Imagine you need a new pair of running shoes. Simple enough. You open a tab, type “running shoes,” and within seconds you’re staring at hundreds of results. You filter by price. Still too many. You open a few product pages, skim reviews, compare specs, start a mental shortlist, abandon it, open three more tabs. One reviewer says a shoe is perfect for long distances. Another says it falls apart after three months. A third insists it runs half a size too small.
Twenty minutes later, you’re no closer to a decision than when you started. The shoes remain unbought.
This isn’t a story about bad products or poor search results. The right shoes were probably there all along. The problem was something harder to solve with better filters or faster search: there was nobody to help you decide.
There used to be someone whose job was exactly that. They stood in the store, knew the inventory inside out, and when you walked in looking vaguely uncertain, they’d ask a few questions and say: “Given what you’ve told me, I’d go with this one.” You bought it. You left satisfied. The decision felt easy because someone knowledgeable helped narrow the field and build confidence in the choice.
Then ecommerce scaled, but the guidance didn’t. What replaced it was a keyword search box, built on the assumption that shoppers already knew what they wanted and how to describe it. For the better part of two decades, ecommerce optimized relentlessly for retrieval: more products, better filters, faster results, more personalization.
Far less attention went toward helping people decide.
That gap — between finding products and feeling confident enough to choose one — is where an enormous amount of commerce friction now lives. And according to Coveo’s 2026 Commerce Relevance Report, based on a survey of 4,000 shoppers across the U.S. and U.K., shoppers are increasingly feeling it.
The Problem Isn’t Finding. It’s Deciding.
The report’s most important insight isn’t a single statistic but a reframe. Shoppers today aren’t struggling to find products. They’re struggling to decide between them.
Coveo’s research reflects this clearly: 59% of shoppers say they’re more likely to buy when a guided experience helps them find exactly what they’re looking for. The product was often there already. What was missing was confidence in the choice.
This distinction matters because much of modern ecommerce infrastructure is still optimized around surfacing products rather than helping shoppers evaluate them. Better search, more filters, faster recommendations — these improvements increase access to options, but they don’t necessarily make decisions easier.
Shoppers today can access virtually infinite inventory, but abundance alone doesn’t create confidence. The more difficult it becomes to narrow choices, compare tradeoffs, and determine what’s actually relevant, the more likely shoppers are to disengage entirely.
That’s the gap conversational commerce is beginning to address.
Conversational Commerce Is Raising the Bar
Into this gap steps a new kind of expectation — one that shoppers are developing not on retail sites, but everywhere else.
As shoppers encounter AI-powered assistance across search engines, apps, and everyday tools — asking ChatGPT which laptop to buy, getting reviews summarized before they read a single one, comparing feature differences in plain English — their standard for what “helpful” looks like in a shopping experience is quietly rising. Two thirds of shoppers (67%) now expect online experiences to evolve with conversational guidance. For Millennials and Gen Z, that number climbs to 80–81%.
The definition of helpful has shifted. It no longer means returning a list of products from a keyword query. It means experiences that explain options, synthesize information, and help shoppers move forward — without taking the decision out of their hands.
This is where conversational product discovery becomes a meaningful differentiator. Unlike traditional keyword search — which requires shoppers to already know how to describe what they want — conversational product discovery lets shoppers express their needs in natural language, ask follow-up questions, and receive guidance that progressively narrows their choices. It’s the difference between a search bar that returns results and an experience that holds a real conversation. For commerce teams, it reframes the central question from “are we returning relevant products?” to “are we helping shoppers reach a confident decision?”
What shoppers actually want from AI is telling. The most valued use cases are practical: help understanding product attributes before shopping begins (32%), guides tailored to a specific buying situation (32%), and real-time assistance during the journey (28%).
Every one of these use cases points toward the same underlying need: reducing cognitive load. The AI experiences that shoppers respond to most positively aren’t the ones adding novelty to commerce. They’re the ones reducing effort, synthesizing complexity, and making decisions feel easier to navigate.
The Exit You’re Not Tracking
Commerce teams spend enormous amounts of time analyzing cart abandonment. Entire optimization strategies exist to recover shoppers after they’ve added products to a cart but failed to complete a purchase.
But there’s an earlier exit that gets far less attention. It happens before the cart. Before the product page. Sometimes before the second search result. It’s the moment a shopper — overwhelmed, unable to quickly determine which of seventeen similar products is actually right for them — quietly closes the tab and goes somewhere else. Or nowhere at all.
The Coveo report puts a number on it: 71% of shoppers abandon a site when on-site search fails. Sixty-nine percent have left a retailer’s site due to poor discovery at least sometimes. Eighty-two percent of shoppers say they encounter product discovery problems when shopping online — the most common being too many choices with poor filtering (31%), difficulty finding what they want through search or browsing (30%), and irrelevant recommendations (28%).
These aren’t edge cases or occasional frustrations. They’re the majority experience, playing out across sites every day.
The search bar is where confidence is won or lost first. When shoppers arrive on a retailer’s site with a specific goal, 44% go straight to the search bar — more than double the share who start by browsing categories. Search isn’t just a feature. It’s the first test of relevance, the moment where a shopper decides whether this experience is worth their time. Fail it, and everything downstream — personalization, recommendations, checkout optimization — never gets a chance.
Trust Is the Ceiling
Shoppers are open to guidance. They’re warming to AI. But they’re drawing a firm line between AI that helps them decide and AI that decides for them — and that line has significant implications for how aggressively commerce teams should be moving toward automation.
Only 16% of shoppers say they’re very comfortable letting an AI assistant find and purchase products on their behalf. This isn’t a minority opinion — it reflects where the majority of shoppers still stand on autonomous AI. Research from the University of Virginia Darden School of Business adds important nuance here: consumers trust AI more for practical, lower-stakes purchases, but prefer human input when decisions feel more personal or consequential. The more the purchase matters, the more shoppers want to remain in the driver’s seat.
Data sharing tells the same story. Three quarters of shoppers (75%) say they restrict what they share online to only what’s absolutely necessary. Willingness to share does exist — just over half will do so for better deals (54%) or with brands they trust (53%) — but it’s conditional. Fewer than half are comfortable sharing data purely to improve personalization (48%).
This is the central tension commerce teams need to sit with: AI is raising shopper expectations faster than it is earning unconditional trust. The brands that navigate it well won’t be those that deploy the most AI. They’ll be those that deploy it where it genuinely helps — explaining, comparing, narrowing — and stop before it starts deciding.
What This Means for Commerce Teams
The 2026 Commerce Relevance Report doesn’t call for a wholesale reinvention of ecommerce but a more honest understanding of where shopper confidence is actually built.
The data points to a consistent pattern: shoppers arrive with expectations already shaped by experiences outside your site. They test relevance immediately and exit without warning when it falls short. The drop-off isn’t happening at checkout. It’s happening at discovery — at the moment when too many options and too little guidance make moving forward feel harder than walking away.
Conversational product discovery is the most promising way to close that gap. Not because it’s technically impressive, but because it mirrors what shoppers have always wanted: an experience that listens, interprets, and guides them toward a decision they feel good about. The same thing the store associate used to do, now delivered at scale.
The brands that get this right will earn something more valuable than a single transaction. They’ll earn the kind of trust that brings shoppers back — and that, in a market where 71% of shoppers will leave the moment search fails them, is the only sustainable advantage left.