Virtual Round Table · Jul 22

View the event
Retail's broadcast era is ending, and zero percent of retailers have finished the conversational AI switch
AI & ML

Retail's broadcast era is ending, and zero percent of retailers have finished the conversational AI switch

New research from Infobip and Retail Economics finds 88 percent of retailers are exploring conversational AI while none report being fully integrated, exposing a wide gap between intent and execution.

PublishedJuly 21, 2026
Read time6 min read
Share

The finding that should worry marketers

Research released July 20 by cloud communications firm Infobip and consultancy Retail Economics lands on a blunt conclusion: the broadcast era of retail engagement, built on one-way notifications blasted to entire lists, is giving way to a conversational, AI-driven relationship layer. The report argues that treating a customer's phone as a loudspeaker for promotions is losing effectiveness as shoppers come to expect two-way exchanges that answer a question, track an order, or resolve a return in the same thread. That shift reframes messaging from a campaign channel owned by marketing into an always-on service surface that cuts across marketing, customer service, and logistics.

Richard Lim, chief executive of Retail Economics, put the demand-side pressure plainly. A one-size-fits-all approach to communication no longer works, he said, with preferences shifting based on age, income, and shopper intention. That variability is the operational problem retailers now face. A single broadcast cadence cannot serve a Gen Z shopper who wants a delivery update over WhatsApp and a Baby Boomer who still reads email, and the report's data suggests the cost of ignoring that split is measured in open rates, conversion, and eroding attention from the customers a retailer most wants to keep.

The channels are fragmenting by generation

The numbers behind the argument are stark. Infobip reports WhatsApp open rates between 85 and 95 percent, against 32.7 percent for email in e-commerce, a gap wide enough to change the economics of any post-purchase communication. Where a retailer reaches a third of an email list, it reaches nearly all of a WhatsApp audience, and that difference compounds across order confirmations, shipping updates, and returns prompts that carry real revenue consequences when they go unread. The channel a retailer picks now determines whether a time-sensitive message actually arrives in front of the customer or dies in an unopened inbox.

The generational split is equally clear. The report finds 24 percent of Gen Z now prefer WhatsApp and SMS for deliveries and returns, while Baby Boomers are twice as likely to want email. That divergence means a retailer cannot standardize on one channel and serve its full base well. It has to route each message to the medium a given customer actually reads, which is an orchestration problem rather than a content problem. Building that routing is where conversational platforms earn their keep, because deciding channel by customer and by message type at scale is precisely the work a rules-and-AI layer is meant to automate.

The readiness gap is enormous

The most quotable figure in the research is the distance between ambition and delivery. Industry polling cited in the report finds 88 percent of retailers have begun exploring conversational AI, yet zero percent report being fully integrated across the entire customer journey. That combination describes a market crowded with pilots and empty of finished deployments. Nearly everyone has a chatbot on a product page or a WhatsApp number for order updates, and almost no one has stitched conversation through discovery, purchase, fulfillment, and post-purchase service as a single connected experience the customer perceives as coherent.

For enterprise buyers, that gap is the opportunity and the warning at once. It is an opportunity because full integration is still open territory, so a retailer that connects the journey end to end can differentiate before rivals catch up. It is a warning because zero percent completion after this much exploration signals that integration is genuinely hard, blocked by fragmented systems, siloed customer data, and channels owned by different teams. The retailers that close the gap will be the ones that treat conversation as an operating capability spanning departments, not a feature that one team bolts onto an existing campaign tool and declares done.

The cost objection, answered

The reflexive objection to conversational retail is that answering messages at scale means hiring an army of agents. Infobip's retail specialist Kim Johal confronted that head on, noting the biggest objection the firm hears is that retailers cannot afford to answer conversational messages at scale, and adding that the industry is no longer building call centres. The point is that AI now handles the high-volume, low-complexity questions that once required human staffing. A where-is-my-order query, the single most common inbound message in retail, can be resolved by an automated agent pulling live order status, freeing human staff for the smaller share of interactions that genuinely need judgment.

The report frames the payoff in time-to-value terms, claiming that using AI to handle routine queries can deliver a 24 hour, seven day a week sales-associate experience that proves its return on investment in as little as 60 days. That figure deserves the usual scrutiny that vendor economics warrant, and buyers should insist on seeing the baseline it measures against. Still, the direction is credible. Deflecting routine contacts away from human queues while improving response times is a well-understood source of savings, and a 60 day payback window, if it survives a real pilot, is short enough to clear most procurement thresholds without a drawn-out business case.

What commerce leaders should take from it

The practical takeaway is to stop scoring messaging by campaign metrics and start scoring it by resolution. A conversational layer succeeds when it answers the customer's actual question in the channel they chose, and that standard requires connecting order, inventory, and customer data to the messaging surface rather than running a chatbot in isolation. The retailers stuck in the exploration phase are usually the ones that deployed a conversational tool without wiring it into the systems of record, which produces a bot that can chat but cannot tell a shopper where their parcel is, the one thing they messaged to find out.

The strategic risk is drift. Conversational engagement is quietly becoming the expected baseline for how a shopper interacts with a brand, and a retailer that treats it as a future project cedes attention to competitors who answer instantly on the channel the customer prefers. This research announces no single product or deal, yet it quantifies a shift enterprise leaders can act on now. The move is to pick the two or three journeys where conversation clearly beats broadcast, deliveries, returns, and order status, and to finish those end to end before attempting the whole journey at once.

Tagged#news#retail#retail-ai#ecommerce#agentic-commerce#cpg#infobip#retail-economics#conversational-ai#conversational-commerce#whatsapp#customer-engagement#richard-lim