According to the Colliers Spring 2026 Retail Research Report, 85.1 percent of U.S. retail sales still flow through physical stores. Furthermore, 71 percent of retailers are expanding their physical footprints this year. The retail apocalypse narrative that defined a decade of industry conversation is over. The data is unambiguous.
But the same Colliers report contains a number that most expansion coverage is not leading with. Early adopters of in-store AI technology report 79 percent higher store sales growth compared to those who have not. Retail leaders who increased IT spending by 52 percent over the past five years will see profits grow nearly three times faster than competitors who increased IT spending by only 13 percent over the same period.
The retail apocalypse was wrong. But the recovery is not evenly distributed. Expanding square footage without rebuilding the technology layer inside it is not a store renaissance. It is the same store problem at a larger scale.
The Numbers Behind the Recovery
The Colliers Spring 2026 data set deserves a careful read before the narrative gets oversimplified. Colliers projects consumer adoption of AI shopping agents will jump from 19 percent to 46 percent by end of 2026. Physical stores now fulfill one in four online orders, a figure Colliers projects will reach 35.4 percent by 2030. Coresight Research forecasts more than 5,000 store openings in 2026, characterized by Retail Brew as strategic refinement rather than rapid expansion.
These are not vanity metrics. They show that the store now does a fundamentally different job. It is no longer just a place where the shopper arrives. It is the node through which a growing share of digital commerce gets fulfilled, returned, and serviced. The physical store is now infrastructure, not just retail.
The problem is that most retailers joining the expansion wave are adding square footage to a store model built for the old job, not the new one. The store that handles walk-in shoppers requires a different operating model than the store that simultaneously fulfills one in four online orders and manages AI-driven customer interactions. Most expanding retailers have the square footage. They do not yet have what goes inside it.
Why Expansion Without Infrastructure Creates a Different Problem
In several deployments across the United States and Latin America, we have watched retailers announce store expansion programs, execute them on schedule, and then spend the next 18 months trying to understand why the new locations underperform the projections. The stores are open. The traffic is there. The unit economics are not.
The pattern is consistent. Three structural gaps explain most of the shortfall.
Gap 1: The Store Format Changed, But the Operations Did Not
The modern store in 2026 now serves as a showroom, a fulfillment center, a returns desk, an AI-assisted service environment, and a brand experience simultaneously. However, most retailers staffing new locations are using playbooks from the old store model. The associate hired for the new location knows how to run a register. That same associate now needs to resolve BOPIS exceptions, manage real-time inventory alerts from the connected floor system, and handle returns from shoppers who completed purchases through a Google AI agent. Those are different competencies. No traditional store opening playbook covers all of them.
Gap 2: AI Adoption Inside the Store Is Where the 79 Percent Lives
The 79 percent higher store sales growth that Colliers attributes to early AI adopters does not come from foot traffic. It comes from what those retailers do with each customer once they arrive. As I described in the context of the AI for retail associates adoption gap, the tool is not the advantage. The operating model that retailers build around the tool is the advantage. Retailers adding AI to existing stores without changing the associate role, the KPIs, or the data architecture will not capture the 79 percent. They will capture whatever improvement comes from having a chatbot answer FAQs.
Gap 3: Fulfillment Infrastructure Is Now a Prerequisite, Not an Add-On
Physical stores now fulfill one in four online orders. By 2030, Colliers projects that ratio rises to one in three. Reaching that ratio requires stores with real-time inventory accuracy above 98 percent, dedicated pick paths, exception handling workflows, and a data layer connecting the store floor to the digital order management system without batch syncs. As I described in the context of the Amazon Now store format analysis, retailers who treat fulfillment as a back-office function will lose the next decade to retailers who treat it as a core store capability. That principle applies to every retailer opening locations in 2026.
The Adoption Layer: What Has to Change When the Store Becomes a Technology Platform
This is the part real estate reports do not cover. The Colliers data documents the expansion. What it does not document is what happens inside the buildings after the ribbon is cut.
The Operating Model Has to Shift
A store serving walk-in shoppers while fulfilling digital orders runs on a different shift structure and a different escalation path than the traditional store model. In the old world, the morning supervisor managed the floor. Today, that supervisor manages the floor, the pick queue, the BOPIS hold area, and the real-time exception alerts from the inventory system at the same time. Retailers who open new locations with the old supervisor playbook will discover the new job on the first Black Friday, not in the planning phase.
