Worldwide spending on conversational commerce channels is projected to reach approximately 290 billion dollars in 2026. The retail and commerce sector now leads all industries in conversational AI adoption, holding a 21.2 percent market share. Most retail technology coverage treats this as an e-commerce story. It is not. It is a redefinition of what a Point of Sale system is for.
In several deployments across the United States and Latin America, we have watched retailers spend years optimizing the POS as a transaction terminal. Faster scanning. Better receipt printing. Tighter EMV integration. Meanwhile, the customer interface around the POS has been quietly migrating to chat windows, voice assistants, and AI agents that handle the entire pre-transaction conversation. The terminal has not changed. The job around it has.
The POS is no longer the system the customer interacts with. It is the system that confirms what the AI agent already decided. The retailers who understand that shift will rebuild their store stack. The ones who do not will keep buying faster scanners for a job that has moved upstream.
The Shift Underneath the 290 Billion Dollar Number
The headline figure is striking, but the structural change underneath it is more important. According to aggregated industry research published in 2026, retail and commerce now lead all industries in conversational AI adoption. Furthermore, the number of voice assistant users in the United States is expected to reach 157.1 million by the end of 2026.
For context, that is roughly half the U.S. population using voice as a shopping interface. Specifically, 49.6 percent of U.S. consumers now use voice search for shopping. The U.S. voice commerce market alone is estimated at 22.4 billion dollars in 2026. That is no longer a niche category.
What this means operationally is that a meaningful share of pre-purchase intent now forms in a conversational interface, not in a search bar and not in a store aisle. By the time the customer reaches the register, the decision has already been made. The product has been compared, the price has been validated, sometimes the order has been placed. As a result, the POS is being asked to do a different job. And most POS architectures are still designed for the old one.
Why Most POS Systems Are Failing the Conversational Test
Walk into a retailer running a modern POS deployment in 2026 and you will see a terminal that does its core job well. It scans. It accepts EMV. It prints. It integrates with inventory at the SKU level. However, ask the terminal to handle a customer who arrived via a conversational interface, and most deployments break down.
From the deployment side, three structural gaps consistently keep POS architectures out of the conversational commerce stack.
Gap 1: No API Surface for AI Agents
If an AI agent cannot query the POS in real time for price, inventory, and promotion data, it cannot complete the transaction it started. Most legacy POS systems were not designed to be consumed by an external agent. Therefore, the entire conversational stack ends at the moment of payment because the POS cannot participate in the conversation.
Gap 2: No Continuity Across Channels
A customer who started a conversation with an AI agent on the brand’s app, continued it in a WhatsApp thread, and finished at a physical kiosk should walk up to the register with the order ready. Specifically, the cart, the customer ID, the loyalty status, and the AI-driven recommendations should all be visible to the POS. In most retailers, none of those are. The terminal sees the customer as a stranger every time.
Gap 3: No Voice or Chat Interface at the Terminal Itself
The customer at the POS in 2026 increasingly expects to be able to ask a question by voice or text, not navigate a touchscreen menu. However, most POS terminals are still pure tap-and-scan interfaces. The conversational layer stops at the doorway. That is a UX gap, but more importantly, it is a data gap because every voice or chat interaction is also a data point about intent that the retailer is not capturing.
The Adoption Layer: What Changes When the POS Becomes Conversational
This is the part vendors will not put on a slide. Adding conversational capability to a POS is not a software upgrade. It is an operating-model change that touches store labor, customer-data governance, and the relationship between hardware vendors and AI platform providers.
The Operating Model Has to Shift
In a pre-conversational world, the cashier or self-service kiosk was the customer’s first and last touchpoint with the retailer’s decision systems. In a conversational world, the customer arrives at the terminal having already interacted with the brand multiple times. By contrast, the cashier becomes a closing role, not a starting role. As a result, training shifts from product knowledge to context recovery, conflict resolution, and exception handling. The retailers running healthy conversational programs are the ones who redesigned the associate role first, the terminal second.
The KPIs Have to Change
The old KPIs around POS were transaction speed, scan accuracy, and uptime. Those measure that the terminal is working. However, they do not measure whether the conversational layer worked. By contrast, the KPIs that matter once the POS becomes conversational are different: conversation completion rate before terminal handoff, AI-resolved exceptions at the terminal, conversational basket size versus traditional basket size, and the rate at which the customer’s pre-arrival context successfully attaches to the transaction. Most importantly, retailers that fail to install the new measurement framework will run a conversational pilot that never proves out.
