How Routing Decisions Shape Dark Store Network Economics for North American Retailers

A VP of Supply Chain at a North American grocery retailer reviews the dark store network plan with the executive team. Real estate has identified eleven candidate sites across two metropolitan areas. Network design consultants have modelled location coverage and inventory distribution. The technology team has begun evaluating warehouse management and pick automation. Finance has built unit economics for each site. The capital request is ready for the board.

The routing layer that will actually operate the network — store-to-order assignment, multi-origin dispatch, inventory-aware decisions, capacity management, surge response — is treated as a downstream implementation detail to handle after the network is built.

This is the most common pattern in North American dark store planning, and it leaves a substantial share of network economics on the table. Dark store conversations focus heavily on location selection and inventory distribution — both genuinely hard problems served by specialist software ecosystems. The routing layer that operates the network once it exists is structurally undercovered. The decisions made at the routing layer — which store fulfils which order, how driver capacity flexes across stores, how inventory state shapes routing in real time, how the network responds to surge — produce a meaningful share of total network economics. Operators treating routing as a first-class design input materially outperform those treating it as an afterthought.

This is a strategic guide for North American VP Supply Chain leaders, Heads of E-Commerce Operations, and Heads of Logistics evaluating dark store network strategy.

The North American Dark Store Reality

Before discussing routing, market context matters. The fifteen-minute delivery category that dominated dark store conversations in 2021–2022 has largely consolidated. Pure-play q-commerce operators including Gorillas, Jokr, Buyk, and Getir exited or scaled back US operations through 2022–2024; Gopuff retreated significantly from earlier expansion. The category survived but reshaped.

The dominant North American fulfilment model in 2026 is one-to-two-hour grocery and convenience delivery — Amazon Fresh, Whole Foods, Instacart partnerships, big-box programs (Walmart, Kroger, Target), and surviving q-commerce operators that pivoted to longer SLA windows. In-store fulfilment — store associates picking from regular store shelves — accounts for the majority of fast grocery delivery in the US. Pure dark stores still exist as one channel within a multi-channel mix.

According to Bain & Company research on retail fulfilment economics, the strategic question for most North American retailers is not whether to operate dark stores but how to optimise the multi-channel mix between in-store fulfilment, dark stores, regional micro-fulfilment, and centralised distribution. Dark store networks live within this mix, and routing decisions shape how well the network performs in context.

Five Routing Decisions That Shape Dark Store Economics

1. Store-to-Order Assignment

In a multi-store metro, every order presents an assignment decision: which dark store fulfils it? The default — assign by zip code, or always the closest store — leaves value on the table because the closest store may be out of stock, capacity-saturated, or operationally suboptimal.

Smart store-to-order assignment considers stock availability per store, pick capacity remaining, batching opportunities with nearby orders, and current driver capacity. The differential between dumb and smart assignment in dense metros is real and measurable in delivery cost per order. According to McKinsey & Company research on grocery delivery economics, last-mile cost variation across operational decisions can produce significant per-order economic differences, particularly at scale.

2. Multi-Origin Dispatch Routing

A network of fifteen dark stores produces fifteen origin points, not one. The routing problem is fundamentally different from single-distribution-centre dispatch. Each store has its own driver pool, demand profile, and capacity dynamics — but network economics depend on how well the system balances across them.

Most legacy routing systems were architected for single-origin operations and adapted to multi-origin contexts through configuration rather than redesign. The multi-origin Vehicle Routing Problem (VRP) is computationally harder than single-origin VRP, and routing engines built natively for it produce materially different operational outcomes than systems retrofitted for it.

3. Inventory-Aware Routing

Routing decisions need real-time inventory visibility per store. When an order arrives for a specific SKU, the system needs to know which stores have it in stock, which have pick capacity, and which combination produces the best delivery economics — at the moment of routing, not from inventory data refreshed four hours ago.

Most dark store operations run inventory updates on cycles measured in hours rather than minutes. Routing decisions made against stale inventory produce stockout experiences customers see, late substitutions, and operational rework. The integration depth between routing engines and inventory or warehouse management systems matters more than dark store vendors typically discuss.

4. Capacity-Aware Dispatch

Every dark store has a finite pick capacity per hour. Saturation at one store while a nearby store has slack is system-level waste — and capacity-blind routing produces this pattern systematically. Popular stores get overloaded; underutilised stores stay underutilised; SLA failures concentrate at the busy stores.

Capacity-aware dispatch routes orders to stores with available pick capacity rather than just to the geographically optimal store. The routing engine’s awareness of capacity state changes operational outcomes during peak periods substantially.

5. Surge Orchestration Across the Network

Promotional spikes, weather events, and viral product moments hit the network unevenly. Some stores absorb surge volume; others remain idle. Cross-store rebalancing through routing happens faster than physical inventory rebalancing, making the routing layer the operational lever that responds to surge in real time.

Operators who treat surge response as a routing-layer problem absorb peak periods with less SLA degradation and lower exception cost. Those who don’t typically experience peak as a recurring crisis their operations team manages through escalation calls.

Composite Illustrative Scenario

Consider a mid-size grocery retailer running fifteen dark stores across a major metropolitan area — Chicago, Boston, or the San Francisco Bay Area — with daily order volume around three thousand, on a ninety-minute delivery promise.

Without routing as a first-class layer. Store-to-order assignment runs static by zip code. Each store dispatches its own drivers without cross-store coordination. Inventory data refreshes every four to six hours. Surge response is manual — when Black Friday or a weather event hits, the operations team works through escalation calls and ad-hoc reassignments. Capacity utilisation imbalances persist across stores, with popular locations saturated and quieter ones underused. SLA hit rates sit at the lower end of operationally acceptable; delivery cost per order trends higher than the network’s potential.

