XPO AI Labor Planning Lifts Productivity Nearly 2.5 Points

XPO said its AI-supported labor planning technology improved second-quarter productivity by nearly 2.5 points year over year. The disclosure, made on July 30, 2026, is the latest qualifying update, but XPO has not disclosed the application's standalone contribution to wage expense or operating profit.[1]

How XPO uses AI to plan LTL labor

XPO's core business is less-than-truckload freight, which combines shipments that do not fill an entire truck and moves them through a service-center network for loading, transfers and delivery. Labor-demand planning sits directly in that core operating workflow. The company's proprietary Smart and related predictive-AI tools forecast demand using current freight volumes, seasonal trends and the probability that sales leads will convert. Service-center managers and shift supervisors use those forecasts to determine the drivers, dock workers and labor hours required, helping schedules track changes in freight volume more closely.[2][3]

XPO's AI labor-demand planning has evolved from using Smart to forecast freight volumes and help align labor hours into a staffing decision tool that works at the service-center and shift levels and can forecast as far as 90 days ahead. In the second quarter of 2026, the company directly linked this technology to a nearly 2.5-point year-over-year productivity improvement for the first time.[1][2][3]

The disclosed development path began on February 6, 2025, when XPO described matching labor hours at the service-center level. On April 30, 2025, it detailed driver and dock-worker allocation at both the service-center and shift levels. By July 30, 2026, the company was still using the same labor-planning workflow and reported a stronger productivity result. The application is currently in limited production.

The productivity result—and its limits

The latest result was a nearly 2.5-point year-over-year productivity improvement, above XPO's 1.5% quarterly target.[1] This comparison shows that management treats the labor-planning technology as a measurable operating tool. However, the disclosure does not define the productivity denominator, coverage or calculation method, and it does not isolate the effects of other productivity initiatives. The figure therefore supports an operating-efficiency improvement, but it cannot be converted directly into labor hours saved, jobs reduced or profit added.

The financial link remains unquantified

The financial line item most closely associated with this application is salaries, wages and employee benefits expense, which ultimately affects operating profit. More accurate demand forecasts could reduce excess staffing and the labor hours required per shipment, potentially easing the pressure that wage inflation and freight-volume changes place on salaries, wages and employee benefits expense and, in turn, supporting operating profit.

However, XPO's salaries, wages and employee benefits expense still rose 7% year over year during the same period, reflecting factors that also included inflation and freight volume. The company said only that its broader productivity initiatives mitigated cost pressure; it did not separately disclose savings from the labor-planning application.

What can be confirmed is that XPO's AI labor-demand planning has developed from freight-volume forecasting into a service-center- and shift-level staffing decision tool, with a productivity improvement explicitly attributed by the company. Investors cannot confirm its actual contribution to the financial statements unless XPO discloses the labor hours or dollar amount saved and separately bridges that impact to salaries, wages and employee benefits expense or operating profit.

Application assessment

• AI Labor Demand Planning | Business position: core operations | Application stage: limited production | Deployment scope: unclear | Value type: cost reduction

Sources

[1] Drillr · XPO, Inc. (XPO) · July 30, 2026 · Earnings call

Original: In the second quarter, we used our workforce planning technology to improve productivity by nearly 2.5 points versus last year, which outperformed our quarterly target of 1.5%.

[2] Drillr · XPO, Inc. (XPO) · April 30, 2025 · Earnings call

Original: And we use AI in that demand forecasting model to be able to help us guide where we believe tonnage is going to be at the service center level and at the shift level.

[3] Drillr · XPO, Inc. (XPO) · February 6, 2025 · Earnings call

Original: Our systems can forecast volume trends using predictive AI so we can quickly align labor hours at the service center level.