Target builds faster hybrid system for digital order allocation at scale

A faster router for digital orders

Target is trying to make one of retail’s messiest jobs, deciding where every digital order should come from, look a lot less messy. Its new hybrid allocator weighs inventory, delivery speed, shipping cost and packaging constraints in seconds, then picks where items should be sourced, how they should be grouped and how fast they should reach the guest. The big claim is striking: it closes about 91% of the fulfillment-cost gap between brute-force and greedy approaches while using only about 8% of brute-force compute time.

For store teams, that matters less as an engineering trophy than as a daily labor question. If the allocator does a better job of placing the right order in the right location, the floor sees fewer last-minute reroutes, fewer split shipments and fewer moments when a team member has to explain why a guest’s order arrived in pieces instead of one clean box.

What changes when the promise gets smarter

The easiest way to think about this system is as a traffic cop for Target’s digital orders. Instead of simply chasing speed at any cost, it tries to balance promise accuracy with the realities of inventory, packaging and shipping. That can reduce the kind of chaos that forces team members to re-pick, re-pack or field complaints when a guest’s order was promised too aggressively and then broken apart across multiple boxes.

That shift could help in three everyday ways. First, it may lower the number of unnecessary substitutions, because better routing should make it easier to source from the most suitable store or node. Second, it can trim wasted labor when teams are not chasing unrealistic fulfillment paths. Third, it may smooth the guest conversation, because a more accurate promise means fewer apologies, fewer delays and fewer surprises at pickup or delivery.

Why Target cares so much about this now

The scale behind the problem is huge. Target said it has built a $20 billion first-party digital business, and same-day services like Drive Up and same-day delivery with Target Circle 360 grew at a double-digit pace in 2024. The company also said more than 65% of digital sales were fulfilled through same-day options in fiscal 2024, while same-day services represented more than 10% of total sales and grew 13.6% in one reported quarter.

That makes allocation logic more than a backend optimization. It is one of the hidden systems supporting loyalty and repeat spending, especially when Target says active Target Circle members spend three times more on average than nonmembers and Target Circle 360 subscribers spend eight times more. When guests pay for convenience, they expect the promise to be right. When the promise is wrong, the friction lands on store teams first.

The store network is the real engine

Target’s fulfillment model works because its stores are not just sales floors. The company says it has about 2,000 stores across the United States, more than 75% of the U.S. population lives within 10 miles of a Target store, and more than 30 million guests shop its stores each week. Target also says nearly all of its sales, including digital, were powered by its stores last year.

That is why digital order allocation hits the workday so directly. Most digitally originated sales are not flowing through a separate e-commerce-only network. They are being handled by the same stores where team members are already stocking shelves, serving guests, prepping pickup orders and keeping the front end moving. Better routing can mean fewer rushed picks and fewer cross-functional disruptions, while poor routing can turn a normal shift into a scramble.

What this could mean for team leads and ETLs

For team leads and executive team leaders, the practical upside is predictability. If the allocator is making better calls up front, there should be less backtracking later, fewer emergency moves to recover a bad promise and less unnecessary labor burned on orders that were never efficient to begin with. In a store environment where digital volume can change quickly, even small gains in promise accuracy can make the difference between a controlled shift and a constant series of interruptions.

There is also a morale angle that gets overlooked. Teams tend to feel fulfillment pain when they are asked to move fast without a realistic plan, especially if that speed creates more guest frustration downstream. A system that reduces split shipments and improves on-time accuracy can make the work feel more orderly, even when the volume is still high.

How this fits Target’s broader supply chain buildout

The allocator is not arriving in a vacuum. Target executives said in a 2023 supply chain discussion that over less than 10 years the company created, acquired and advanced sortation centers, upstream distribution centers, food distribution centers and logistics innovation. That long buildout suggests the company is trying to make its stores, supply chain and digital operations act like one connected system instead of separate pieces.

Target has also said that lower digital fulfillment costs have partly come from changes in the mix of digital sales fulfilled through lower-cost same-day services. The new hybrid allocator pushes in the same direction, but with a more sophisticated decision engine that aims to preserve speed without wasting compute or forcing bad routing choices. That matters because Target’s 2024 earnings release said gross margin improvements were partly offset by higher digital fulfillment and supply chain costs.

Who benefits most if it works

The biggest winner should be the guest, if orders arrive on time, in fewer pieces and with fewer substitutions. But the people who feel the change most immediately are the teams in stores and supply chain, because they are the ones absorbing the cost of bad routing, late promises and avoidable rework. If the system really rebalances speed, cost and accuracy, it can reduce friction on the floor instead of just moving it around.

That is the real test of this kind of AI at Target. Not whether it sounds sophisticated, but whether it makes a digital order easier to fulfill, a shift easier to run and a promise easier to keep. If those three things improve together, the algorithm is doing more than saving money. It is changing what the work feels like.