Salesforce’s AI Bet Is Starting to Pay Off as Software Companies Move Past the ‘SaaSpocalypse’
Salesforce has delivered a strong rebuttal to the “SaaSpocalypse” narrative that dominated software-sector discussion for much of 2026. In its latest quarterly results the company reported revenue of $11.35 billion, up 11% year-on-year, and raised full-year guidance. More striking still was the performance of its AI products.
Agentforce, Salesforce’s platform for autonomous AI agents, reached $1.5 billion in annual recurring revenue, growing 240% year-on-year. Combined AI and data ARR approached $3.9 billion, more than tripling in twelve months. Customers generated 3.2 billion “agentic work units” in the quarter, nearly double the previous period. Net new average order value hit its strongest level in four years, customer attrition remained near record lows, and seats continued to expand rather than contract.
CEO Marc Benioff has been characteristically blunt. “This SaaSpocalypse narrative has been such nonsense,” he told CNBC after the results. Nine of the top ten frontier AI companies now run on Salesforce and Slack, with their combined spending up 435% year-on-year. A new partnership with Anthropic, branded Claudeforce, integrates Claude models directly into the Salesforce environment, reinforcing the idea that leading models need enterprise data, workflows and distribution rather than replacing them.
The broader fear that large language models would allow companies to build their own agents and discard traditional software licences has not materialised for Salesforce. Instead, agents appear to be driving greater platform usage. Data generated by agent activity is stored and analysed within Salesforce systems, creating a reinforcing loop that deepens the company’s competitive position. Slack has also benefited, delivering its fastest net-new annual order value growth since the acquisition as more users and external agents connect to the platform.
Not every software company will navigate the shift as successfully. Those whose products consist mainly of discrete tasks that agents can fully automate may still face pressure. Salesforce’s advantage lies in owning the customer data, process context and human collaboration layer that agents still require. As Benioff put it, frontier models depend on CRM; they do not replace it.
For UK and European software firms watching the sector, the Salesforce results offer a clearer picture of where value is accruing. Platforms that combine proprietary data, workflow orchestration and reliable agent deployment are demonstrating pricing power and retention. Pure feature-for-feature automation plays look more vulnerable. The SaaSpocalypse, at least for the moment, appears to have been overstated.
