Burq bets on the last-mile decision layer as giants retreat

The national parcel carriers that once set retail delivery economics have spent a decade shedding their least profitable packages. What filled the gap is a deeper bench: regional carriers, 3PL cross-dock networks, gig courier platforms and private fleets. As it turns out, almost none of them talk to each other.

Burq, a last-mile delivery technology company, wagers that last-mile orchestration — the decision layer sitting above that bench — is the product enterprise retailers will actually pay for. The shift is already underway.

FreightWaves reported in July that carrier diversification is eroding the last-mile delivery duopoly: 55% of retailers now use carriers outside FedEx, UPS and the U. S. Postal Service, and more than a third are actively moving volume away from the two national giants. The payoff is measurable.

In one example, home goods brand Caraway cut total parcel costs 20% after its 3PL began shopping every order across a network of regional and national carriers. Capturing that value requires knowing which provider performs in which ZIP code on which day.

Most retailers do not know, because the systems that pick a carrier stop watching the moment the order leaves. A deeper bench, and no referee Jake Stein joined Burq four months ago to run growth for its retail sector, after four and a half years at Uber launching ship-from-store, same-day and on-demand programs on Uber Direct.

The limitations of using a single network are what pushed him out. “It still came down to that Uber couldn’t be the single source for everything,” Stein told FreightWaves. “Mostly gig, reasonably shorter distances, some batching but not large scale, no big and bulky. And they can’t be 100% of every possible location.”

The fallback pitch was to add providers, which created a new challenge. “Our sales pitch sort of sometimes fell on ‘well, you should have more than one provider,’ and then what does that mean?” he said. “How does that get orchestrated? What intelligence is behind making those decisions?”

“There’s been a movement probably over the last at least 10 years now where more and more, especially enterprise-scale companies, are looking at not just one or two or even three national providers,” Stein said. “It’s a whole host of regional providers just for their parcel.”

Layered with 3PL cross-docking and a private fleet, the delivery promise starts to close on Amazon. The management of it does not. “How do you actually optimize all of that through one source is sort of the ideal flow of what we’re trying to help merchants get to,” Stein said.

Where last-mile orchestration starts: After the OMS hands off Large retailers already run distributed order management through platforms such as Manhattan or IBM Sterling. Those systems decide where an order ships from based on inventory location, distance, cost and the service levels a retailer wants to promise. They are also expensive.

Companies that run one start a large step ahead, Stein said, “compared to a lot of the mid-market or small companies that can’t afford those enterprise multi-year builds into their systems.” The gap Burq is targeting opens the moment that decision is executed.

“Once things leave the OMS and they’ve gone out into, let’s say, the parcel carrier or to your own fleet to deliver, even to a third-party gig platform that is like an Uber or a DoorDash,” Stein said, “they don’t know what happens after that. There’s no constant monitoring of what’s going on.”

A package promised in two days that has not been scanned at the pickup point is, in most operations, a customer service call waiting to happen.

“We could say, you know what, we’re going to redispatch or cancel the order with the first courier/carrier and send it to the other carrier who’s a better performance level and we know is more than likely to pick it up,” Stein said.

“There’s really nothing out there that exists that can orchestrate across all of those different nodes and modes and SLAs,” he said. “That’s really the big enterprise solution that we are working to build.” Real-time orchestration: Nine minutes, then reassign Stein’s working example runs on a nine-minute threshold.

“We know based on weather, traffic, history of the providers in the area that provider A typically gets there to pick it up within nine minutes, but nine minutes has passed and it hasn’t even been assigned to a driver,” he said. “We just immediately take it and reassign it to the next available courier.” The replacement courier may cost more.

That is a policy question, rather than a technical one, and it can cut against volume commitments already on the books. “Maybe you’ve had a commitment to a specific carrier. You have to give them a certain number of packages otherwise you get penalties on costs,” Stein said. “So that could all be factored into the decision-making process.”

Adoption runs in stages: background recommendations showing what last month could have saved, then real-time prompts, then full automation “once yo