The 3PL market has split into two camps and most operators are in the wrong one.
On one side, a small group of tech-enabled fulfillment operators is building software moats. Stord, fresh off its January acquisition of Shipwire from CEVA, now sits on top of seven acquisitions in four years. Cart.com keeps adding volume and capabilities. Radial, the veteran of the group, launched Fast Track in March 2025 and has signed 40+ enterprise brands in twelve months according to CEO Tom Schmitt.
On the other side is every 3PL that couldn’t compete on technology. UPS sold Ware2Go to Stord. CEVA sold Shipwire to Stord. Shopify flipped Deliverr to Flexport, where it was absorbed into freight forwarding. Flexe and Flowspace still exist but they’re running niche plays in temp warehousing and project overflow, not mainstream commerce fulfillment. The 4PL pure-play category effectively collapsed in 24 months.
The tempting conclusion for a 3PL boardroom right now is that it’s time to invest in tech. That conclusion is correct. The follow-on conclusion, that the 3PL should hire developers and build it, is the mistake that will kill most of them.
A picture of the market
The dynamic is easier to see as a 2×2.
Operational excellence runs up the vertical axis. Technology focus runs across the horizontal. Four quadrants, one winning position.
STRUGGLING sits bottom-left: low operational capability paired with slow or missing technology. These 3PLs don’t make enterprise shortlists.
FAST BUT LIMITED sits bottom-right: strong tech, weaker operations. ShipBob, ShipMonk, LVK. They win mid-market deals on speed to value and ease of integration. Operational depth catches up as they scale, or it doesn’t and they get acquired.
POTENTIAL is top-left, and this is the quadrant most operators reading this piece sit in. Traditional best-in-class 3PLs. Decades of operational excellence. Real warehousing discipline. Proven service levels. And no commerce technology to speak of. These are often the best operators in the industry, and they are losing enterprise deals every quarter to tech-enabled competitors with lower operational scores.
WINNING is top-right: Stord, Cart.com, Radial. High operational capability paired with commerce-native technology. These operators show up with both halves of the equation on day one.
The hard truth for a best-in-class 3PL reading this: operational excellence alone isn’t the ticket anymore. The POTENTIAL quadrant used to be a defensible position. It isn’t now. Brands pick the WINNING quadrant even when the POTENTIAL quadrant has better warehouses, tighter labor, and stronger SLAs. Tech is the difference. The gap is widening every quarter. A best-in-class 3PL that doesn’t move right on this chart loses.
The build-it-yourself math doesn’t work
There are maybe twenty 3PLs in North America with the scale to justify a real engineering team. Everyone else is kidding themselves. And even for that top twenty, the math is ugly.
Stord has made 50+ product and engineering hires across 2025 and early 2026 on top of an existing team. Cart.com has been hiring engineers for years. These companies treat software as the product, not as a cost center bolted onto a warehouse. They’re funded explicitly to widen the technology gap every quarter.
A 3PL at position 50 or 250 that decides to hire developers is committing to compete with that math: five engineers against five hundred, on commerce integrations that get rewritten every quarter, against a moving target funded by private capital specifically to stay out of reach.
A small team can ship a Shopify connector, maybe update it once a year, and build a TikTok Shop connector when enough prospects ask for one. The 3PL will always be behind, always reactive, and always explaining to the next enterprise prospect why the feature they need is on the roadmap.
The deeper problem: you don’t know what to build
There’s a worse issue than pace. A 3PL building its own integrations doesn’t know what to build next.
A prospect walks into the pipeline next month running Acumatica. Maybe they’re the first Acumatica shopper the 3PL has ever seen. The 3PL has a Shopify connector and a NetSuite connector in production. The prospect needs Acumatica connected and live in six weeks, or they’re going with a competitor.
The 3PL has two bad options. Tell the prospect “we’ll build it” and burn eight to twelve weeks of developer time, during which the prospect goes shopping somewhere else. Or tell the prospect “we don’t support Acumatica” and watch the deal walk out the door the same afternoon.
Now multiply that across every system a real enterprise stack contains. Acumatica, Sage, Microsoft Dynamics, SAP, Oracle Fusion, Zoho, Odoo, plus the long tail of custom-built ERPs that mid-market brands still run. TikTok Shop, Faire, Walmart Marketplace, Amazon Vendor Central, Amazon Seller Central. Extensiv, Logiwa, Manhattan, Blue Yonder on the WMS side. Every EDI variant that retail partners require. Every returns platform, every tax engine, every fraud tool, every subscription platform.
The 3PL doesn’t know the next prospect’s stack until the prospect walks through the door. Building integrations on demand is how 3PLs go from pitch to “we’ll get back to you” to no deal, every time.
