C.H. Robinson Worldwide Touts “Lean AI” at Morgan Stanley Conference, Citing 40% Productivity Gain

C.H. Robinson Worldwide (NASDAQ:CHRW) executives used a Morgan Stanley TMT Conference appearance to outline how the logistics provider is applying what it calls “Lean AI” to drive productivity, improve service levels, and expand its competitive position in a fragmented transportation market.

Positioning in a fragmented logistics market

CEO Dave Bozeman described C.H. Robinson as a major logistics platform that connects shippers with carrier capacity, characterizing the company as “in the center of a really complex, fragmented market.” Bozeman said the company handles about 37 million shipments annually, serves roughly 75,000 customers, and works with more than 450,000 carriers.

Bozeman said the company’s goal is “to make simplicity out of the complex,” and argued that its operating model, technology stack, and “logisticians” combine to create a competitive moat that is difficult for others in the industry to replicate.

“Lean AI” as operating model and technology strategy

Bozeman, CFO Damon Lee, and Chief Strategy and Innovation Officer Arun Rajan emphasized that the company’s approach pairs lean operating principles—focused on rigor, operating discipline, and continuous improvement—with AI-enabled technology and human expertise. Lee said the company is often asked how much of its recent performance is attributable to lean versus technology, and added that the two are “so combined, so symbiotic” that it is difficult to separate their effects.

According to Lee, AI-driven initiatives have delivered 40% productivity since the end of 2022 across the enterprise. He also argued that Robinson’s AI efforts are not limited to cost savings, citing revenue impacts including “outgrowing the end markets for over 10 quarters” and improvements in revenue management capabilities such as price optimization and cost-of-hire optimization.

Rajan framed the long-term objective as building a scalable platform in which incremental volume requires minimal incremental cost, comparing the target economics to an “Amazon-like model.” He suggested this could strengthen the company’s value proposition for customers considering whether to outsource logistics and transportation functions.

AI infrastructure approach: application layer, model flexibility, and cost discipline

Management described C.H. Robinson as operating primarily in the AI “application layer,” with large language models treated as a commodity input. Rajan said the company uses Azure AI Foundry and selects models based on price-performance and token costs, which he said are declining rapidly over time.

Lee and Rajan said the firm’s architecture allows it to switch among models because its applications are “abstracted from the underlying LLMs.” Rajan noted that observability tools and test harnesses enable teams to adjust models if costs increase, and that in many cases the company does not need the newest model versions to achieve desired outcomes.

On cost control, Lee said the company’s overall spending has not increased since it began investing in AI, and that AI investments have been contained within existing spending levels. He also argued that because the company builds its own technology rather than purchasing off-the-shelf tools, the marginal cost of deploying an additional agent is close to zero beyond token usage. Rajan offered a data point on efficiency: over the past year, he said token usage increased 85x while costs rose about 1.5x.

Competitive moats: data, context, internal building, and change management

When asked why competitors may struggle to replicate these capabilities, executives pointed to several factors:

• Scale and uniqueness of data: Management said Robinson has the largest dataset in the industry, built over decades, and argued that it cannot simply be purchased.
• Context engineering: Rajan described building a “context layer” that captures standard operating procedures and “tribal knowledge” from employees, enabling AI agents to execute workflows with guardrails that reduce hallucinations and contain costs.
• Builder vs. buyer approach: Bozeman emphasized speed and “velocity experiments,” arguing it is difficult to achieve similar iteration cycles when relying on third-party tools.
• Operating model adoption: Lee argued that lean is hard to implement deeply and sustainably, and said combining it with AI provides a “delivery mechanism” for prioritizing and executing automation efforts.

Lee also said companies often build agents to automate individual tasks rather than optimizing entire workflows, which he described as necessary to achieve meaningful productivity gains.

Top-line benefits and operating leverage in an upcycle

As freight markets begin to show early signs of improvement after what participants described as a multi-year downturn, executives said their automation efforts could allow Robinson to capture volume without traditional headcount expansion. Lee said the industry has historically responded to cycles by reducing staff in recessions and rehiring in upcycles, but Robinson believes its automation will reduce the need to add headcount as volume returns, creating “substantial” operating leverage.

One example offered was the company’s quote process. Lee said that historically, when quotes were managed primarily by humans, Robinson responded to about 60% to 65% of transactional quote requests. With an “orchestrating agent” managing the quote cycle, he said Robinson now addresses 100% of those requests. He also said the response time fell from 17–20 minutes to about 31 seconds, with the agent using roughly 1,000 data points to generate more sophisticated responses that have increased win rates and supported margin expansion.

Bozeman and Lee both stressed the company views its transformation as early-stage. Rajan said the company has not automated “nearly as many processes” as it ultimately expects to, describing current efforts as “second inning.” Lee added that while automation has been concentrated in its North American surface transportation business, the company is “just now starting” to deploy the same technology into forwarding, with results expected to emerge in the second half of 2026.

On workforce implications, executives said they are not managing the business to a headcount KPI. They described a shift in job focus from running day-to-day operations to managing SOPs and context for agents, while freeing employees to solve more complex customer problems. Bozeman noted turnover in some roles runs about 11%–14% and said the company has been transparent with employees about the evolution of work.

About C.H. Robinson Worldwide (NASDAQ:CHRW)

C.H. Robinson Worldwide, Inc is a third-party logistics provider founded in 1905 and headquartered in Eden Prairie, Minnesota. Originally established as a produce brokerage firm, the company has since expanded its offerings to become one of the world’s largest freight and logistics intermediaries. C.H. Robinson leverages a global network of transportation providers, technology platforms, and in-house expertise to connect shippers and carriers across multiple modes of transportation.

The company’s primary services include truckload, less-than-truckload (LTL), intermodal, air and ocean freight, and managed transportation solutions.