The long reign of Software-as-a-Service (SaaS) is facing a fundamental challenge, as the rise of artificial intelligence reshapes the economics of digital business. For two decades, SaaS companies have thrived on predictable subscription models, but a new reality is emerging where the cost of delivering value is directly tied to usage – a utility model that threatens the traditional SaaS playbook.
The Unit Economics Shift
Historically, software operated under a favorable economic paradigm: build once, scale at near-zero marginal cost. AI fundamentally alters this equation. Every interaction with an AI-powered service incurs a cost, effectively turning software into a utility where the “meter is running” with each user request, as highlighted in recent analysis. This creates a paradox – the very engagement that businesses seek to drive with AI is simultaneously increasing their expenses.
This shift is already causing financial surprises. Companies piloting AI tools have experienced cloud bills tripling unexpectedly, a scenario described as a sign of finding something valuable, but also an entry point into a new era of AI unit economics. The traditional software model is being disrupted, and ignoring these changing economics could lead to unsustainable financial models.
SaaS Bloat and Underutilization
The SaaS model itself has become increasingly inefficient, ironically mirroring the problems it initially sought to solve. A McKinsey report from indicated that only approximately a quarter of purchased software licenses are consistently used. This “SaaS fat” – paying for underused applications and idle seats – is a significant drain on resources. Many platforms also function as glorified databases, requiring human input and subsequent data export for analysis, further diminishing their return on investment.
Despite the increasing integration of AI add-ons, which can drive price hikes of up to , fewer than one-third of firms are able to demonstrate a measurable return on these investments, according to McKinsey. This imbalance between expansion and efficiency is now being exposed by AI.
The Pressure on Pricing Models
AI is applying pressure to SaaS companies from both sides. Enterprise buyers, as reported by Bain & Company, now expect automation to deliver outcomes, not just display data. Platforms unable to meet this expectation risk becoming “shelf-ware.” Simultaneously, the very nature of AI is challenging the traditional seat-based pricing model.
As AI automates tasks, the number of “seats” required by a company decreases. However, seat-based pricing remains the default, creating a misalignment between value creation and cost. EY’s paper, “Agentic AI: How SaaS Companies Can Embrace the Future,” warns that traditional subscriptions are ill-suited to this new landscape and that hybrid or consumption-based models will likely become dominant. SaaS vendors are facing a critical choice: adapt their pricing to reflect outcomes, or risk being priced out of the market.
Loom’s Strategic Shift
The need for pricing innovation is not theoretical. Loom, a video messaging platform, has already begun to transform its SaaS pricing strategy with AI at its core. While specific details of Loom’s changes were not provided, this move underscores the growing recognition that traditional models are unsustainable in an AI-driven world.
The Rise of Agentic AI and the Future of Software
The emergence of “agentic AI” – AI systems capable of autonomous action – is further accelerating this disruption. While the complete elimination of software layers is unlikely, even with AI’s ability to generate code and perform tasks, businesses will still require security and governance. The software market is undergoing a transition, but it is a messy, non-linear process governed by technical limitations.
The value proposition is shifting from software itself to a company’s own intelligence layer. The most valuable digital asset of the next decade will be a company’s ability to leverage its own data and AI capabilities, rather than relying on generic software solutions. This suggests a future where businesses prioritize building internal AI infrastructure and integrating AI into their core operations, rather than simply subscribing to SaaS platforms.
Implications for Investors
The transition from SaaS to an agentic AI ecosystem presents both risks and opportunities for investors. A sensible approach, as noted in recent analysis, is to think in terms of probabilities and scenarios, recognizing that no one can predict the future with certainty. While dismissing AI as a bubble is shortsighted, proclaiming the death of software is equally premature. The key will be identifying companies that can adapt to the new economic realities of AI and successfully navigate the evolving landscape.
As of , the software sector is at a critical juncture. The traditional SaaS model is under pressure, and the future belongs to those who can embrace the challenges and opportunities presented by artificial intelligence.
