AI

AI is accelerating commerce, but service-led AI is where brands win

If agentic commerce scales into a trillion-dollar opportunity, success will hinge on infrastructure.

Adobe Stock

April 17, 2026 by Matt Price — CEO, Crescendo

Summarize with

Retail loves a breakout forecast — from livestream commerce to QR codes, to one-click hardware. The current candidate: AI and agentic commerce. But unlike past trends, this one is reshaping not just how people shop, but how brands engage customers once they arrive.

There is promise, and the numbers are hard to ignore. AI-referred shopping traffic surged 693% year over year in November and December 2025, with January 2026 still up 527%. Consumers increasingly use AI assistants to research products, ask questions, and narrow options before making a purchase. These interactions are shaping shopping behavior in meaningful ways.

Yet AI has not yet surpassed social commerce, where referral traffic still grew roughly 20% year over year. Instead, it is unlocking something brands have historically struggled to scale: turning service interactions into revenue moments.

The most immediate shift in retail AI is happening inside the shopping experience itself.

Consumers want AI, on their terms

Most predictions on AI-driven commerce hinge on the idea that conversational assistants will sit between brands and buyers, orchestrating discovery and conducting checkout.

Evidence supports this: 49% of consumers say they would delegate both routine and research-heavy purchases to an agentic assistant. But among consumers who primarily use search engine AI summaries, 59% say they prefer an AI assistant on the merchant's website.

Consumers aren't rejecting AI. They are deciding where it belongs. Increasingly, they prefer it embedded in the retail environments they already trust.

Search is changing, so is skepticism

Forty-four percent of users who have tried AI-powered search say it has become their primary and preferred way to search the internet, compared to just 31% who still prefer traditional search.

McKinsey projects that by 2030, U.S. B2C retail alone could see up to $1 trillion in orchestrated revenue from agentic commerce. But growth projections don't eliminate consumer scrutiny around where AI fits in the purchase journey. Discovery may evolve, but hesitation still happens closer to the point of purchase.

Customers still have practical questions before buying: shipping timing, bundling options, personalization, sizing, or returns. These are the moments that determine whether a shopper proceeds to checkout or abandons the purchase.

AI-enabled service solves this hesitation layer. Even if discovery begins elsewhere, those moments of uncertainty are most effectively resolved within the environments consumers already trust.

Engagement is up, but control isn't shifting

AI is improving how shoppers navigate information and narrow choices. But narrowing the funnel is not the same as owning the customer relationship. Consumers may be willing to delegate decisions to AI agents, yet preference for merchant-site assistants indicates that delegation works best when anchored to a known brand ecosystem — complete with clear return policies, loyalty programs, and accessible human support.

This is where the next competitive divide will emerge: not over who builds the smartest AI assistant, but over who builds the strongest customer context that informs it. Retailers that embed AI directly inside their digital experience gain something external assistants cannot replicate: contextual understanding of the customer relationship.

The economics of service-led growth

Brands have historically been cautious when proactively inviting engagement. Every additional conversation required human labor, making proactive engagement more costly. Service was necessary, but economically constrained and rarely treated as a scalable growth channel.

AI changes that equation. When routine questions and pre-purchase inquiries can be handled autonomously, the marginal cost of conversation drops dramatically. What was once defensive support becomes a proactive, conversion layer across the entire shopping experience.

In early deployments, nearly half of AI-driven inquiries are tied directly to purchase intent: sizing, shipping timing, product comparisons, or availability. The other half are traditional support interactions that can be resolved instantly without incremental labor costs.

This shift introduces what many retailers now describe as service-led sales. A customer begins with what appears to be a support interaction — a shipping question, sizing clarification, return policy inquiry, or product comparison. The assistant removes friction, answers the question, and recommends relevant products or bundles that ultimately lead to purchase.

In early deployments, AI-assisted chat interactions have converted to orders at rates between 23% and 27%, compared to roughly only 12% for baseline web search. Customers who engage through conversational assistance also tend to spend more, with average order values roughly 25% higher than typical sessions.

When retailers move from reactive support pages to proactive, site-wide assistance, engagement can increase by more than 60%.

In other words, AI doesn't just narrow the funnel. It changes the economics of engagement. Instead of limiting conversations to control cost, brands can surface assistance across the entire shopping experience, capturing intent that previously never translated into interaction.

Rather than shifting trust away from brands, AI allows brands to operationalize trust at scale, turning hesitation into revenue.

Persistent context and clienteling

If agentic commerce scales into a trillion-dollar opportunity, success will hinge on infrastructure.

Loyalty in an AI environment requires hyperpersonalized experiences — building persistent customer context layers accessible by agents and exposing loyalty services and eligibility engines via APIs.

An AI agent that surfaces the right product but doesn't know that the customer has elite status, a pending return, or a size preference isn't clienteling; it's a slightly smarter search function. That preference for a merchant site isn't just for the UX, but rather something that other platforms can't replicate: a context layer embedded in the brand relationship.

Consumers may delegate tasks to AI. But the brands that win will be those that embed AI directly inside their service layer, where context, loyalty, and transaction history already live. The retailers that understand this won't just ride the growth curve, they'll own the relationship long after the transaction is automated.

About Matt Price

Matt Price is the Founder & CEO of Crescendo, the world’s first AI-native contact center. He previously spent more than a decade at Zendesk Labs, where he founded the company’s European operation and led global marketing of products and strategic projects as SVP and Executive Leader. Earlier in his career, he held leadership roles at several high-growth software companies, three of which went public and two that were acquired.

Connect with Matt: