Saaspocalapse: How AI Is Reshaping SaaS Valuations and Igniting a Hardware Resurgence
Over the past decade, software-as-a-service (SaaS) businesses dominated both public and private market valuations. Recurring revenue, high margins, and scalable distribution created one of the most compelling investment narratives in modern markets. That narrative has now materially shifted. Over the last five years, a combination of macroeconomic tightening and, more importantly, the rapid advancement of artificial intelligence has triggered what many investors are now referring to as the “saaspocalypse”, a structural repricing of SaaS.
The data is unequivocal. Public SaaS valuation multiples peaked at approximately 18.6x EV to revenue in late 2021. By early 2026, that figure has compressed to roughly 6.4x, representing a decline of around 65 percent. This is not a cyclical correction. It is a structural reset.
The Compression of SaaS Valuations
The initial phase of this decline was driven by interest rates. As capital ceased to be free, long-duration assets such as high-growth SaaS companies were repriced downward. However, what followed is more significant. Even as markets stabilized, SaaS multiples did not recover.
Instead, they bifurcated.
Today, traditional SaaS companies typically trade between 2.5x and 7x revenue, while AI-native companies command materially higher multiples, often exceeding 10x and in some cases far beyond that. This divergence reflects a fundamental reassessment of future defensibility.
AI as a Deflationary Force in Software
Artificial intelligence is compressing the value of traditional software by reducing the cost and time required to build it. This is not theoretical. It is already happening at scale.
Recent data indicates that 35 percent of companies have already replaced at least one SaaS tool with an internally built or AI-generated alternative. This single statistic captures the core issue. The “buy versus build” equation, which historically favoured SaaS vendors, is being rewritten.
AI coding tools and autonomous agents now enable smaller teams, or even non-technical operators, to replicate large parts of the SaaS stack. The implications are direct:
- Increased competition across nearly every SaaS category
- Reduced pricing power due to commoditisation
- Lower switching costs as AI simplifies migration and integration
In parallel, markets are pricing in the expectation that headcount will decline across many industries due to AI-driven productivity gains. This directly impacts seat-based SaaS models, which have historically scaled with employee growth.
The $1 Trillion Reset in Software Markets
The scale of this repricing is substantial. Estimates suggest that public SaaS companies have collectively lost over $1 trillion in market value since their peak. Importantly, this has occurred while the overall software market continues to grow, with global spending projected to reach approximately $1.4 trillion in 2026.
This is not a decline in demand. It is a decline in perceived scarcity.
Software is no longer viewed as inherently defensible. Instead, investors are distinguishing between:
- Software as interface, which is increasingly replaceable
- Software as infrastructure, which remains deeply embedded and defensible
Only the latter continues to command premium valuations.
The Rise of AI Infrastructure and Hardware.
As value compresses in application-layer software, it is expanding in infrastructure and hardware. AI is not lightweight. It is compute-intensive, storage-heavy, and power-hungry. Every meaningful advancement in AI capability requires exponentially more physical resources.
This has triggered a reallocation of capital toward companies that supply those resources.
A clear example can be seen in the resurgence of storage and memory providers. Following its recent re-emergence as an independent entity, SanDisk now reports approximately $14 billion in revenue and $6 billion in EBITDA, with strong profitability metrics that reflect rising demand for high-performance storage. While historically such companies traded at modest multiples, the strategic importance of storage in AI workloads is driving renewed investor interest.
The underlying dynamic is straightforward. AI models require vast datasets, high-throughput memory, and low-latency systems. Without this infrastructure, AI does not function at scale. As a result, hardware is no longer a commodity layer. It is becoming the constraint.
Why Hardware Is Being Repriced
Three structural forces are driving the revaluation of hardware companies:
First, demand is becoming more predictable. Hyperscalers and enterprises are committing tens of billions of dollars to AI infrastructure, creating long-term visibility for hardware suppliers.
Second, supply remains constrained. Advanced chips, memory systems, and storage solutions cannot be scaled instantly, creating pricing power in the short to medium term.
Third, strategic importance has increased. In an AI-driven economy, control over compute and data infrastructure is equivalent to control over production capacity.
This is a meaningful shift. For the past twenty years, software captured the majority of value in the technology stack. Today, that value is redistributing downward.
A Rebalancing of the Technology Stack
What is emerging is not the collapse of SaaS, but its maturation. The sector is transitioning from a high-growth, narrative-driven asset class to a more measured, efficiency-focused one.
At the same time, AI-native software and infrastructure providers are capturing the premium that SaaS once commanded. The spread between these categories is widening, and it is likely to persist.
For investors, the implication is clear. The question is no longer whether a company is a SaaS business, but whether it is defensible in an AI-first world.
The rise of artificial intelligence is reshaping the economics of the technology sector. SaaS valuations are compressing as software becomes easier to build and harder to defend. At the same time, hardware and infrastructure providers are experiencing a resurgence, driven by the immense computational demands of AI.
For founders, operators, and investors, the implication is clear. The era of easy SaaS premiums is over. The next phase of value creation will be defined by those who control the infrastructure layer or build truly differentiated, AI-native software that cannot be easily replicated.
The market is not abandoning software. It is simply pricing it more realistically in a world where intelligence is becoming abundant and the resources to power it are becoming the true constraint.