A new benchmark grades retailers on how ready they are for AI shoppers
On July 15, Digital Commerce 360 and ReFiBuy launched the AI Commerce Rankings, a quarterly benchmark that scores how prepared the Top 1000 North American retailers are for AI-driven shopping and agentic product discovery. It bolts a readiness layer onto the industry's most-cited ecommerce ranking, which until now measured one thing: online sales. "Digital Commerce 360's Top 1000 has long helped the industry understand ecommerce scale," said Brian Warmoth, editor-in-chief at Digital Commerce 360. The reframing matters. For a decade, the Top 1000 told retailers where they stood on revenue. This new cut tells them whether AI systems can even find and read their catalog, which is fast becoming the more urgent question for anyone selling online.
The timing tracks a real shift in how shoppers arrive. Adobe Analytics reported traffic to US retail sites from generative AI tools rose 693 percent year over year during the 2025 holiday season and stayed elevated into 2026, up 393 percent year over year in the first quarter. "The AI Commerce Rankings add an essential new layer of insight to the Top 1000, helping leaders evaluate readiness," said Jon Love, research data manager at Digital Commerce 360. We would put it more sharply. A rising share of demand now originates inside AI tools, and most retailers have no instrumentation to measure it, let alone influence it. A public benchmark forces the conversation into the open.
Four signals that decide whether an agent can sell your product
The rankings score retailers on four signals. Bot friendliness measures whether AI agents can actually access and read a retailer's catalog data. AI source traffic measures the share of web visits originating from AI-powered discovery. Diversity of AI sources checks whether that traffic comes from multiple engines rather than one. And 90-day momentum tracks whether AI-source traffic is rising or falling. None of these appears on a traditional ecommerce dashboard. All four are now leading indicators of whether an autonomous shopping agent will surface your product or route the sale to a competitor whose data it can parse cleanly and trust enough to recommend.
Bot friendliness is the one that should sting. Many retailers spent the past two years hardening their sites against bots, and with reason. Akamai recently put AI-driven traffic at nearly half of all commerce traffic, and a lot of it is scraping and fraud. The same defenses that block bad bots also block the shopping agents you now want to reach. This benchmark exposes an uncomfortable tension inside most security and ecommerce teams: the crawler you throttled last quarter may be the buyer you court this quarter. Retailers need policies that separate legitimate commerce agents from abusive ones, and most do not have them yet.
Scale stopped guaranteeing discovery
The headline finding is that revenue rank and AI readiness have come apart. The retailer sitting at number 814 by online sales ranks first in AI readiness. Early leaders named by Digital Commerce 360 include Online Labels, Nixon, Fashionphile, Everlane and Brooklinen, a list of mid-sized specialists rather than the usual giants. That should reframe how large retailers think about their moat. Scale still wins on price, logistics and marketing budget. It confers no automatic advantage when an AI agent evaluates products on clean data, clear attributes and machine-readable availability. A focused brand with a tidy catalog can outrank a national chain in the channel that is growing fastest.
We read this as good news for challengers and a warning for incumbents. The retailers on that early leaderboard tend to sell well-defined catalogs with rich structured product data, which is exactly what agents need to reason about a purchase. Sprawling assortments with inconsistent attributes, thin descriptions and messy availability signals are harder for a model to trust, and agents route around uncertainty. The lesson for a large retailer is that catalog hygiene is now a growth lever, and it deserves to be funded like one. The work of standardizing product data across thousands of SKUs is unglamorous. In an agent-mediated market, it is also where discovery is won.
Why a benchmark beats another AI opinion piece
The market is full of commentary on agentic commerce and short on measurement. Until now, a retailer asking whether it was ready for AI shopping got projections and vendor decks, not a score against peers. A quarterly, comparative benchmark changes the internal conversation. It gives a head of ecommerce a number to bring to the board, a way to show movement quarter over quarter, and a competitive frame that budget committees understand. We have long argued that what gets measured gets funded. Putting AI readiness into the same database that already tracks online sales is the most practical thing anyone has done to move it from slideware to roadmap.
The caveat is that a benchmark is only as useful as the behavior it drives. It would be easy to treat a good AI readiness score as a vanity metric and chase it with superficial fixes. The signals that matter, catalog access, structured data, and genuine agent-sourced demand, require sustained engineering and merchandising work. Retailers should use the ranking as a diagnostic and a prioritization tool, then instrument their own funnel to confirm that improved readiness actually converts into AI-sourced revenue. A rising score that never shows up in sales is a warning that the effort went to the wrong layer, and it is better to catch that early.
What to do on Monday
For the CxOs we write for, this benchmark is a prompt to stand up a new KPI. Start by measuring how much of your traffic and revenue already originates from AI tools, because most teams genuinely do not know. Then audit whether your catalog is readable by agents: structured product data, accurate real-time availability, and bot policies that admit legitimate shopping agents while still blocking abuse. Those three fixes move the needle on the rankings and, more importantly, on actual discovery. None of them requires a moonshot. All of them require someone to own the problem, which today is nobody's explicit job at most retailers we talk to.
The strategic point is that AI discovery is becoming a distinct channel with its own economics, and it deserves the same rigor retailers gave paid search fifteen years ago. The companies that treated search as a discipline early compounded an advantage that laggards never recovered. Agent-mediated commerce is at the same inflection. A public benchmark that decouples readiness from scale tells every mid-sized retailer that the window is open and every large one that its size will not hold it. The safer place to be is measured and improving. Large and invisible to the systems that increasingly decide what shoppers see is the position to avoid.
