Agentic Commerce Is Here. What Do You Build First?

AI agents now find, compare, and recommend products for shoppers, and on Shopify, they can complete the checkout too. For most merchants, the question has moved on from whether this is real to what to build first.

The shift is already in the numbers.

For a while, agentic commerce was something to prepare for "eventually." As we now know, eventually has arrived. 

Adobe found that traffic from AI sources to US retail sites grew 393% year-over-year in the first quarter of 2026, and by March that traffic converted 42% better than non-AI traffic. On Shopify, AI-driven traffic and orders tripled year-over-year in Q2.

The holiday season looks set to continue the trend. After last year's Cyber Week, Salesforce estimated that AI and agents influenced 20% of global online orders, a broad measure that includes AI-powered recommendations and customer service. This year, Adobe expects traffic from AI assistants to US retail sites to grow 130% over the holiday season.

With growth like that, most merchants know they need to act. The real challenge is deciding what to build, and in what order.

There's more to build than any team can take on at once.

Shopify has given merchants a lot to work with: Shopify Catalog, the Universal Commerce Protocol, Agentic Storefronts, and a more capable Sidekick. Around all of that sits a growing list of advice on structured product feeds, agent-readable content, and new ways to measure a channel most analytics weren't built to see.

Most of it is good advice on its own. Put together, it's more than one team can tackle at once, and not all of it matters equally for every business. A brand with a focused, well-described catalogue starts from a different place than one with thousands of SKUs and thin product data. A B2B seller with long, relationship-driven sales cycles faces a very different agent landscape than a DTC brand with frequent repeat orders.

Some merchants are already making these calls differently. A PYMNTS Intelligence survey of 60 US merchants found that merchants plan to build a separate shelf for AI Agents; only 28% said they'd give agents their full product range on the same terms as other channels. About a third would offer the full range with different prices, promotions, or delivery choices, and another 28% would limit agents to selected products.

That's why a single agentic readiness list doesn't help much. The better question is: for your business specifically, what's the first thing worth building, and why that one?

Start with what your data already shows.

Before we'd recommend a specific build, we'd start with what the business's own data already shows us. For example, how much AI-referred traffic is already coming in without being counted? Could an agent recommend this store with confidence today? And where would getting this right make the biggest commercial difference?

Starting with the business's data matters even more for agentic commerce. It's the newest and fastest-moving channel, and the one most merchants are least set up to measure.

In our experience, the gap usually isn't the technology. Most stores already have the basics AI agents need to find and recommend products, or are close. The harder part is knowing which piece matters most for a specific business, and doing that first.

Three questions to consider before you commit a budget.

1. Do you know how much AI-referred traffic you're already getting?
   Most analytics tools weren't built to recognize visits that come from AI assistants, so they often get lumped in with "direct" or "unknown" traffic. That makes the channel look smaller than it is. Some agent sales may not show up at all: Shopify notes that Google Analytics and custom pixels don't fire in direct checkout on Meta's AI surfaces. Without a reliable baseline, every decision after this one is more of a guess, so it's worth looking into off the bat. If you're unsure about how your analytics are tracking, we can help with that.

   
   

2. If an agent looked at your store today, could it recommend a specific product with confidence?
   That depends on structured product data, complete feeds, and readiness for the newer discovery surfaces Shopify and others have built. It's a separate question from how your site looks to a person, and often a more basic one. Adobe found that product pages scored lowest for machine readability, averaging 66% against 75% for homepages. That matters, because on Shopify, half of AI-referred sessions land straight on a product page.

   
   

3. Where would getting this right move your business most?
   The best first step depends on your catalogue, average order value, sales cycle, and how much of your traffic already looks agent-driven. A merchant with thin product data across thousands of SKUs may need to start with feeds. A merchant already seeing AI-referred visitors land on well-described products may get more from fixing measurement first. The general answer and the right answer for your business are rarely the same.

What's next?

Agentic commerce will keep growing, with new protocols, new places to discover products, and new ways for agents to buy. That makes the underlying question more important: for your business, what's the most valuable thing to build first, and how sure are you before the budget goes out?

That's the thinking behind Signal, iamota's consulting practice, which reads what the business is already telling you before deciding where to invest. We dig into a merchant's own data and come back with a clear read on the one or two moves to make first, and why.

If you're looking at everything agentic commerce makes possible and aren't sure where to start, we're happy to help you work it out.

Reach out to us at hello@iamota.com.