Four years ago, the formula for retail success was shifting from operating big stores and a large physical network to providing customers with multiple touchpoints, building local density, and using data to encourage shoppers to spend more with a single brand.
Since then, individual stores have come to contribute value in very different ways, well beyond their own four walls. As AI becomes embedded in more retail workflows, real-time data on each store's contribution has become easier to obtain. That visibility has finally made it feasible to manage a retail network as a portfolio of nodes with distinct roles.
According to BCG’s newly updated Win the Town retail strategy, companies that recognize this opportunity no longer measure financial performance and allocate capital solely on the basis of the traditional store-level four-wall profit and loss (P&L). Instead, they’ve adopted an approach that accounts for the broader value that individual locations contribute to the whole.
Retailers struggling to make this shift can follow the lead of cutting-edge companies that optimize their store networks according to one of three patterns: treating locations as infrastructure that supports the broader ecosystem; designing store locations to have distinct but well-coordinated roles; or following a rigorous, disciplined path to implementing in-store technology at scale.
How AI Updates the Win the Town Retail Strategy
In 2022, BCG introduced the Win the Town approach to help companies understand how to succeed in a dynamic retail environment. At the time, growth based on large-format locations and large physical networks was being overtaken by growth from business models that prioritized customer density and local ecosystem mastery. Retailers gained an edge by serving customers across sales touchpoints, building local operating density, and using data to unlock a larger share of wallet.
Flash forward to 2026, and none of that has changed—but individual stores can now play a variety of roles in a retailer’s total network, and AI is the reason. In a relatively short period of time, AI has evolved from being tested in pilot-stage experiments to being embedded throughout retail operations. (See Exhibit 1.)
This process has significantly affected multiple areas of retail:
- Frontline Workflows. In addition to providing analytics, AI now powers real-time decisions about pricing, labor, inventory replenishment, and handling exceptions. It has shifted the basis of store operations from scheduled routines to data-based, continuously optimized real-time actions.
- In-Store Technology. The integration of AI into digital shelf labels, radio frequency identification (RFID), computer vision, integrated point of sale (POS), and Internet of Things (IoT) technology makes it easier and faster to track data and act on trends.
- Retail Media. AI-based data on store performance, traffic patterns, and audience density offers measurable value for retail media—the advertising that appears within a company’s owned physical or digital channels. As cross-channel engagement matures, physical stores are at the center of a US retail media industry that’s growing at a compound annual rate of 26%.
- Fulfillment. Retailers are improving order fulfillment and delivery by integrating AI into delivery coverage, route density, and click-and-collect throughput. They are combining that activity with efforts to partner with global logistics providers that operate strategically located and largely automated warehouses.
Why the Traditional Four-Wall P&L Falls Short
As stores’ roles change, leading retailers have altered how they evaluate individual locations’ financial performance and how they make investment decisions. The traditional four-wall P&L is still useful for enforcing accountability and driving operating discipline, but it no longer captures all of the economic value that stores create. A store that appears to produce marginal value when assessed in isolation may generate substantial value to the network by anchoring click-and-collect traffic that increases spending or drives brand awareness and thereby raises demand in sales channels that it never directly touches. A store could contribute to a higher density of nearby shoppers that makes it possible for the company to monetize retail media. Or it could serve as an inventory buffer ensuring that other stores in the network don’t leave orders unfilled because they’ve run out of stock.
When a retailer applies the same KPIs to stores with vastly different roles, stores that should receive investment on the basis of what they contribute to the network may instead be penalized because of a narrow focus on their direct margins. Conversely, stores that underperform on the network level may receive unearned benefits because of how the retailer measures the value they deliver.
The Portfolio Approach to Determining a Store’s Value
Retailers that take a portfolio approach design and manage stores in a way that accounts for each location’s role, resilience, and value to the ecosystem (see Exhibit 2):
- Role. Traditionally, retailers used a store’s size or format as a proxy to determine its role within the network. Today, however, stores can play multiple roles. A large-format store may serve as a fulfillment hub, a browsing destination for shoppers who ultimately buy online, and an advice or service center. Having a clear vision of the role or roles that a store plays in the network will inform not only its design and operation but also things such as technology investment and labor model. Leading retailers identify well-defined roles for their stores and apply relevant KPIs to each. They use metrics to translate each store’s role into decisions about its day-to-day operations, and they tie its performance to its primary financial contribution to the network.
- Resilience. Consumer preferences and expectations are evolving at such a rapid pace that volatility in demand, supply, and labor has become a given. Retailers that optimize their store networks for steady-state efficiency are at risk when conditions shift. Cutting-edge retailers design resilience into their networks through inventory pooling, rapid redeployment, dense coverage, and flexible operating models. Resilience protects margins by minimizing markdowns, improving service levels, and avoiding costs associated with disruptions.
- Ecosystem Value. Some individual locations create system-level value that flows to other parts of the business. They may contribute to delivery speed, route density, retail media reach, brand amplification, customer engagement, or data generation. In such instances, retailers should measure the value that the store generates with the same rigor they apply to value generated by direct margins, and they should take that value into consideration when making decisions about investments and closures.
This philosophy is in keeping with the Win the Town approach of prioritizing local ecosystems and density. But AI supercharges the ability to measure, act, and reallocate capital, technology, and operational attention at the network level.
How Leading Retailers Optimize Store Networks
Optimizing a retail network to account for role, resilience, and ecosystem value can follow one of three patterns, as the following examples of leading retailers illustrate.
