Google’s UCP Explained: Will "Agentic Commerce" Replace Search in 2026?

Episode Transcript

00:00:00:02 - 00:00:19:20
Mike: Let’s go, Chris. Handshake first. I missed it, man. All right, sir. Hey, Chris, welcome to Growing E-commerce. We are your hosts, Mike Ryan of smarter eCommerce and Christian Scharmüller. It’s been a minute, Chris. Yeah, we’re in February, and a whole month has gone by.

00:00:19:21 - 00:00:38:22
Chris: Yeah, what was the reason, actually? So much stuff coming in. We are locked in again. I’ve already done two webinars this year, I’m working on a report now, and everything else in between. I know exactly what your calendar looks like, and that brings me to a story.

00:00:38:22 - 00:01:05:18
Mike: I’m going to use that as a transition. Our producer, Kevin, kindly reminded us that we left a cliffhanger on the last episode—our holiday episode for 2025. Would we wear those ridiculous Christmas pajamas? They were beautifully designed and actually, this was an audience question as well. Someone asked me what happened. I’m going to tell the story.

00:01:05:18 - 00:01:29:00
Mike: Hell no, I put my clothes back on. But Chris’s calendar, for people who don’t know, is a brick wall. Every second of every minute is booked or double-booked. We finished recording and he looked at his watch and had to jump straight into a car. One thing happened after the next.

00:01:29:03 - 00:01:47:23
Mike: I saw him hours later, still wearing those pajamas. I was proud; I was really very proud of him.

Chris: I even had a client call! But I’m not that guy. Good thing he didn’t see my awesome Christmas socks, just the upper part of my body. I had to explain to the client why things were a little bit different this time.

00:01:48:00 - 00:02:09:15
Mike: It was a great episode. It would hurt to lose a customer that late in the year.

Chris: I think quite the opposite happened. I think he is more into us now than ever. So, I think it’s a great retention strategy.

00:02:09:17 - 00:02:30:07
Mike: More for us next time. But we’ve got another busy day ahead of us. Let’s get right into it. Did anything noteworthy happen in January?

Chris: I would say at least one very astounding, substantial thing happened. Three letters: UCP.

00:02:30:07 - 00:02:53:23
Mike: Yes, another acronym we have to care about. What is UCP about?

Chris: It stands for Universal Commerce Protocol. There are all these acronyms now. There’s ACP from OpenAI (Agentic Commerce Protocol) and these are inspired by MCP from Anthropic, which is the Model Context Protocol.

00:02:54:00 - 00:03:18:15
Mike: It sounds very positive. Is it? The technology is neutral; it depends on how it gets accessed. It’s perfectly connected to what we have been preaching throughout 2026: Google is very smart in embedding their AI capabilities into their core product, which is Google Search.

00:03:18:17 - 00:03:51:07
Mike: We think the future of search will look significantly different. Agentic commerce will be at the center of it. The idea of the “zero-click buy” environment is something we’ve both talked about. UCP is the engine of agentic commerce. Anyone dismissing that hypothesis as of today is on the wrong side of history.

00:03:51:13 - 00:04:13:10
Chris: Yeah, there are three layers here. Agentic commerce is even bigger than a single channel like search, but it will absolutely change the way people buy through and interact with search. It’s poised to change e-commerce.

00:04:13:10 - 00:04:40:07
Mike: UCP is the infrastructure layer under all that. You have the models and the AI powering the agents, and if you want those agents to do something productive together in a commerce context, UCP is the standard for that.

00:04:40:09 - 00:05:00:13
Mike: The goal is to modularize commerce—to figure out the “Lego bricks” like checkout, product discovery, and payments. These key capabilities are defined, along with extensions like discounting.

00:05:00:19 - 00:05:29:17
Mike: Different platforms will have different capabilities. If you’re PayPal, payment is your capability. But PayPal doesn’t care about product discovery. When agents act on behalf of platforms, merchants, or websites, they go through a “negotiation” where they agree on common capabilities and structured ways to interact. It’s basically guardrails.

00:05:29:21 - 00:05:52:17
Chris: Zooming out from a strategic perspective, what’s the aim here for Google? Let’s fast forward two years. What changes to the commerce environment will we be facing?

00:05:52:19 - 00:06:13:17
Mike: Google isn’t new to building infrastructure. There have been infrastructure races in the past. At first, people thought there would be moats around AI models, but those models are becoming a layer of infrastructure. It’s valuable to own that, just like owning railroad lines or undersea internet cables.

00:06:13:19 - 00:06:33:08
Mike: This is a standards war. It would be awkward for Google to tell their merchants to integrate with OpenAI’s ACP. They want to own the standard to bring the center of gravity closer to their stack.

00:06:33:08 - 00:06:59:15
Chris: Again, I don’t want to let you off the hook. Two years from now, what will be different? It will impact everyone buying online. My friends who don’t care about “standard shopping” or PMax will notice this.

00:06:59:17 - 00:07:17:23
Mike: This depends on the product strategy of the platforms. Zalando is already building with UCP in Europe. In the US, Walmart is building with it. Google is integrating it with AI mode, Merchant Center, and Google Ads.

00:07:17:23 - 00:07:40:02
Mike: In 2023, they announced SGE, which became AI Overviews, and they were talking about ads then. Now it’s 2026 and these things are still rolling out. While there’s hype, this stuff doesn’t move as fast as you think.

00:07:40:08 - 00:08:11:10
Mike: In a year, I think we will see full merchant integrations with UCP leaving beta in the US. What we don’t know is the consumer uptake. Once Google is ready to scale, it depends on how fast people use it.

