New Salesforce research finds near-universal AI adoption alongside rising frontline turnover, fragmented data, and measurement uncertainty
Field service organizations have moved past the AI pilot phase. According to State of Field Service: The Road to Revenue in the Agentic Era, a new Salesforce report based on responses from more than 2,300 field service professionals across nine countries, 95% of organizations have deployed some form of AI and 85% plan to increase that investment over the next one to two years.
The harder question the research raises is whether the workforce operating those tools is being brought along at the same pace. Two-thirds of leaders (66%) say mobile worker turnover has risen over the past two years, and when asked what’s driving it, they pointed to a single factor above all others: insufficient training or support when new technology is introduced.
For enterprise marketing and CX leaders, the finding lands in familiar territory. Field service is often the only in-person touchpoint a customer has with a brand, which makes the technician’s experience — and the data they carry into the appointment — a direct input to the customer experience.
Where the investment is going
Leaders named a clear set of priorities for the next 12 months: improving customer satisfaction (35%), improving mobile worker productivity (31%), improving safety (27%), and increasing revenue (25%).
Deployment patterns track closely to those goals. Fifty-four percent are using AI-driven tools for customer communication, and 51% are applying AI to assist employees working in the field — an allocation weighted toward the two constituencies that most directly shape service outcomes.
The ROI case is strongest in scheduling and dispatch
Eighty-five percent of leaders say they have measured ROI on their AI investment. The reported returns align with stated priorities: 43% cite higher mobile worker productivity, 40% cite improved customer satisfaction, and 34% report fewer safety incidents.
The clearest signal is in scheduling and dispatch. Among organizations using AI-powered scheduling and dispatch, 57% report higher revenue per job. The same cohort reports gains beyond revenue: 57% see higher mobile worker productivity, 52% report reduced emissions, and 49% report lower labor costs.
That concentration is worth noting. Scheduling and dispatch is a bounded, data-rich, high-frequency decision — the profile of a workflow where AI tends to produce measurable results early. It also suggests where organizations still in evaluation may find the most defensible business case.
The workforce and data gaps
Alongside the turnover finding, the report identifies a context problem at the point of service: 61% of organizations say mobile workers have limited access to the relevant customer data they need. Training a technician on an AI tool has limited value if the underlying customer record isn’t available when the recommendation surfaces.
Other operational barriers surfaced in the research:
- 49% lack a clear process for converting service visits into sales leads
- 44% have limited ability to quote in the field
- 38% struggle to accept payment in the field
Together these represent revenue that field organizations are positioned to capture but structurally unable to act on — a gap that should interest marketing leaders looking at service as a growth channel rather than a cost center.
Fragmentation sits underneath most of it
Only 16% of organizations report that their field and back-office technology are unified on a single platform. For the rest, customer data, schedules, and asset records live in separate systems.
The asset data picture illustrates the sprawl: 63% use mobile apps, 63% use inventory management systems, 59% use GPS trackers, 58% have connected sensors, 52% still use spreadsheets, and 43% still rely on manual or paper logs.
That fragmentation shows up in measurement as well. While 85% say they’ve measured AI ROI, 40% say they struggle to determine whether AI is working. The two figures aren’t contradictory so much as revealing: measurement is happening, but confidence in it is uneven when the underlying data is distributed across apps, sensors, spreadsheets, and paper.
What buyers are evaluating
When selecting a partner for AI agents, cost was not the leading criterion. Organizations prioritized transparency into how AI makes decisions, data security and privacy practices, quality of ongoing support and service, and external validation.
That evaluation profile is consistent with what enterprise buyers have signaled across adjacent categories over the past two years. As AI moves from feature to infrastructure, procurement criteria shift from price toward explainability, governance, and vendor durability.
What it means for enterprise leaders
Three implications stand out for marketing and CX leaders whose organizations include a field component:
Enablement is now a CX investment, not an HR line item. The report frames insufficient training as the top driver of frontline turnover. In a service model where the technician is the brand, retention and enablement spending has a direct downstream effect on customer satisfaction scores.
Data access at the edge is the constraint on AI value. With 61% reporting limited field access to customer data and only 16% operating on a unified platform, integration work — not model capability — is the likely gating factor on returns.
Service-to-sales conversion remains largely unbuilt. Roughly half of organizations have no defined process for turning a service visit into a lead, and substantial minorities can’t quote or take payment onsite. For teams evaluating where field service fits in the revenue picture, those are addressable workflow gaps rather than technology limitations.