The KPIs Have to Change
The old store KPIs were sales per square foot, conversion rate, and average transaction value. Those measure the store as a transaction venue. They do not measure the store as a fulfillment and AI-assisted service platform. The KPIs that matter for the 2026 store are different: digital order fulfillment accuracy rate, BOPIS exception rate, real-time SKU-level inventory accuracy, and the share of customer interactions the AI layer resolves before the associate gets involved. Most retailers opening new locations in 2026 still measure the old KPIs. The gap between what they measure and what actually drives profitability is where the expansion economics break down.
The False Success Mode
The most common failure pattern in 2026 store expansion is a location that opens on time, hits its first-year traffic targets, passes internal audit, and still underperforms the financial model by year two. The traffic is real. The technology is live. But nobody redesigned the operating model for what the technology requires. Opening is not operating. Traffic is not revenue. Square footage is not infrastructure.
Three Architecture Decisions Every Expanding Retailer Has to Make
The Store as a Fulfillment Node
Retailers should design every new location as a fulfillment-capable node from day one. That means a dedicated BOPIS hold area, a pick path that does not conflict with the shopper floor, inventory accuracy infrastructure (electronic shelf labels, RFID, or cycle count automation), and a real-time connection to the order management system. As I described in the context of the Walmart Mexico ESL rollout and the inventory robot deployment analysis, the connected floor layer is the foundation everything else runs on. Retrofitting after opening costs more than building it in from the start.
The AI Layer Built Into the Store Design
Retailers should specify AI-assisted associate tools, self-service kiosk infrastructure, and the data feeds connecting in-store behavior to the broader customer data platform during the store design phase, not after opening. The 79 percent growth advantage that Colliers documents for early AI adopters did not come from stores that bolted AI onto an existing model. It came from stores where retailers designed the AI layer in. The difference in unit economics between those two approaches compounds every year the store stays open.
The Data Architecture That Connects the Store to Everything Else
A store that fulfills digital orders, serves AI-assisted shoppers, and feeds performance data back to the merchandising team runs on a real-time data architecture. Batch syncs that update overnight do not support a store that must fulfill a same-day order placed at 6 PM. The data architecture decision made during store design determines the store’s operational ceiling for the next decade. Retailers who choose the cheap batch-sync approach during the 2026 expansion wave will face an expensive architectural migration at exactly the moment competitive pressure peaks.
What This Means for LatAm Retailers
The Colliers report focuses on the U.S., but the same pattern plays out across Latin America with a two-year lag. From the deployment side, I have walked retail expansion projects in Mexico, Colombia, Chile, and the Dominican Republic where teams executed location selection, lease negotiation, and store design at a high level but leaders treated the technology architecture as a procurement afterthought. The stores opened. The operating models did not scale. The economics never hit the underwriting assumptions.
For LatAm grocers, department stores, specialty chains, and drugstore operators planning expansion in 2026 and 2027, the strategic question is not “how many locations?” Rather it is “are the locations we are opening capable of operating as AI-enabled fulfillment nodes, and if not, what is the three-year cost of retrofitting them compared to building them right the first time?” The CFO and the CIO should answer that question together before signing the first lease.
The Question Nobody Is Asking in the Expansion Planning Meeting
Every retailer announcing a store expansion program in 2026 has a real estate committee, a construction timeline, a lease negotiation team, and a grand opening playbook. What most lack is an honest answer to this question.
When this location opens, what percentage of its revenue will come from digital order fulfillment? Does the store design physically and technologically support that volume without degrading the walk-in shopper experience at the same time?
Retailers who answer that question honestly in the planning phase build different stores. The ones who defer it discover the answer in the first year of operations, at which point the cost of changing it is already built into the lease.
85 percent of U.S. retail sales still flow through physical stores. But not through all of them equally. The stores winning in 2026 were designed for the job retail is doing now, not the job it was doing five years ago.
If you are planning a store expansion or a technology architecture refresh for your retail network, connect with me here or reach me on LinkedIn. I am happy to walk through the deployment framework we use across the U.S. and Latin America.
Adriana Rivas is a retail technology executive and AI strategist, and the founder of a U.S.-based hardware company specializing in self-service kiosks, POS systems, electronic shelf labels, and digital signage deployed across the United States and Latin America. She is the award-winning author of How to Implement Self-Service Without Failing (Amazon #1 Hot New Release, Silver Nonfiction Book Award 2025) and recipient of the Gold Stevie® Award, Thought Leader of the Year 2026. She is also recognized by Thinkers360 as a Top 10 Thought Leader in Retail and a Certified Master Expert in Retail.