00:08:11:12 - 00:08:32:08
Chris: On the technical layer, it sounds like they did a good job and collaborated with the right companies. They might be ready to scale sooner than later. The big question is the adoption rate. Are consumers willing to let the buying process be handled by an agent?

00:08:32:08 - 00:08:54:01
Chris: I’m rather conservative here. It won’t happen overnight. That’s the big risk: is it really driving value for the client? Google has miscalculated adoption rates before; they forced adoption of PMax because it wasn’t growing organically as fast as they wanted.

00:08:54:04 - 00:09:18:23
Chris: Give me your take on US e-commerce revenue. In two years, will 50% of revenue go through this new way of shopping, or will it be closer to 10%?

00:09:19:00 - 00:09:39:22
Mike: It will be closer to 5%.

Chris: 5%? That’s a statement.

00:09:39:23 - 00:10:04:07
Mike: Things can change quickly, but in two years, maybe 10%. It won’t be 50%. I don’t care what anyone says. There’s massive hype now, and CMOs are asking what they should do. You have to have a strategy, but it’s not happening overnight.

00:10:04:09 - 00:10:23:20
Chris: It’s a tricky balance. Not long ago, CMOs were worried about NFTs or the Metaverse. That was embarrassing. UCP is a different game. The technology is real and credible. Google has done a better job than OpenAI here. Their partnerships with PayPal and Walmart are immaculate.

00:10:23:22 - 00:10:46:06
Chris: Who faces the biggest risk with UCP? The small retailer, the big online retailer, or the marketplace?

00:10:46:08 - 00:11:16:13
Mike: I think it gets painful in the middle. Small brands and niche products can still do well if they ensure the AI can discover them. But for a medium-sized, multi-brand retailer, it’s challenging. If your brand isn’t well-known, you have to compete on speed, availability, and cost.

00:11:16:15 - 00:11:38:18
Mike: Those advantages tend to concentrate in larger players. If the AI fulfills its mission, it will be doing ruthless price and convenience comparisons. You won’t even get that traffic on your site. This is why I’ve been a “bear” on retail media networks; if retailers don’t get the traffic, they don’t get the first-party data, and the whole thing falls apart.

00:11:38:19 - 00:12:00:10
Chris: I cosign that hypothesis. Google is now moving responsibility to the retailer by launching 10 to 12 new data feeds—and they’ve mentioned dozens more are coming.

00:12:00:10 - 00:12:30:05
Mike: That means the AI is not as capable as it should be yet. They need more structured information. They are looking for conversational attributes—detailed pieces of information and even Q&As provided in your Merchant Center feed.

00:12:30:07 - 00:12:50:04
Mike: Product discovery in AI mode will work largely based on your feed, not just your website. Google could crawl landing pages, but this is a silent admission that the results are not good enough or the risk of hallucination is too high.

00:12:50:06 - 00:13:16:07
Chris: If the assistant hallucinates, people won’t give it a second chance. This is a safety net for Google. It’s a great chance for retailers to impact their organic visibility through a smart data feed strategy. It will require work and testing, but it puts control back in the retailer’s hands.

00:13:16:09 - 00:13:35:12
Mike: If the AI handled everything, it would be completely commoditized. Now, you can have an advantage by doing a great job with your data.

00:13:35:13 - 00:13:54:19
Chris: I spoke with a successful online shop owner in Germany. He asked if they need to restructure or create new roles to conquer these challenges. The big question is: how much energy should they put into this now? If it’s only 5% of revenue, is the trade-off worth it?

00:13:54:22 - 00:14:25:06
Mike: The honest answer is we don’t know for sure. But look at what’s happened: in 2023, 2024, and 2025 people were predicting AGI, and meanwhile, it took the most capable company a long time just to put ads in AI services. That “boat” is moving slowly. You can’t overreact, but you can’t under-react either.

00:14:25:08 - 00:14:45:20
Chris: Savvy retailers might be more flexible and quicker to adapt. But don’t go “all in” yet if you have limited resources. Europe will likely lag behind the US by 6 to 18 months anyway.

00:14:45:22 - 00:15:11:16
Chris: One last hypothesis. We have ACP from ChatGPT, UCP from Google, and MCP from Anthropic. This is the “war of protocols,” like Tesla vs. Edison or Blu-ray vs. HD DVD. Is it “winner takes all”?

00:15:11:16 - 00:15:38:05
Mike: No one is going to concede victory, but Google is in the best position to win. You can bet on Google pretty easily right now. UCP is open source, but it’s backed by all their partners and fits their ecosystem perfectly.

00:15:38:05 - 00:16:00:06
Mike: It’s like the paid search market: Microsoft Bing is still there, but Google owns the lion’s share. Also, Google is smart—UCP is compatible with MCP. They want you in their Model Garden.

00:16:00:08 - 00:16:28:19
Chris: Google has the perfect plug-in and the best starting position. But even they need time. The message to retailers is: there is time. Don’t overreact, but stay connected to the topic.

00:16:28:19 - 00:16:56:24
Mike: What else happened in January? Actually, we are out of time.

00:16:57:01 - 00:17:23:00
Chris: It was a pleasure. Good to be back in front of the microphone.

00:17:23:06 - 00:17:46:16
Mike: Good to see you again, Chris. You’re still my favorite American.

00:17:46:18 - 00:18:05:04
Chris: I’m so relieved to hear that! A weight off my chest.

00:18:05:06 - 00:18:32:10
Mike: Thanks everyone for listening. This has been another episode of Growing E-commerce, brought to you by Smarter Ecommerce. To learn more, visit smarter-ecommerce.com. We appreciate the shout-outs online—it helps a lot. Thanks, and we’ll see you next time.

00:18:32:12 - 00:19:03:05
Chris: See you guys. Bye bye.