A new Gartner report is making waves across the software industry, and if you run a business or manage a technology budget, it deserves your attention. Published July 1, the research firm warns that agentic AI is on track to put $234 billion of enterprise application software spending at risk by 2030. That is roughly 20% of the entire enterprise SaaS market.

The term Gartner uses is “agentic arbitrage.” It describes what happens when an AI agent completes tasks that used to require a person toggling between multiple software tools. When an AI can pull data from your CRM, update your ERP, generate a report, and send a notification — all without a human opening a single application — the traditional model of selling software by the seat starts to fall apart.

The Seat Licensing Model Is Breaking

Enterprise software has been built on one core assumption for decades: more users equals more revenue. You pay per seat, per user, per month. Vendors grow by landing more seats and expanding into more teams.

Agentic AI breaks that assumption completely.

If an AI agent handles the work that five people used to do across three different platforms, why would you pay for five seats on each platform? You would not. And that is exactly what Gartner is flagging.

The displaced spending does not just disappear. Gartner says it will be cannibalized by incumbents and captured by new entrants delivering what they call “horizontal agentic platforms.” These are systems that deliver measurable outcomes rather than feature access, and that can act as the agentic layer across an entire enterprise stack. The per-seat licensing model does not fit this world, and vendors who have built their entire revenue model on it are facing a structural problem.

Who Gets Disrupted First

Gartner’s research points to the categories most exposed: workflow tools, reporting platforms, middle-office automation software, and any application where the primary value is data entry and retrieval rather than genuine decision support.

Think about how many software subscriptions your business carries where the main job is moving information from one place to another. That is the category most at risk. Agentic AI does not care about your UI. It reads APIs, executes tasks, and returns results. The software sitting in between becomes optional.

Tools with deep workflow logic, proprietary data models, or domain-specific decision support are more insulated. The commoditised middle — software that exists mainly as a connector between other systems — is the most exposed.

What This Means for Business Leaders

For business leaders, the Gartner finding is not just a warning about what happens to software vendors. It is a signal about where to direct your own investment and attention right now.

The businesses that capture value in the next four years will be the ones building or deploying the agentic layer — not buying more seats on top of it. That means evaluating your current software stack not just for cost, but for replaceability. Which tools does your team use because the software genuinely cannot be replaced, and which do they use because no better option existed until now?

A few questions worth putting to your team:

• Which of our current tools exist primarily to move data between other tools?
• Where are people spending time on tasks that an AI agent could complete in seconds?
• Which vendors are actively building agentic capabilities, and which are defending their seat model with pricing changes?

The Shift From Software Spend to Outcome Spend

The bigger picture here is a shift in how businesses should think about technology budgets. The question is changing from “what software do we need?” to “what outcomes do we need, and what is the most direct path to get there?”

That is the premise behind AI agent workforces. Instead of buying another seat in another tool, you deploy an agent that handles the entire workflow — consistently, around the clock, without the overhead of per-seat contracts. The ROI case is direct: fewer seats, faster execution, measurable output.

Gartner’s data suggests the market is already moving faster than most enterprise software vendors anticipated. By 2030, businesses that have made this shift will carry a structural cost advantage over those still managing stacks of per-seat subscriptions.

The question is not whether this shift will happen. According to Gartner, it already is. The question is whether your business is positioned to benefit from it or to be disrupted by it.

---

Enterprise DNA put together a free field guide on exactly this: the full Claude ecosystem, Claude Code, and how to roll agents out without breaking things. Get the guide.