Salesforce study finds field service firms deploying AI at scale, but struggling to keep workers trained

Field service organizations are rapidly rolling out artificial intelligence, but Salesforce says the bigger obstacle is not the technology itself. In a new report released Wednesday, Salesforce argues that organizations are outpacing their ability to train and support the mobile workforces that must act on AI recommendations, creating a talent risk that can weaken the promised economic gains.

The report, State of Field Service: The Road to Revenue in the Agentic Era, is based on a survey of 2,317 field service professionals across nine countries. The survey was conducted April 22 through May 12, 2026. Salesforce frames the findings around an “agentic” future, meaning software agents that can take actions on behalf of users, rather than simply answering questions or generating suggestions.

According to the survey, nearly all field service organizations (95%) have already deployed some form of AI. The commitment is also increasing, with 85% of leaders saying they plan to raise AI investments over the next one to two years. Salesforce says AI is no longer treated as a pilot effort, but as a core strategy tied to near-term business outcomes.

Leaders’ priorities for the next 12 months cluster around customer and worker outcomes: 35% cite improving customer satisfaction, 31% focus on mobile worker productivity, 27% highlight safety, and 25% prioritize increasing revenue. On the deployment side, 54% of leaders said they use AI-driven tools for customer communication, while 51% said they use AI to assist employees working in the field.

Salesforce reports that when organizations connect relevant systems and prepare their teams, AI is delivering measurable results. Eighty-five percent of leaders say they have measured the return on investment from their AI investments, and the reported ROI appears aligned with the stated priorities. In scheduling and dispatch, 57% of organizations using AI-powered scheduling and dispatch said they see higher revenue per job. Many also reported operational spillovers, including 57% citing higher mobile worker productivity, 52% pointing to reduced emissions, and 49% reporting lower labor costs.

The report’s central warning is that workforce disruption is rising even as AI adoption expands. Two-thirds of field service leaders said mobile worker turnover has increased over the past two years. When asked what is behind the increased turnover, leaders pointed to insufficient training or support when new technology is introduced, suggesting that workers may be leaving when systems and workflows change faster than they can adapt.

Compounding the training problem is what the report describes as a data access gap. Sixty-one percent of organizations said mobile workers have limited access to the customer data they need. The implication, as Salesforce frames it, is that even trained employees can struggle to act on AI-generated recommendations if the underlying information is not readily available where work happens.

Salesforce also attributes some of the measurement and operational difficulties to fragmentation across platforms. Only 16% of organizations say their field and back-office technology are united on a single platform. When customer data, schedules, and asset records are stored in separate systems, the report says it becomes harder to ensure that AI tools can draw on the right information, and harder to determine what is driving results.

The tension shows up in measurement itself. While 85% of leaders say they have measured AI ROI, 40% said they struggle to measure whether AI is working. Salesforce links that uncertainty to the same fragmentation problem, including the reality that data can be spread across apps, spreadsheets, sensors, and paper, making it harder to isolate which AI improvements are delivering value.

In selecting vendors for AI agents, Salesforce says organizations are not primarily driven by cost. Instead, they prioritize transparency into how AI makes decisions, data security and privacy practices, quality of ongoing support, and external validation. The company’s broader argument is that AI should be treated as critical infrastructure, where outcomes depend on people, data, and operational integration as much as on the model or tool itself.

The report does not quantify how much turnover is reduced when organizations improve training, nor does it provide company-by-company performance benchmarks. It also does not specify which AI technologies are used most often, beyond scheduling and dispatch and basic customer communication and field assistance categories. What remains clear is the linkage between adoption, training capacity, data access, and the ability to demonstrate sustained ROI.

For field service leaders, the next phase of AI deployment may hinge less on expanding features and more on workforce enablement. Watch for how companies standardize training and support during rollouts, improve access to customer and asset data for mobile workers, and consolidate platforms enough to make AI performance easier to verify over time.