Salesforce says its AI agents have helped generate more than 88 million dollars in revenue by doing a job its sales staff simply could not keep up with: making first contact with potential customers who would otherwise have gone unanswered.
The figure was shared by Itai Margalit, Salesforce's area vice president for sales in Israel, speaking at the Israel Information Technology Conference 2026 in Tel Aviv, in remarks reported by CTech on October 18. It is one of the more concrete revenue claims yet attached to a company's own internal use of AI agents, a category where vendors often talk about efficiency in general terms rather than naming a dollar amount.
The scale of the underlying problem helps explain why agents fit it. According to Margalit, Salesforce had 100 million leads, and before it integrated AI agents, about 75 percent of them went unanswered. Customers wanted to hear from the company, he said, but Salesforce lacked the capacity to reach everyone in a timely manner. Handling a comparable volume of initial responses through conventional staffing, he said, would have required approximately 400 additional employees.
The design, as described, keeps humans in the loop rather than replacing them. The agent initiates the first contact and hands the interaction off to a human whenever a personal touch is required. Margalit also noted that agents can now be deployed directly from within the work environment, including Slack, where Salesforce's team collaboration takes place, rather than forcing staff to switch to a separate application.
There is an important caveat, and the reporting states it plainly. Salesforce has not provided further details on the calculation behind the 88 million dollar figure, including the period over which the revenue was generated or how much of it can be attributed directly to the agents. That does not make the claim meaningless, but it does mean it should be read as a company-reported result, not an independently audited one.
Margalit's broader message was less about the headline number than about how companies get adoption wrong. Many AI projects fail, he argued, because they start from the end, rushing toward results before defining what they actually aim to achieve. His advice was to start small, choosing use cases that deliver significant business value while also being relatively easy to implement.
He also warned against treating AI as a layer that can simply be added on top of an unchanged organization. Companies must adapt their ways of working, he said, so the technology can actually move the organization forward. Underpinning that, in his account, is data: an accessible, reliable infrastructure that provides a single source of truth about customers, employees and organizational activity, shared by humans and agents alike. Without that foundation, automation can reproduce errors or act on an incomplete picture of the business.
For other companies weighing agent deployments, the Salesforce example is useful in both directions. It shows a bounded, high-volume task — first contact with leads — where agents can plausibly create measurable value, and it shows the limits of vendor-reported numbers. The result worth watching next is not another headline figure, but whether Salesforce, and companies like it, begin publishing the time periods, attribution methods and handoff rates that would let outsiders judge how much of the gain the agents truly earned.