Logistics

COMMENTARY

By Paul Tonsager

The rate is not the cost. The rate plus the data exposure is the cost.

Amazon Supply Chain Services( ASCS) launched on May 4. By the end of June, it will land on every shipper’ s desk as either an inbound sales call or a CFO-mandated cost study.

The market reaction was dramatic. FedEx shares tumbled 9 % on launch day, while UPS, GXO and Forward Air all took double-digit losses.

The question for shipper procurement teams is more practical. What do we do about it?

At Maersk, I ran a multi-billion-dollar procurement portfolio. The benchmarking tools we used were built to evaluate rate, service tier, and capacity reliability. None were built to evaluate a trade in which the cost of a service is partial control of our own demand data.

ASCS forces that trade, and most procurement organizations do not have a playbook for it. Here is the framework I would run today:

What data does Amazon need and retain? The rate is not the cost. The rate plus the data exposure is the cost.

Before any commercial conversation, map exactly what data Amazon ingests across forecasting, transportation, inventory and fulfillment, and what Amazon retains and aggregates outside your account. Patrick Tyler Brown at Raymond James flagged this as the central shipper concern in his post-launch note.

The contract language matters more than the rate sheet.

What is the exit cost in 24 months?

Every logistics decision has a switching cost. ASCS’ s switching cost is structurally higher than a typical third-party logistics provider( 3PL) because the data dependency compounds. Once Amazon’ s forecasting models are tuned to your demand pattern, leaving means rebuilding that capability internally or with another partner.

If you cannot articulate a clean 24-month exit, you do not have a contract. You have a dependency.

Does ASCS make sense for your network?

Three categories where the answer is yes: long-tail small and medium-sized lanes where 3PL service is uneven, and volume does not justify direct carrier contracting; regional fulfillment positioning where Amazon’ s 200-plus US nodes create a service tier you cannot replicate; and specific intermodal corridors where Amazon’ s 24,000 containers provide density your intermodal marketing company can’ t match.

Three categories where the answer is no: lanes with enough volume to negotiate directly with the steamship line, Class I railroad, or lessthan-truckload( LTL) carrier; lanes that touch your most strategic customer relationships; and lanes where your demand data is itself a competitive advantage in your industry.

What hybrid model protects optionality? The decision is not binary. Most shippers should run a hybrid. The mistake is going all in on rate compression and discovering at the next contract cycle that you no longer have a benchmarking baseline outside the Amazon network.

LTL is a useful test case. Amazon Freight already operates inbound LTL into its own fulfillment network and is expanding shipper outreach. FedEx Freight spun off as an independent public company as of June 1. Shippers who lock LTL into ASCS without evaluating FDXF will be repricing that decision in 2027.

Is your procurement organization equipped to make this decision?

The honest answer for most teams is“ no.” Procurement organizations were built around request-for-proposal cadence, contract management, cost-per-mode benchmarking, and supplier scorecards. None of those tools price the data-exposure trade.

A real ASCS evaluation needs three things most teams do not have: a data governance reviewer with authority to veto data terms, a contract attorney with experience in cloud-style master service agreements rather than transportation contracts, and a scenario modeler who can run the eight-year dependency curve, not just the 12-month rate comparison.

Most thought USPS had a moat on rural deliveries. Think again. The shippers using 2010 frameworks to evaluate 2026 procurement decisions are about to repeat that mistake.

The bottom line ASCS is a real product with real network capability. For some lanes and some shippers, it will be the right answer. For others, it will be a strategic mistake dressed up as a cost saving.

The difference will not be visible on the rate sheet. It will be visible at the eight-year mark when one set of shippers has built optionality, and another set has built dependency.

CFOs are going to ask the question. You need to be ready with the framework before the question arrives.

email: paul. tonsager @ multimodalsolutions. io

22 Journal of Commerce | June 1, 2026 www. joc. com