Agentic AI makes building worse, not better
The pitch landing in every 3PL CEO’s inbox this year is that AI solves all of this. Point an agentic system at the prospect’s ERP, let it generate the connector, deploy it, move on. Integrations become free.
It’s a trap.
Generalist AI doesn’t know commerce. It doesn’t know that a dual-allocation in a split-shipment scenario means the brand ships twice and takes the margin hit. It doesn’t know the specific failure pattern of a BOPIS cancellation inside a Shopify + NetSuite stack. It doesn’t know why an EDI 856 that passes validation can still break a retailer’s receiving dock.
AI that writes net-new integration code on the fly is AI with a hallucination problem running in production order flows, where hallucinations cost real money. The failure modes are double shipments, inventory blowouts, and returns that don’t reconcile. A 3PL that builds its tech stack on generated code is one bad quarter from a major brand exit and a customer reference that kills the next ten deals.
The 3PLs that will win with AI are the ones using it to configure pre-built, proven commerce flows, not generate new ones. 100% proven beats AI-generated every time. That’s the entire game.
The “free software” trap
The tech-enabled 3PLs have a second move, and every competing 3PL and every brand needs to understand it clearly.
Stord bundles in OMS. Cart.com bundles in integration and order management. The pitch to the brand is that the technology is free when the brand signs for fulfillment. It reads like a gift. It isn’t.
The 3PL’s OMS becomes the brand’s OMS. The integration layer running every order, every inventory sync, every returns flow, every channel connection is now owned by the 3PL. When the brand wants to switch 3PLs eighteen months later because service levels dropped, pricing moved, or a better operator opened a facility closer to the customer base, the brand isn’t switching a 3PL. The brand is replatforming its commerce stack.
Rip out the OMS. Rebuild every integration. Retrain every ops person. Reconfigure every channel. Re-test every workflow. Take a peak-season risk during the rebuild. Explain the eighteen-month project to the board.
Most brands run that math and stay. The “free” software wasn’t free. It was the entire switching cost priced at zero today and charged later, in a currency the brand doesn’t get to negotiate.
For competing 3PLs, the implication is uglier. Once a brand adopts Stord’s Order Management System, that brand isn’t a Stord fulfillment customer anymore. The brand is locked into Stord’s commerce infrastructure. No competing 3PL wins that account on operations alone, no matter how much better the warehouses, the labor, or the pricing. The operational decision and the commerce stack decision have been permanently merged.
That’s the design. That’s why it’s worth it for Stord and Cart to give the software away. They’re not in the software-licensing business. They’re in the customer-lock-in business, and free software is the cheapest lock on the market.
What 3PLs should actually do
Invest in tech. Buy it. Don’t build it.
The 3PLs that will win the next decade buy their commerce technology from a commerce-native platform and focus engineering budget on the parts of the business that actually compound: warehouse automation, network design, operational excellence. Integration, order orchestration, channel connectivity, and self-service configuration tools arrive pre-built, commerce-specific, and already connected to the 300+ systems that real enterprise stacks run on.
That’s the Pipe17 thesis, and it’s why leading 3PLs partner with Pipe17 instead of hiring developers. The 3PL doesn’t become a software company. The brand gets the same day-one connectivity it would get from Stord or Cart, without the lock-in. The 3PL competes on operations, service, and geography, which are the things the 3PL can actually win on, instead of trying to out-engineer a venture-funded competitor with 500 engineers and a two-year head start.
The alternative is the position every non-top-20 3PL will find itself in within eighteen months: slower to integrate than the tech-enabled operators, stuck explaining why the ERP or channel the prospect needs is on the roadmap, watching deals walk out the door, racing to the bottom on price, and eventually selling the business to one of the consolidators at a multiple no one on the cap table wants to talk about.
What brands should watch for
Two questions on any 3PL evaluation separate the investable partners from the dead-end ones.
First, when the 3PL is asked “do you connect to my ERP and my channels?” the answer matters. “We’ll build it” doesn’t count as a yes. It’s a maybe in eight to twelve weeks, which counts as a no for any brand with a go-live date.
Second, if the 3PL bundles in free OMS, store ops, or integration software, understand the exit cost before signing. Stickiness that comes from the 3PL owning the brand’s commerce stack is stickiness bought at the brand’s expense, not the 3PL’s.
The bottom line
The 3PL technology arms race is real and the gap is widening every quarter. The operators who win the next cycle will be the ones who understand what they’re actually competing on.
Hiring five engineers to catch a competitor with 500 is a losing strategy. Feeding prospects into a custom-build integration pipeline while tech-enabled operators close deals in two weeks is a losing strategy. Letting a tech-enabled competitor install their software inside a prospect’s stack because “free” looked like the right answer is a losing strategy.
Buy the tech. Compete on operations. Win the customer.