The False Success Mode
The most common failure pattern in 2026 is bolting a chatbot onto the existing POS and declaring the upgrade complete. Specifically, the chatbot answers FAQs. The POS still processes the transaction the same way. None of the AI-driven intent data flows into the terminal, and none of the terminal-side transaction data flows back into the AI layer. Six months later, leadership cannot explain why the conversational investment never moved basket size. A chatbot is not a stack. A voice interface is not an architecture. Bolting on is not building in.
The Hardware and Architecture Decisions That Decide the Outcome
Three architectural decisions separate the retailers who will earn real value from conversational commerce from those who will keep adding chatbots that do not move the P&L.
POS as an API, Not a Terminal
The POS of 2022 was a closed system that exposed limited integration points to specific certified partners. By contrast, the POS of 2026 needs to be an API-first system that any AI agent, conversational platform, or third-party shopping assistant can query and transact against. Cloud-native POS platforms have a structural advantage here. Legacy on-premises systems require significant middleware to participate in the conversational stack at all.
Customer Context Continuity
A customer’s pre-arrival conversation must persist into the in-store moment. That requires a customer-data platform that the POS can consume in real time, not a CRM batch sync that updates overnight. Furthermore, the customer-data architecture is now upstream of the POS, not downstream. Retailers who move on this in 2026 will have a different basket size and a different attribution model than retailers who treat customer data as a marketing concern.
Conversational Interface at the Terminal
Adding voice or chat to the terminal itself is the visible part of the change, but it is the smallest part. The retailers running serious conversational programs are adding microphones and natural language interfaces at the kiosk, the staffed register, and the mobile POS. Meanwhile, they are training the conversational layer on the same intent data the upstream AI agents use. This is the same operating-system shift I described in the context of the Walmart Mexico ESL rollout and the self-service kiosk ROI framework. The technology surfaces look different. The pattern underneath is the same: connected, real-time, and context-aware.
What This Means for LatAm Retailers
Latin America has a structural advantage in conversational commerce that most CIOs in the region underestimate. The conversational interface most LatAm consumers already use daily is WhatsApp, with penetration above 80 percent in most major markets. That is a conversational layer the U.S. retail industry is still building from scratch.
From the deployment side, I have walked stores in Mexico, Colombia, and Chile where the customer’s entire journey already runs through WhatsApp: product inquiry, price negotiation, order placement, delivery scheduling, and post-purchase service. However, the POS at the physical pickup point treats that customer as a walk-in. As a result, the conversational data evaporates at the moment of payment. The retailer collected nothing from the most valuable part of the journey.
For LatAm grocers, drugstore chains, department stores, and specialty retailers, the strategic question is not “should we add a chatbot?” Rather, it is “is our POS architecture capable of consuming the conversational context our customers are already creating, and if not, what is the 18-month plan to make it so?” Importantly, that question goes to the CIO and the CFO at the same time. It is not a marketing decision.
Where to Start: The 18 Month Playbook
The sequencing playbook for retail leaders evaluating a conversational POS strategy over the next 18 months is concrete.
Next 30 Days: Audit the API Surface
First, document every integration point your current POS exposes. Identify which AI agents, conversational platforms, and customer-data systems can query the POS in real time. The size of that gap is the size of your conversational debt. Retailers running on cloud-native POS platforms typically find smaller gaps than retailers running legacy on-premises stacks.
Two Quarters Out: Pilot Context Continuity
Next, pilot one channel where the customer’s pre-arrival conversation persists into the in-store moment. Specifically, WhatsApp for LatAm retailers, brand-app chat for U.S. retailers. Then measure the basket size of customers with conversational context attached versus those without. Most importantly, stop measuring chatbot usage alone. Start measuring the lift across the journey.
The 18 Month Horizon: Rebuild the POS as a Node
Finally, rebuild the role of the POS in your overall architecture. The terminal is no longer the system of record for the customer interaction. Rather, it is one node in a conversational mesh that includes mobile, voice, chat, kiosk, and the physical register. As a result, the budget envelope, the talent profile, and the vendor relationships all shift. The retailers who think about this in 2026 will operate a different store stack in 2028.
In conclusion, the retailers who treat the POS as a transaction terminal will keep buying faster scanners for a job that has migrated upstream. By contrast, the retailers who treat it as a conversational node will capture the full value of the 290 billion dollar shift now underway.
The POS used to be where the conversation ended. In the conversational era, it is where the conversation gets confirmed. The retailers who build for that will compound advantage every year.
If you are evaluating a POS modernization or conversational commerce strategy for your store 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 – Retail and a Certified Expert – Retail.