With routing as a first-class layer. Store-to-order assignment runs dynamically per order, considering inventory state, capacity remaining, batching, and delivery cost. Multi-origin dispatch routes drivers across the metro with system-level optimisation. Inventory integration runs continuously rather than on hours-long refresh cycles. Surge response is automated — capacity-aware routing diverts orders before saturation produces SLA failures. Capacity utilisation balances across stores. SLA hit rates improve into the higher operational range; delivery cost per order trends toward the network’s economic potential.

The qualitative difference is meaningful. The quantitative difference depends on metro density, network configuration, demand patterns, and inventory dynamics — and varies enough across operators that publishing a single number would mislead.

Where the Routing Layer Lives Architecturally

Routing platforms like Locus that handle multi-origin dispatch, inventory-integrated routing, capacity-aware decisions, and cross-network surge orchestration as core capabilities provide the routing-layer architecture dark store networks need to capture full economics. The architectural choice — routing platforms purpose-built for multi-origin operations versus single-origin systems adapted to dark store deployment — determines whether the routing layer produces network value or leaks it.

According to INRIX congestion data, urban routing in dense North American metros faces variable conditions that further reward routing architectures with continuous reoptimisation rather than static assignment. Routing decisions that matter most for dark store economics are made many times per hour, against changing inventory, capacity, traffic, and demand state.

The Real Question for Supply Chain Leaders

Dark store network design is mostly about location and inventory. But the routing layer that operates the network produces a meaningful share of total economics — and most North American operators under-invest in it relative to its impact. The strategic question is: given the network we have or are building, is the routing layer a first-class design input or an afterthought we’ll handle later?

FAQs

What is a dark store and how does it differ from in-store fulfilment?
A dark store is a retail location dedicated exclusively to online order fulfilment, with no walk-in customers — picking, packing, and dispatching online orders within tight delivery windows. Dark stores typically locate within three to five miles of dense customer clusters to support fast delivery promises. In-store fulfilment, by contrast, uses regular retail stores where store associates pick online orders from the same shelves serving walk-in customers. Walmart, Kroger, and Target run substantial in-store fulfilment operations across the US, while Amazon Fresh, Gopuff, and grocer-specific programs operate dark stores. Most North American retailers run a mix of both, with dark stores serving urban density and in-store fulfilment serving broader geographic coverage.

Why is the routing layer undercovered in dark store conversations?
The routing layer is undercovered in dark store conversations for two reasons. First, network design (where to locate stores) and inventory distribution (which SKUs go where) are visible, capital-intensive decisions with specialist software ecosystems and consulting practices around them — they get the attention they deserve. Second, routing is often treated as a downstream implementation detail to be handled after the network is built, rather than as a design input that shapes network economics. This sequencing produces operationally suboptimal outcomes: networks designed without routing-layer thinking often need to be retrofit for capacity-aware dispatch, inventory-integrated routing, and surge orchestration after launch — at higher cost and lower performance than designing for these capabilities upfront.

What is multi-origin dispatch routing and how does it differ from single-origin? Multi-origin dispatch routing is the routing problem of dispatching orders from multiple origin points — typically five to fifteen dark stores in a metropolitan area — rather than from a single distribution centre. Single-origin routing solves a Vehicle Routing Problem with one starting point; multi-origin routing solves a more complex optimisation across multiple starting points, considering which origin should fulfil each order alongside the route from origin to destinations. The multi-origin VRP is computationally harder than single-origin VRP. Most legacy routing systems were architected for single-origin operations and adapted to multi-origin contexts through configuration rather than redesign — producing different operational outcomes than routing engines built natively for multi-origin.

Why does inventory-aware routing matter for dark store economics?
Inventory-aware routing matters because routing decisions made against stale inventory data produce stockout experiences customers see, late substitutions, and operational rework. When an order arrives for a specific SKU, the routing system needs to know which stores have it in stock, which have pick capacity, and which combination produces the best economics — at the moment of routing, not from inventory data refreshed hours ago. Most dark store operations run inventory data updates on cycles measured in hours, but routing decisions are made on cycles measured in minutes. The integration depth between routing engines and inventory or warehouse management systems determines whether routing decisions reflect operational reality or operate against stale data.

How should North American retailers evaluate dark store network strategy?
North American retailers should evaluate dark store network strategy by recognising that location and inventory are necessary but not sufficient. Beyond location selection and inventory distribution, retailers should evaluate the routing layer that will operate the network: store-to-order assignment logic, multi-origin dispatch architecture, inventory integration cadence, capacity-aware dispatch capabilities, and surge orchestration mechanisms. The routing layer should be a design input to network strategy rather than an implementation detail handled afterwards. Retailers should also recognise that pure dark stores are one option within a multi-channel mix that typically includes in-store fulfilment, regional micro-fulfilment, and centralised distribution — and that the routing layer needs to operate across this mix rather than within dark stores alone.

What’s the difference between dark stores in North America and Europe?
Dark stores in North America and Europe differ across several dimensions. Urban density: European cities are typically denser, supporting smaller dark store catchment areas (one to three kilometres) and more locations per metro. North American urban sprawl requires larger coverage areas and fewer, larger stores in many cases. Channel mix: European q-commerce operators (Getir UK, Flink Germany, Gorillas before exit) ran pure dark store models more aggressively than most North American operators, who lean more on in-store fulfilment from existing retail footprint. Real estate: dark store-suitable industrial spaces are particularly scarce in dense European urban cores. Regulatory: European labour rules and zoning restrictions affect dark store operations differently than US regulations. The strategic question for North American retailers is rarely “follow the European q-com playbook” and almost always “what’s the right multi-channel mix for our market.”