According to the Colliers Spring 2026 Retail Research Report, Buy Online Pick Up In Store now accounts for 8.9 percent of total retail revenue in the United States. That figure will rise to an estimated 12 percent by 2028, driven by the same fulfillment shift that is pushing one in four online orders through physical store locations today. The number is significant. What sits next to it in the same report is more significant. Only 18 percent of retailers report having fully optimized the BOPIS process.
That gap, between 8.9 percent of revenue and 18 percent optimization, is one of the most consequential mismatches in retail operations right now. BOPIS is not a feature in the testing phase. It is a live revenue channel generating real dollars in most large retail organizations today. And most of those organizations are running it on infrastructure that was never designed to support it at scale.
BOPIS is not a button. It is an operating model. And most retailers have built the button without building what the button requires to work.
What 8.9 Percent of Revenue Actually Means Operationally
For a retailer generating one billion dollars in annual revenue, 8.9 percent is 89 million dollars flowing through BOPIS. That is not a pilot. That is a business unit. And it is a business unit that, in most retailers, does not have a dedicated P&L, a dedicated operational playbook, a dedicated workforce model, or a dedicated technology stack.
The Colliers data also shows that BOPIS drives 34 percent stronger performance among early AI adopters compared to the broader retail market. That 34 percent advantage is not coming from a better mobile app checkout flow. It is coming from retailers who built the operational infrastructure that makes BOPIS work at the pick, hold, and handoff layer: real-time inventory accuracy, a hold area that does not conflict with the shopper floor, a notification system that reaches the associate before the customer arrives, and an exception handling workflow that catches the out-of-stocks before the shopper does.
The retailers in the 18 percent who have fully optimized BOPIS built all of those things. The 82 percent who have not are running 8.9 percent of their revenue through a process that breaks regularly and gets patched manually every time.
Why Most BOPIS Programs Break at the Same Three Points
In several deployments across the United States and Latin America, we have walked BOPIS operations ranging from grocery chains to electronics retailers to department stores. The failure modes are consistent. Three structural gaps account for the majority of BOPIS exceptions, customer complaints, and associate workarounds.
Gap 1: Inventory Inaccuracy at the Moment of Pick
A BOPIS order is accepted by the system when inventory shows available. The associate goes to pick the item and finds the shelf empty. The item was sold from the floor between the time the order was placed and the time the pick was attempted. In a system with batch inventory updates, that gap can be hours. For a retailer with high-velocity SKUs, that gap produces a predictable failure rate that compounds with order volume.
Real-time inventory accuracy, driven by electronic shelf labels, RFID, or cycle count automation, is the only reliable solution. As I described in the context of the Walmart Mexico ESL rollout and the GS1 RFID TDS 2.3 web-resolvable standard, the connected floor layer is the infrastructure prerequisite for any fulfillment operation that cannot tolerate overnight batch sync windows. BOPIS is exactly that operation.
Gap 2: The Hold Area That Conflicts With the Shopper Floor
In most stores that launched BOPIS without redesigning the floor plan, the designated pickup area is a repurposed customer service desk, a section of the returns counter, or a shelf near the front register. These solutions work at low BOPIS volume. At 8.9 percent of revenue, they do not. The hold area becomes a source of associate congestion, customer confusion, and lost orders. The shopper who arrives to pick up an order should be able to identify the pickup zone, retrieve their order, and exit in under three minutes. In most stores running BOPIS through a repurposed service desk, that experience runs five to eight minutes on a busy day and involves at least two associate interventions.
Gap 3: Notification Timing That Gets the Sequence Wrong
The BOPIS notification sequence should be: order received, pick confirmed, order ready for pickup. What most retailers are running in practice is: order received, automated ready notification fired at pick start (not pick completion), customer arrives before pick is complete, associate is still pulling the order, customer waits. That failure is a software configuration issue, not a hardware issue. But it produces the same customer experience outcome as a pick failure and it is the most common single source of BOPIS complaints in retail operations research.
The Adoption Layer: What Has to Change to Hit the 34 Percent
The 34 percent BOPIS performance advantage that Colliers documents among AI-enabled retailers did not come from a software upgrade. It came from retailers who rebuilt the operational model around BOPIS as a primary revenue channel, not as an extension of the e-commerce checkout flow.
The Operating Model Has to Shift
BOPIS at 8.9 percent of revenue is a warehouse operation running inside a retail store. It requires a shift structure that accounts for peak pick times (which correlate with digital order volume, not foot traffic), a dedicated associate role or rotation for pick and hold, and an escalation path for substitution decisions that does not require a manager intervention for every out-of-stock. Retailers running BOPIS as a task that any available associate handles between customer interactions will find that their exception rate rises linearly with volume. As I described in the context of the AI for retail associates adoption gap, the operating model has to be designed for the new job before the technology can deliver its value.
