Last-mile costs rise 12% for a second year

The survey gathered more than 3,000 data points from U.S. delivery operators in the first half of 2026. Control figures are based on the 41 operators who answered the control question. Fleet-mix figures are based on 84 operators.

Last-mile costs stay in double digits

Operators most often placed fuel (70 percent), driver cost and availability (51 percent), and vehicle operating cost (40 percent) among their top three cost pressures. Inefficient routing made the top three for 21 percent. Failed deliveries and returns made it for 18 percent.

“This twelve percent is obviously a combination of about six percent, which is the public rates that were increased by FedEx, UPS,” Nahata said. “But along with that, this additional six percent which gets contributed by the operational inefficiencies that we have in our operations.”

That 6-and-6 split is his on-stage accounting, not from the survey data. One takeaway is these continued increases in costs continue to place pressures on these businesses.

“Neither the CFOs nor the business allow us to operate at any cost,” he said. “The cost needs to come down as well.”

That mandate to tackle costs showed up in the survey data. Forty-five percent of operators put reducing delivery cost in their top three investment priorities, more than any other item. Companies above $1 billion in revenue posted a 13.8 percent median cost increase, the highest of any size band.

Reliability draws the promise

Predictable delivery time and successful first-attempt delivery together accounted for 55.7 percent of the promise operators said matters most. Fastest possible delivery accounted for 11.4 percent.

“What consumers are looking for, as we see it now, is speed along with certainty or reliability,” Nahata said. “The delivery promise is probably the biggest thing. That’s where consumers have moved.”

The eight operators who put maximum speed first posted 76 percent on-time performance and a 24 percent cost increase. Operators who put predictable delivery first posted 88.4 percent on-time and a 10 percent cost increase. Visibility-first operators posted 90.3 percent on-time and a 4.9 percent cost increase.

Where-is-my-order contact, or WISMO, moved with that cost pattern. Operators with WISMO rates above 30 percent posted 17.2 percent cost inflation and 81.4 percent on-time performance. Operators at 5 percent WISMO or less posted 7.3 percent inflation and 86.5 percent on-time. The high-WISMO group carried 2.4 times the cost inflation of the low group.

“WISMO probably is a good economic indicator of your last-mile cost as well,” Nahata said. “It’s not just about customer experience.”

Proactive delay notifications were the only customer-experience capability in the survey tied to better on-time performance, at 88.1 percent with the notifications against 81 percent without them.

“If we’re trying to make a delivery promise, it’ll impact the cost side as well,” Nahata said. “We can’t solve one for the other.”

Control tracks on-time performance

Among the 41 operators who answered the control question, on-time performance ran from 65.5 percent in the low-control group to 93 percent in the middle group and 95 percent in the high-control group. WISMO ran 20.8 percent, 10.4 percent and 6.2 percent. Median cost inflation ran 14.5 percent, 10.5 percent and 8.3 percent. Median investment was the same across the three groups.

Higher technology maturity sat with a lower cost-per-delivery band and had almost no relationship with year-over-year inflation or on-time performance.

The dominant network is already mixed. Fifty-seven percent of operators run a hybrid fleet, 26 percent are fully outsourced and 17 percent run an owned fleet only. Forty-seven percent of hybrid operators plan to outsource more. Hybrid networks posted 93 percent on-time performance and 7 percent WISMO.

These mixed network demands are placing greater pressure on visibility, as customers become more accustomed to real time tracking updates and want that as part of their buying experience. “Can your team answer these questions without picking up a phone?” Nahata said. “Where is the shipment?”

AI adoption is ahead of execution

Operators at an implementation or operational stage of AI rose from 46.2 percent in 2025 to 66.3 percent in 2026. Extensive operational adoption rose from 4 percent to 13.8 percent.

“Last year when we looked at AI implementation, about forty-six percent of businesses said they are doing it,” Nahata said. “Now we hear it’s sixty-six percent.”

“There were four percent last year who said full scale. Now almost 14 percent are saying it’s a full-scale one.”

Leading use cases were ETA prediction (39 percent), demand forecasting (36 percent) and customer support (31 percent). Real-time dynamic routing, the only operational-decision use case in that list, sat last at 21 percent. Mean trust in AI for real-time operational decisions was 1.98 on a 4-point scale.

“Where we see the biggest change is moving from prediction to execution,” Nahata said. “It’s not just about answering, it’s about getting things done as well.”

Operators with the least roadmap backlog posted 93 percent on-time performance and 5.2 percent WISMO. Operators with the most capabilities still on the roadmap posted 85.5 percent on-time and 17.4 percent WISMO. Operators running three or more named systems posted 80.9 percent on-time. Those on a single system posted 85.6 percent.

“The winners will not be the companies that promise the most,” Nahata said. “They will be the companies that can keep the promises that matter, at an economics the business can scale.”