BeaconSpec

← All articles

Agentic commerce • Market briefing

The rule became a scoreboard. Your store now has an AI readiness score.

A few weeks ago the market said the rule out loud: a store an AI agent can read gets recommended, one it can't gets skipped. This cycle it did something new. It started keeping score, in public, and most businesses on a custom stack aren't on the board.

David Soden  •  Market briefing  •  6 min read
A stadium scoreboard listing team names down two ranked columns under a clear blue sky, standing in for the new public ranking that grades retailers on how ready they are for AI shopping agents.
For a while, being readable to an AI agent was a quality nobody could see or measure. This cycle, someone put it on a board and started keeping score.

If you read the last briefing, you know the rule the market finally said out loud: a store with a clean, agent-readable feed gets recommended, and one without it gets skipped. That was the market describing how AI shopping agents already choose. It was true, but it was still a description, the kind of thing you could nod at and put off dealing with.

This cycle the description turned into a number. Digital Commerce 360 and ReFiBuy launched the first public AI-readiness ranking, grading 1,000 top retailers on a 0-to-100 scale for whether an agent can find, read, and buy from their catalog. The rule you could shrug off two weeks ago now has a scoreboard attached, and a lot of businesses are about to learn where they land on it.

The rule stopped being something you could nod at and defer. It became a score, in public, that a competitor or a customer can look up.

What's happening

An AI agent doesn't browse your store. It queries, compares, and picks.

When someone asks an AI assistant to find and buy something, it doesn't shop the way a person does, clicking around, reading a review, squinting at a photo. It queries a set of sources it can read, compares what comes back, and picks. If your product data is messy, incomplete, or simply unreadable to the systems agents use, you aren't a worse option in that comparison. You aren't in it. The agent has no way to include a store it can't parse.

Until now, "can an agent read you" was an invisible quality. You couldn't see it, so it was easy to assume you were fine, or to treat it as a someday problem. That's the thing this cycle changed. The market took the invisible quality and made it a public grade, which means it stops being a matter of opinion and starts being a matter of record.

What's changed since our last briefing

Being big does not make you agent-ready. The scoreboard just proved it.

In our last briefing the news was that the market had named the rule in plain language: recommended or skipped, decided by whether an agent can read you. This cycle two names you may not know, Digital Commerce 360 and ReFiBuy, turned that rule into a quarterly ranking. They score 1,000 of the largest retailers 0 to 100, and the first thing they measure is exactly the thing that matters here: whether AI agents can access, read, and transact with your catalog.

The detail worth sitting with: the retailer ranked #814 by sales took the top readiness spot. A store far down the size list beat household names on the one measure agents actually use. Size and readiness turned out to be different things, and now there's a public list that shows it. If the biggest names can score badly, a custom-built store that nobody has optimized for agents is not scoring well by accident.

A person at a desk studying charts and rankings on a monitor while making notes, the new routine of checking where a business lands on a public readiness score.
A public score changes the conversation. Now a buyer, a partner, or a competitor can look up where you rank before you get a chance to explain.

Three numbers behind why the score matters

How big the agentic-commerce prize is getting, how fast the buyers are arriving, and what being unreadable already costs in plain web traffic.

$3.5T projected agentic commerce value by 2031, up from $8B in 2026 +350% growth in agentic commerce users by 2031, to 1.3 billion people ~50% drop in clicks a site gets once an AI answer sits above the results

Sources: Juniper Research (agentic commerce transaction value and user growth to 2031); Pew Research Center (click-through study across roughly 68,000 searches).

Why this matters to you

A slow site costs you some visitors. Being unreadable costs you the whole sale.

Think about the kinds of risk a business already plans for. A slow website loses you some visitors. A weak search ranking loses you some traffic. You watch those, you spend to fix them. Being unreadable to a shopping agent is a different animal, because the agent never presents you as an option in the first place. You don't lose a slice of the sale. You lose the whole thing, and you never see it happen, because you were never in the comparison to begin with.

A car's illuminated gauges and speedometer glowing in low light, standing in for a readiness score that reads from zero to a top number you can't move by advertising harder.
A readiness score behaves like a gauge, not a billboard. You don't move it by advertising harder; you move it by making your data readable to the agents doing the reading.

Here's what a public score adds on top of that. Until now, being invisible to agents was at least private. Now the grade sits on a list a buyer, a partner, or a competitor can pull up. And the number in front of it keeps growing: Juniper Research puts agentic commerce transaction value at $8 billion this year, rising to $3.5 trillion by 2031, with the buyer base going from under 300 million people to 1.3 billion. Meanwhile Pew found that once an AI answer sits above the search results, a site loses roughly half its clicks. Being unreadable isn't a someday worry. It's measured lost traffic today, now with a scoreboard next to it.

Why we're built for this

A clean, agent-readable feed is a specific, buildable thing. That's the layer we work in.

The top signal in that new ranking, whether an agent can access, read, and transact with your catalog, is the exact problem BeaconSpec was built to solve, and it's the one the big platforms already handled for their own merchants. If you're on Shopify or a similar platform, you have a path. If you run a custom or home-grown stack, you don't, and that's who scores low or doesn't appear at all.

You almost certainly already have the raw material: a product catalog, an inventory system, a booking engine, an API some internal tool already talks to. What's missing is the translation layer that turns what you have into the structured, agent-readable presence the emerging standards expect. That's what we build. We turn your existing APIs into one curated server that agents discover and buy from over the open UCP standard, so an agent reads your real product data instead of guessing at a scraped page. You keep your own systems and decide exactly what's exposed. No replatform. You just stop being the store an agent can't parse.

You already have the APIs. What you're missing is the layer that turns them into something an agent can read, trust, and buy from.

A separate, related free tool

Different question, worth two minutes: which AI crawlers are even allowed to read your site in the first place. Our free AI Visibility Checker writes you a correct robots.txt for that. It matters most if you're on Cloudflare, whose defaults change on September 15, 2026.

Set your AI crawler rules →

No sign-up, about two minutes, and it doesn't touch your search ranking.

The one thing to remember

The market stopped describing readiness and started scoring it.

There's now a public scoreboard for whether AI agents can read and buy from you, and being big doesn't put you near the top of it. If you run a custom stack, you're probably not scored well, or scored at all, and to an agent, unscored reads the same as skipped.

This is a market briefing we run every cycle, tracking what's actually changing in agentic commerce so you don't have to piece it together yourself.

The rankings will get more detailed, more of them will launch, and the number in front of the whole thing keeps climbing. The rule hasn't changed since last cycle, readable gets recommended, unreadable gets skipped. What changed is that it's now written down where people can see it. The only real choice left is whether you find out your score before it costs you, or after.

BeaconSpec exists for exactly this shift: making your business discoverable and transactable by AI agents over the UCP standard, instead of invisible to them.

See what we do  •  Read more articles

Related reading: Recommended or skipped: the new rule for AI shopping agents


David Soden writes about agentic commerce, automation, and building durable technical systems for businesses. Photography via Pexels (Phil Evenden, Vitaly Gariev, Griffin Wooldridge); figures cited are drawn from the public reporting named above.