high-quality platforms struggle to sustain historical valuations as pricing models transition and growth visibility declines.
The most likely outcome lies between the first two. Dispersion persists, and business model resilience becomes the primary determinant of returns.
Implications for Software Investors and Operators
The software sector is not experiencing a routine correction. It is undergoing a structural reset.
Seat-based pricing, long viewed as a durable and scalable model, is increasingly misaligned with how value is created in an AI-driven economy. Some software businesses will adapt successfully. Others will not.
For investors and operators alike, the central question is no longer whether software will recover, but which business models remain economically viable as AI reshapes the unit of work itself.
