Gartner has issued a stark warning to the enterprise software industry. Up to $234 billion of enterprise application software spending, around 20% of the SaaS market, is at risk from agentic AI by 2030 as autonomous agents take over work that used to mean logging into a piece of software.
The analyst firm says legacy SaaS market share will be "cannibalized" as organisations shift towards AI agents that carry out work spanning several systems at once, rather than requiring employees to interact directly with individual applications. Gartner calls this "agentic arbitrage": less need for people to touch multiple software interfaces one by one.
George Brocklehurst, Managing Vice President at Gartner, said agentic systems "deliver outcomes directly, bypassing traditional user experience (UX)-heavy applications and making the software invisible." That breaks the ‘more users equal more revenue’ link that enterprise vendors have relied on for years.
The software itself isn't going away. What's changing is who's actually using it. Increasingly, that's an AI agent rather than a person, a dynamic that echoes how OpenAI recently positioned HP as a blueprint for AI as an operating layer across enterprise systems.
Why Customer Experience Platforms Are Directly Affected
Few corners of enterprise software look more exposed than customer experience. CX platforms are already moving towards agentic AI, with vendors including Salesforce, Microsoft, ServiceNow and NICE embedding agents capable of completing customer service, sales and operational tasks with limited human involvement.
In practice, that means agents updating CRM records, resolving customer enquiries end to end, triggering downstream workflows, and coordinating actions across several enterprise systems at once. Where an employee once logged into a CRM, then a ticketing tool, then a billing system, an agent can now run through that same sequence on its own.
Gartner's Brocklehurst put it bluntly: "As organizations increasingly use agentic AI systems, the user interface is no longer a differentiation." Software becomes infrastructure that AI works through, rather than a screen someone sits in front of. Gartner's view on the consequences of this is that legacy market share could ultimately be “cannibalised by incumbents” adapting their own products, or taken outright by new entrants building agentic platforms from scratch.
Either way, the vendors best placed look to be the ones moving fastest to embed agentic capability into how their software actually executes work, rather than defending existing dashboards and seat-based pricing. You can see this already playing out through Salesforce's headless enterprise bet, where agents interact directly with backend systems rather than through dashboards built for people.
The SaaS Business Model Could Be the Bigger Story
Strip away the technology and Gartner's forecast is really a question about money. Traditional SaaS pricing has always rested on named users and seat licences, on the assumption that value scales with headcount.
That assumption doesn't hold once one agent can do the work of several employees. Fewer people logging in means seat-based pricing captures less of the value those agents are actually creating. Gartner expects the vendors under most pressure to stop pricing around dashboards and screens altogether, instead tying revenue to the results those agents produce and baking agentic functionality into the work itself, not into features layered on top of it.
How that pricing question gets resolved may end up mattering as much as the technology itself. Brocklehurst believes the next few years will be “less an apocalypse and more of a metamorphosis" of the legacy SaaS market.
What CX Leaders Should Watch
A few things are worth keeping an eye on. Agents are becoming an operational layer that runs across enterprise software, not a feature bolted onto the side of it. Platform integration and orchestration look set to matter more than any single application's feature list. And the vendors that manage to shift from human-centric software to AI-first platforms are the ones best placed as buying models change underneath them.
Gartner's forecast, in the end, isn't really about SaaS declining. It's about a change in who, or what, is using enterprise software by 2030. If AI agents become the primary consumers of these applications, the next phase of this industry may be influenced as much by new commercial models as by new technology, reshaping what CX looks like in 2030 along the way.