There’s a lazy narrative forming in the market right now. On one side: “AI is going to kill SaaS.” On the other: “Nothing changes — SaaS will evolve like it always has.” Both are wrong. What’s actually happening is more important — and more disruptive: AI is expanding the value software can deliver, while disrupting the way that value is monetized.
The Model That Built SaaS
For the last two decades, SaaS ran on a remarkably consistent formula: charge per seat, drive value through feature depth and UI workflows, and grow revenue by expanding adoption. The logic was elegant. More users meant more seats. More workflows meant more time in product. More time meant more perceived value — and more revenue. It printed money, and it scaled.
AI Breaks the Link Between Usage and Value
AI changes one fundamental thing: users no longer need to use software to get value from it. Instead of logging in, navigating workflows, and clicking through processes, they can ask a question, trigger an action, and get an outcome. From a user perspective, that’s a massive improvement. From a business model perspective, it introduces serious friction — because fewer clicks, fewer workflows, and potentially fewer users required means you’ve just decoupled usage from value creation. Most SaaS pricing models were built on exactly that coupling.
The Quiet Risk: Revenue Compression
If one AI-enabled user can do the work of three, why does a company need three seats? This creates a very real dynamic: customers can get more value while paying less. It won’t show up overnight. It will emerge gradually — in flat or slowing seat growth, lower expansion revenue, and declining usage metrics that don’t match output. If you’re only looking at ARR, you’ll miss it.
Gartner has put numbers to this: by 2030, at least 40% of enterprise SaaS spend is projected to shift toward usage-, agent-, or outcome-based pricing. Deloitte Insights That’s not a distant forecast. The repricing has already started.
The shift is already visible in the data — and it’s accelerating faster than most pricing teams have adjusted for (see chart below).
The Metrics Are Breaking
Most companies still measure success through monthly active users, feature engagement, and time in product — metrics that assume more interaction equals more value. That assumption is breaking. In an AI-enabled product, the best experience may require less interaction. The most valuable users may be the least active. That flips how product performance should be measured, and most teams haven’t adjusted yet.
The Real Moat Is Shifting
For years, SaaS moats were built on feature sets, UX, and switching costs. AI compresses all three. The moat is no longer feature velocity — it’s data density, the unique contextual layer a product sits on: customer records, operational logs, domain-specific metadata. Webapper
The new moat has three pillars. The first is workflow ownership — are you embedded in how work actually gets done? The second is proprietary data — do you have context that improves outcomes over time? The third, and most underestimated, is the risk and trust layer. In financial systems, payments, and regulated environments, accuracy matters, controls matter, and trust matters. You can’t just “plug in AI.” You need systems that can operate safely and reliably at scale. That last point is where most of the actual competitive differentiation will be built.
Ranked by competitive durability in an AI-first environment, the gap between old moats and new ones is stark (see chart below).
What the Smart Money Is Actually Doing
If SaaS were truly in terminal decline, capital would be leaving the space. It isn’t — but the story is more nuanced than simple confidence. Vista Equity Partners CEO Robert Smith told investors that the vast majority of Vista’s portfolio companies are not experiencing meaningful churn or losing clients to AI alternatives, arguing that AI will enhance software, not replace it. Yahoo Finance Vista has also launched what it calls an Agentic AI factory, with roughly a third of its more than 90 portfolio companies having already used those tools to automate tasks and improve productivity. Yahoo Finance
Meanwhile, Thoma Bravo — the world’s largest software-focused investment firm with more than $183 billion in assets under management — announced a strategic partnership with Google Cloud on April 15, 2026, specifically to help its portfolio companies accelerate AI transformation and deploy agentic AI capabilities. PR Newswire
This is not the same SaaS model they were backing five years ago. The bet is on AI embedded into real workflows — systems that replace work, not just support it.
This Isn’t the End of SaaS — It’s a Reset
Gartner predicts that by 2030, 35% of point-product SaaS tools will be replaced by AI agents or absorbed within larger agent ecosystems of major SaaS providers. Deloitte Insights That still leaves 65% that survive — but likely in evolved forms that look nothing like the seat-based subscription businesses of today.
The takeaway isn’t that SaaS disappears. It’s that the rules change: pricing will evolve beyond seats, metrics will shift from activity to outcomes, and the gap between companies that adapt and those that don’t will widen sharply. Some companies will face serious margin pressure. Others will expand by owning higher-value workflows. The divergence will be dramatic.
The Behavioral Layer No One Is Talking About
Here’s what gets lost in the infrastructure debate: the hardest part of this transition isn’t technical — it’s behavioral. Enterprises don’t reprice vendor contracts because a technology changes. Buyers don’t shift budget lines because Gartner publishes a prediction. Organizations move slowly, protect incumbents, and often optimize for familiarity over efficiency.
The SaaS companies that win this transition won’t just be the ones with the best AI. They’ll be the ones who understand that behavior change is the actual bottleneck — and who design their products, pricing, and go-to-market motions around how humans actually make decisions, not how they theoretically should. That’s not a technical problem. It’s a strategy problem. And most companies aren’t treating it like one.
Early Read
Strip out the noise and the signal is clear: software is moving from a tool you use to a system that does the work. If the last decade of SaaS was about digitizing workflows, the next decade will be about automating outcomes. That’s not a collapse — it’s a shift in where value sits, and who captures it. And right now, most companies are still measuring the wrong things.
💡 This piece was sparked by a reel from @atlasberry008 on Instagram — go check it out.
Sources
- Gartner / Deloitte — “SaaS meets AI agents: Transforming budgets and workforce dynamics” (February 2026) — deloitte.com
- Bain & Company — “Will Agentic AI Disrupt SaaS?” (Technology Report 2025) — bain.com
- Bloomberg / Yahoo Finance — “Thoma Bravo, Vista Reassure Investors as AI Selloff Hits Software” (February 2026)
- Thoma Bravo + Google Cloud Partnership Press Release (April 15, 2026) — prnewswire.com
- PinkLime — “The Death of Per-Seat Pricing” (February 2026) — pinklime.io