Why Agentic Commerce Has a Prioritization Problem.

AI agent access to your store is largely settled.

In October, Shopify's Fall '26 Edition made every Shopify storefront and checkout readable by AI agents. An agent shopping on someone's behalf can now search a store, build a cart, and complete checkout once the shopper approves. There was no setting to switch on; it arrived with the platform.

Shopify’s move reflects a wider expectation that AI agent traffic will become a significant revenue lever for merchants, and early data points in the same direction. In September 2026, shoppers who reached US retail sites from AI tools converted 90% better than other visitors and generated 83% more revenue per visit, according to Adobe. The volume is set to grow, too: Salesforce predicts that AI chat agents will account for 20% of all ecommerce traffic this holiday season.

With numbers like these, you might expect investment decisions to get easier. Instead, they add urgency to a growing list of options for merchants to weigh.

The largest retailers are placing different bets.

The data may be encouraging, but there's no agreement yet on what to do about it. Even some of the biggest names in retail haven't settled on where AI belongs in the shopping journey.

Prada recently launched a ChatGPT app focused on product discovery, which it describes as a pilot. Gap has moved in the other direction and says shoppers will be able to buy its products directly in Google's AI Mode and Gemini. Target launched an AI shopping assistant in beta and says it's too early to tell whether it grows basket size.

Even some of the best-known names in retail, with experienced teams and resources to invest in tooling, are making very different bets. 

Why your agentic plan shouldn’t mirror your competition.

When there's no clear playbook, the natural instinct is to watch what competitors do and match it. In agentic commerce, that instinct runs into three problems.

1. Their business isn't built like yours.
   Agentic channels pay back differently depending on how a business makes its money. For a subscription brand, most of a customer's value arrives on later orders, so the useful question is whether AI-referred shoppers go on to subscribe and keep subscribing. A multi-brand group needs to know which of its brands agents recommend and which they pass over. A merchant selling to both consumers and wholesale buyers has to weigh agentic work against the rest of its B2B roadmap.

   
   

2. It spends your budget on their priorities. Every dollar used to match a competitor's feature is a dollar that can't go to the opportunity your own data points to, and that opportunity may have nothing to do with agents at all.

   
   

3. You see what they launch, rarely what it earns.
   Launch announcements are public. Results almost never are, and many of this year's AI features are still labelled betas and pilots. Copying a competitor's launch means copying a bet whose outcome you can't see.

Matching a competitor means borrowing their answer to a question about your business.

Begin with your own numbers.

Before funding a new agentic capability, we'd encourage merchants to size the decision against their own business. Building first and measuring afterwards often leads to investments that look promising on paper but don't always deliver once they're live.

A handful of questions will tell you most of what you need to know:

• How much revenue already comes from shoppers referred by AI tools, and is it growing?

• Which of your products do agents recommend, and which do they miss?

• Do customers who arrive through AI come back at the same rate as everyone else?

• What would it cost to fix what agents get wrong on your store, and what would that fix be worth?

• Does a tool you already pay for cover some of this?

With those answers in hand, agentic commerce becomes a set of line items you can rank against modeled ROI, alongside everything else competing for budget. Some will merit funding now, some can wait a year, and a few may not merit funding at all. Knowing which is worth as much as anything you end up building.

How Signal helps.

We built Signal for decisions like this one. Between apps, platform features, rebuilds, campaigns, and now agentic channels, most merchants have more options than they can act on, and few dependable ways to estimate the return on each.

Signal is iamota's consulting practice. It works from a merchant's own commerce data, combining AI-assisted analysis with our years of Shopify experience, and produces a ranked, costed list of opportunities with the confidence level of each number stated. Imperfect data is the normal starting point, and we work with it as we find it. From there, we help put the findings into practice, with our team or another partner, and revisit the results every 90 days, which matters in a market changing this quickly.

We've spent years building on Shopify, and Signal adds the other side. We help you find the investments worth making, then build them with you.

Agentic commerce will keep changing, and the list of things you could fund will only get longer. The merchants who come out ahead will be the ones who sized each option against their own business and funded the right ones first.

If agentic commerce is on your planning list and you'd like to know where it belongs, we'd be glad to start with your data. Reach out to our team anytime at hello@iamota.com.