Amazon is not a customer of the parcel industry anymore. It is a competitor that still buys a little capacity on the side.
An internal Amazon planning document reviewed by Business Insider, reported September 3, puts hard numbers on that. Supply Chain Dive picked it up September 9. Amazon's own network is projected to handle 86.3% of its US packages next year. By 2029 that figure reaches 88.7%.
The share left for everyone else is the part worth reading twice. USPS is projected at 10% of Amazon's volume in 2027, falling to 8% by 2029. UPS is projected at 1.8%, falling to 1.4%.
Amazon spokesperson Steve Kelly told Supply Chain Dive the projections are preliminary and subject to significant revision. Fair enough. The direction has been consistent for two years, and the last set of internal projections was lower: 83.8% for 2027 and 85% for 2028. Amazon revised the plan upward, not down.
Why this is not just an Amazon story
Amazon already passed USPS on volume. ShipMatrix data cited in the same report put Amazon at 6.7 billion US packages in 2025, ahead of USPS at 6.6 billion and UPS at 4.4 billion.
When the largest shipper in the country pulls volume out of a network, the network does not shrink neatly. Fixed costs stay. Routes that were dense because Amazon filled them get thinner. The remaining shippers on those routes carry more of the overhead.
UPS said it would cut its Amazon business by more than half by mid-2026, and that decision was UPS's own. USPS went a different way. It negotiated Amazon's proposed volume reduction down from 67% to roughly 20%, which tells you how much the Postal Service wanted to keep those packages.
Two carriers, opposite reactions, same underlying pressure. Neither one ends up with cheaper economics per package.
What this means for your brand
If you ship 1,000 to 25,000 orders a month, none of this shows up as a headline on your invoice. It shows up as drift.
Here is where to look.
Your ground rates in 2027 negotiations. Carriers replacing Amazon volume will chase mid-market shippers harder in some lanes and price defensively in others. That cuts both ways. A brand with clean, predictable volume out of a single origin has more leverage in that conversation than it did two years ago. A brand splitting 4,000 orders across scattered warehouses has less.
Zone density. The reason carrier costs fall when your warehouse sits near your customers has nothing to do with Amazon. It has to do with how far each package travels. Our Canal Winchester facility reaches 60% of the US and Canadian population within 600 miles, and 90% of customers within two days by ground. That math holds no matter which carrier is winning the volume war.
Carrier diversification. Brands that route everything through one carrier are exposed to that carrier's strategy. Brands that can move a package between USPS, UPS, DHL, and a regional option based on weight and zone are not. This is a routing question, and it is one of the practical reasons brands hand fulfillment to a 3PL with multiple carrier accounts instead of managing four relationships in-house.
Amazon channel costs specifically. If you sell on Amazon, Amazon's delivery network is now most of your delivery network on that channel whether you chose it or not. That makes your FBA prep accuracy more important, because rejected or mislabeled inbound freight is one of the few parts of that chain you still control.
The opinion part
The parcel duopoly that shaped DTC shipping economics from 2015 to 2022 is over, and it is not coming back. The replacement is a market where the biggest player mostly delivers its own packages and the national carriers fight over what is left.
That is better for mid-market brands than most people expect. Carriers that lost their anchor tenant have to win volume from someone, and brands with disciplined operations and consistent daily counts are exactly who they have to win it from.
The brands that get hurt are the ones with messy volume: unpredictable spikes, bad address data, high dimensional weight per order, inventory scattered with no zone logic. Those brands were quietly subsidized by a carrier market flush with Amazon revenue. That cushion is going away.
FAQ
Does this mean USPS rates go up for small shippers?
Not automatically, and not because of this document. USPS prices through the Postal Regulatory Commission on its own schedule, and its 2026 peak rates were filed months ago. The longer-term concern is that a network losing 20% of its largest customer's volume has fewer packages to spread fixed costs across. Watch the 2027 and 2028 filings, not this week's.
Should I move volume off UPS?
No, and that would be the wrong lesson. UPS is choosing to trade low-margin Amazon volume for higher-margin business, and mid-market DTC is part of the business it wants. That is a reason to have the rate conversation, not a reason to leave. What you should not do is stay single-carrier. Route by weight and zone, and keep more than one account live.
Is Amazon giving up outside capacity entirely?
No. The same document projects 87.4% self-delivery in 2028, which leaves roughly an eighth of a very large number with outside carriers. Amazon keeps third-party capacity for the addresses and the surge days its own network handles poorly. The volume is shrinking as a share, not disappearing.
How does a 3PL change any of this?
Volume aggregation and routing. A 3PL negotiating on behalf of many brands prices differently than one brand at 5,000 orders a month, and a shared warehouse means your packages start from a location chosen for zone reach. If you want to see what your specific mix looks like from central Ohio, send us your numbers.