Pattern 1: Stores as Infrastructure. Some retailers treat stores as infrastructure that supports the broader ecosystem, an approach that affects how they interpret individual location economics and justify investments. For example, Walmart views its physical footprint as last-mile infrastructure and evaluates individual store economics in the context of the entire system. The company’s vast network of stores yields pickup and delivery economics that a purely digital operator could not achieve. The network’s density leads to cost-to-serve advantages that compound as fulfillment volumes grow. In fact, about 50% of the items that the company sells online are fulfilled from stores. (See Exhibit 3.)
Walmart’s locations contribute to its retail media reach, too—a second justification for investing in store networks beyond the direct transaction volume they generate. Walmart’s global advertising business, which includes its Walmart Connect retail media network in the US, generated $6.4 billion in fiscal 2026, a sign of how much high-margin revenue a store network can produce beyond direct sales.
Alibaba measures the performance of stores in its Freshippo grocery chain (known as Hema in China) not only by traditional supermarket metrics but also with regard to their value as retail floors, fulfillment hubs, and digital engagement nodes. According to public reports, the stores register three to five times as many sales per unit area as traditional supermarkets do. More than half of all orders at the stores are online, and stores deliver in 30 minutes within a radius of roughly 3 kilometers.
Pattern 2: Different Roles in One Coordinated System. Some retailers design store networks with special attention to locations that have distinct but well-coordinated roles. Coolblue, the $2.8 billion Dutch electronics retailer, treats physical stores as amplifiers of its broader business. New stores increase online sales within the location’s catchment area. A store’s physical presence anchors trust for high-ticket online purchases: customers convert more readily because they know that they can pick up merchandise or get expert advice from a nearby store in the same network. Physical stores also serve as a convenient channel for returns, repairs, and post-purchase support, which benefits the company’s public image.
Pattern 3: Industrialized Execution. Advantages don’t always come from the most visible or innovative store concepts. Retailers can also build a durable competitive position by following a disciplined path to implementing in-store technology at scale.
How Decathlon, the French sporting goods retailer, deploys RFID in its store network offers an example of industrialized execution. RFID technology is not novel, but Decathlon applies operational discipline to its stores’ use of the technology, including to improve inventory accuracy, reduce manual counting labor, and accelerate checkout. Individual use cases are tied to a clear operational purpose and measurable economic return, yielding quantifiable improvements in productivity and available inventory.
Domino's, the US pizza chain operator, builds measuring and monitoring into every step of a simple operating model, including order accuracy, make time, delivery time, and store-level labor productivity. The company integrates and continually optimizes ordering, kitchen workflow, and delivery dispatch, thereby reducing friction and variability throughout the network, even though individual franchisees run its locations. The company tests new capabilities against core metrics and scales only those that have a clear demonstrable impact on unit economics. As a result, execution is measurable and repeatable across a large network.
Measuring and monitoring operations to support unit economics and scaling only capabilities that demonstrate a clear return are what separate retailers that build durable capability from those that generate innovation without compounding advantage. The advantage does not come from building the most futuristic store but from designing and governing an orchestrated portfolio in which each node has a defined role, a measurable contribution, and an accountable return.
Implementing a Winning Portfolio Approach
AI may have precipitated the change, but running an orchestrated store network isn’t just a matter of technology or real estate. It requires senior leaders to rethink how they define strategy, allocate capital, and govern tradeoffs across the network. The following no-regret moves can help you start shifting toward a portfolio approach in as little as one quarter:
- Ask store managers about the role that their location plays beyond selling. If the answers clash or if the head office takes a different view, you’ve identified gaps in role clarity. Fill them by giving each store one clear primary role beyond selling—even if it plays several—plus a handful of KPIs to match.
- Put network value at the same level as four-wall P&L. Capital follows numbers that leaders can see. To determine which initiatives are worthy of investment, list your last ten major store-level capital decisions. Note whether you justified them using four-wall payback or citing their contribution to the network. If most decisions fall into the first category, start reporting the two most important network measures, such as fulfillment cost-to-serve or retail media contribution. By emphasizing these measures as one basis for investment decisions, a store can earn its place in part through the fulfillment or media value it adds, for example, even if its standalone margins look thin.
- Test store closures against their potential impact on the network. Before you close or relocate a store, consider what the network would lose if it went away—delivery coverage in the area, for example, or the inventory buffer it provides to nearby locations. A store that supports other network locations nearby may be worth keeping open even though its own P&L looks weak.
- Analyze catchments for new opportunities. Opening new stores isn’t the only way to meet increasing demand. Before deciding that a new location is warranted, map a single catchment across everything it serves: stores, delivery nodes, pickup points, and partner locations. Look for coverage gaps and duplicated effort that a store-level P&L might hide. Act on what you find by filling gaps with the cheapest node available and redeploying any overlaps.
- Scale only capabilities that hit the target return. In a portfolio approach, new capabilities are worth adopting only if they create value for the entire network, not just for a single location. When deciding what to scale, pick a single capability, such as a new labor model or in-store technology, and establish a target return that it must reach before you will introduce it at additional locations. Track how the rollout performs against the return, and continue only if it is clearing the bar.
As companies embed AI in more store operations, retailers will prosper by taking a deliberate approach to setting up store networks and by moving from traditional measures to more holistic metrics that take stock of the value that individual locations create.
The authors wish to thank Lia Gregor for her contributions to this article.