The KPIs Have to Change
Most retailers measure BOPIS through a single metric: order fulfillment rate. That metric captures whether the order shipped. It does not capture pick time, hold time, customer wait time, substitution rate, or the rate at which the notification sequence fires correctly. A retailer with a 94 percent BOPIS fulfillment rate can still have a broken BOPIS operation if the 94 percent that succeeds takes 45 minutes to pick and holds the customer at the counter for 6 minutes. The full KPI set for a mature BOPIS program is: pick time from order confirmation, hold duration before customer notification, customer dwell time at pickup, substitution rate, and post-pickup NPS compared to in-store purchase NPS. Most retailers are measuring one of those five.
The False Success Mode
The most common BOPIS failure pattern in 2026 is a program that passes internal review because the headline fulfillment rate is above 90 percent, while the operational reality behind that number includes daily manual workarounds, associate overtime on high-volume days, and a customer experience that is measurably worse than in-store purchase. Fulfillment rate is not BOPIS health. Completed orders are not satisfied customers. A passing metric is not an optimized operation.
Three Infrastructure Decisions That Separate the 18 Percent From the 82 Percent
Real-Time Inventory Accuracy as a Non-Negotiable
BOPIS at scale requires inventory accuracy above 98 percent in real time, not at end of day. That threshold is achievable with electronic shelf label infrastructure, passive RFID at the item level for applicable categories, or high-frequency automated cycle counts supported by computer vision. Retailers who accept batch inventory updates as the foundation for their BOPIS program are accepting a structural exception rate that will grow proportionally with BOPIS volume. There is no software fix for a hardware and process gap at the inventory layer.
Dedicated BOPIS Infrastructure in the Store Design
A mature BOPIS operation requires a dedicated hold area sized for peak day volume, a designated pick path that does not route associates through the highest-traffic shopper areas, and a notification system that fires on pick completion, not pick initiation. For retailers currently running BOPIS through repurposed customer service desks, the infrastructure question is whether to retrofit existing stores or build the right design into net-new locations. As I described in the context of the store expansion infrastructure analysis, building it in from the start is materially cheaper than retrofitting after opening.
AI-Assisted Exception Handling at the Pick Layer
The substitution decision is the highest-friction moment in BOPIS operations. When the primary item is out of stock, the associate has to identify an acceptable substitute, confirm availability, update the order, and notify the customer, all before the customer arrives. AI-assisted exception handling, integrated into the handheld device the associate is already carrying, can surface substitution options ranked by price, availability, and customer purchase history in real time. As I described in the context of the conversational POS reframe, the AI layer surfaces inside the existing workflow or it does not get used. BOPIS exception handling is exactly the use case that rewards getting that integration right.
What This Means for LatAm Retailers
BOPIS adoption in Latin America lags the U.S. by two to three years, which is an advantage for retailers who use that window to build correctly rather than retrofit hastily. The most common pattern I have seen in LatAm deployments is retailers launching BOPIS as a digital feature, driving adoption through app promotions, and then discovering the operational gap when volume reaches a level that the repurposed service desk cannot handle.
For LatAm grocers, pharmacies, department stores, and electronics retailers, the strategic question is not “should we offer BOPIS?” Rather it is “do we have the inventory accuracy, the physical infrastructure, and the operational model to run BOPIS as a primary revenue channel rather than as a convenience feature, and if not, what is the cost of the gap?” At 8.9 percent of revenue and rising, that is a question with a dollar figure attached. Most LatAm retail planning processes have not calculated it yet.
The Number That Matters for Your Operation
The industry average is 8.9 percent of retail revenue flowing through BOPIS. Only 18 percent of retailers have fully optimized the process. That is not a benchmark to compare against. It is a baseline to start from.
The number that actually matters is yours. What percentage of your revenue flows through BOPIS today? What is your pick time, your hold time, your exception rate, and your post-pickup NPS? And if your current BOPIS infrastructure were asked to handle 12 percent of revenue, which is where Colliers projects the category will be by 2028, what would break first?
The retailers who answer those questions with real operational data in 2026 will build the infrastructure changes into their 2027 capital plans. The ones who do not will discover the answers during their next peak season, when the cost of fixing them is highest and the window to do so is closed.
BOPIS is already 8.9 percent of retail revenue. Most retailers are running it at 18 percent optimization. The gap between those two numbers is not a technology problem. It is a decision problem. And the decision window is open right now.
If you are evaluating a BOPIS infrastructure upgrade or a connected store deployment 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.