Boscov’s E-Commerce Fulfillment Reset: How to Measure a Store-and-DC Network Redesign

Boscov’s is resetting the operating foundation of its e-commerce network. The department store retailer is implementing a unified order and inventory platform intended to connect inventory data and processes across its fulfillment operation. The technology matters, but the real test is operational: can the redesigned network make a more reliable customer promise at a sustainable cost?

That question cannot be answered with a single before-and-after delivery metric. A retailer must separate inventory placement, node execution, transportation performance, and customer outcomes. Otherwise, one improvement can conceal another problem—for example, faster picking achieved by sending more multi-item orders from separate locations.

What Boscov’s is changing

Supply Chain Dive reports that Boscov’s is using Manhattan Associates’ Active Order Management platform to unify inventory management data and processes. CEO Jim Boscov said the company expects the system to reduce out-of-stocks and improve customers’ omnichannel experience.

That is more than a software installation. A unified inventory view can change which node receives an order, whether a store or distribution center fulfills it, and how inventory is reserved across channels. It can also alter the number of packages, carrier services, and handoffs behind one checkout promise.

Retailers already have substantial experience using stores as fulfillment nodes. A Supply Chain Dive report cited a Zebra Technologies study in which 76% of retailers said they were already using stores for fulfillment, while 91% of merchants expected more retailers to do so. The opportunity is real, but store fulfillment adds variability that a conventional DC-centric scorecard may miss.

Separate placement from execution

The redesign should be measured as four connected layers.

Inventory placement asks whether the right products are positioned near likely demand. Track inventory accuracy by node, in-stock rate at order promising, forecast error, and the percentage of orders that can be filled completely from one location. If this layer is weak, the order system will merely make better-informed compromises.

Node execution measures what happens after routing. Track release-to-pick time, pick duration, item-not-found rate, pack completion, cancellation after allocation, and orders ready before carrier cutoff. Stores and DCs should have separate baselines because their labor, layouts, and primary missions differ.

Transportation execution begins at tender. Measure first carrier acceptance, pickup compliance, origin dwell, transit time, delivery exceptions, and first-attempt delivery. These events show whether an order that left the building on time actually reached the customer as promised.

Customer outcomes connect the layers. Track promise accuracy, complete-order delivery, cancellation rate, returns tied to fulfillment errors, and customer contacts per order. A faster average is not a win if variability or partial deliveries increase.

Establish the baseline before changing routing

Three measures deserve special attention.

1. Order cycle time

Measure from order acceptance to final delivery, then retain timestamps for allocation, release, pick, pack, tender, pickup, and each carrier event. Report the median and the 90th or 95th percentile, not only the average. The tail reveals nodes, products, or carrier lanes that routinely miss the promise.

Also distinguish “first package delivered” from “complete order delivered.” The first can improve while the customer waits for the remaining item.

2. Split-shipment rate

Calculate the percentage of customer orders dispatched in more than one package or from more than one node. Then measure packages per order and incremental freight cost. Category relationships matter: one analysis highlighted by Supply Chain Dive found that 76% of multi-item orders in the studied dataset crossed product categories. That is a strong argument for placing items based on what customers buy together, not only SKU velocity.

A routing engine can reduce distance while increasing splits. The scorecard should therefore compare avoided miles with added packaging, handling, and parcel charges.

3. Fully loaded cost per order

Include pick and pack labor, materials, internal transfers, parcel or freight charges, accessorials, refunds, reships, and fulfillment-related customer service. Compare cost by node type, service level, destination zone, and order profile. An apparently expensive DC may still be the better node for large baskets if it avoids two store shipments.

Use shipment events as the proof layer

Inventory and order systems decide what should happen. Shipment events establish what did happen. The retailer needs one record linking the customer order, fulfillment node, packages, tenders, carrier tracking numbers, costs, exceptions, and final delivery.

That record enables useful comparisons: promised ship date versus actual pickup, planned service versus tendered service, estimated versus actual transportation cost, and promised delivery versus the final package event. Exception codes should identify whether misses originated in inventory accuracy, store or DC execution, carrier capacity, weather, address quality, or customer availability.

Run the redesign as a controlled rollout. Establish at least several weeks of baseline performance, activate new routing rules for defined nodes or order segments, and compare like-for-like cohorts. Protect the customer with guardrails: cap the number of fulfillment nodes, suppress locations with poor inventory accuracy, and avoid routing orders too close to carrier cutoff.

The most useful executive view is a balanced scorecard: complete-order promise accuracy, split rate, fully loaded cost per order, cancellation rate, and 95th-percentile cycle time. If one improves at the expense of the others, the team can identify the tradeoff rather than declaring victory too early.

Turn network decisions into measurable execution

Boscov’s unified platform can create a stronger inventory and order foundation. The lasting advantage will come from connecting those decisions to shipment-level evidence across stores, DCs, and carriers.

CXTMS gives logistics teams a common execution record for orders, tenders, tracking events, exceptions, and freight costs. Request a CXTMS demo to see how measurable transportation workflows can support a retail fulfillment redesign.