How will AI agents change the way brands do business online? We’re getting closer to finding out.
This week, Shopify introduced the Universal Commerce Protocol (UCP), co-developed with Google — the latest in a string of deals and announcements from the e-commerce platform as it bets on a new AI agent paradigm for shopping. The new system allows brands to sell directly in Google’s AI Mode and Gemini app, so that shoppers can go from research to purchase through an integrated checkout, without ever having to leave their AI chat.
The UCP means US brands on any platform can now use Shopify’s Agentic Storefronts infrastructure via its “Agentic plan” to sell on AI channels, without needing to have a Shopify-hosted online store. This new “open standard” approach is geared towards a future where AI agents from all the different AI chat providers, like ChatGPT, Google AI and Perplexity, can connect with each other and transact with any merchant online. The UCP allows brands to offer customers discount codes, loyalty plans and different billing options.
“We’ve never done something like this before, but this feels like the right time for us to welcome non-Shopify merchants in,” says Shopify president Harley Finkelstein. “This isn’t like social commerce, something that some people use some of the time. It’s the idea that commerce will be offline, online and agentic,” he adds.
Shopify currently hosts around 12% of US e-commerce sites, including brands like Skims, Victoria Beckham and Loewe, according to the company. As well as ensuring its existing customers have easy access to all the latest AI shopping developments when they happen, the universal nature of Shopify’s new UCP provides the company with a huge new merchant acquisition stream for itself.
“As we thought about the rollout, what soon became obvious was that if we’re going to help make agentic commerce a real thing, from a consumer perspective, they should be able to see more than just Shopify brands,” Finkelstein says. “UCP was really a way to make sure that the experience you have on the agentic AI application feels as good as it does on the online store.”
The world’s biggest tech companies are investing in developing so-called AI “agents”, where the tech can autonomously complete complex multi-step tasks on behalf of humans. We’re not quite there yet — Finkelstein, for example, believes we’re a “couple of years” away from most consumers shopping via agentic AI. Meanwhile, Morgan Stanley estimates that agentic shoppers could carry out up to $385 billion in US e-commerce spending by 2030, a forecast that translates to an almost 20% share of online retail for agentic shoppers.
But when this fully autonomous tech does exist, one of its biggest applications is expected to be online shopping on consumers’ behalf. For the tech companies that host these end-to-end AI purchase journeys, there’s a lot of money to be made via new in-chat advertising models and through hosting in-platform transactions.
It’s no surprise, then, that Google, OpenAI and Perplexity are racing against each other to release new shopping-focused features for their AI chatbots to prep for when these agents do arrive. This week, Google became the latest to announce an integrated checkout within its AI Mode, following moves from Perplexity and OpenAI in late 2025. The big common thread in all these AI platform announcements has been a partnership with Shopify.
As consumers increasingly turn to AI search for shopping, tech and marketing teams have been scrambling to relaunch their brands’ DTC websites and optimize their product description pages according to the emerging practice of AIO (AI optimization). Anecdotally, Finkelstein says that in the last few months, Shopify has seen more legacy brands migrate onto the platform than it has in five years. This paradigm shift will also change how brands are discovered online, with implications for advertising budgets and how they’re spent across platforms like Google and Meta.
Shopify, for its part, helped to facilitate retail’s direct-to-consumer brand boom, making it cheaper and easier for anyone to open an online store. Now, Finkelstein sees its role as bolstering the era of AI shopping. “They’re realizing that it’s not like where you can say you don’t do social commerce. You can’t really say you don’t do AI,” Finkelstein predicts. “You need to work with a partner that’s laying these rails for agentic AI, because you cannot opt out of this.”
Merit-based brand equity over pay-to-play
Alongside the launch of this new agentic commerce UCP with Shopify, Google also became the first major AI platform to introduce advertising within AI chat this week, via a pilot of personalized “direct offers” that brands and advertisers can set up for shoppers within Google’s AI Mode. It also announced a new feature that will allow brands to integrate a branded “business agent” within Google’s AI search. The feature, which is already being used by retailers including Poshmark and Reebok, helps brands customize how shoppers’ product questions are answered in their brand voices.
This follows Amazon’s introduction of sponsored prompts in its AI assistant Rufus in late 2025, where brands can bid on follow-up questions within consumers’ AI conversations as they shop. Meanwhile, rival OpenAI is reportedly testing different in-chat advertising models that CEO Sam Altman has promised will be “thoughtful and tasteful”, but the startup is having to build all this from scratch, where Google and Amazon have well-established ads consoles.
“People will expect when they talk to an AI answer engine that they get what the AI “agent” feels is the best product for them personally. They won’t want to see an ad, so brands will most likely need to prompt people to consider their unique qualities in follow-up prompts they can pay for,” says Max Sinclair, CEO of Azoma AI. “There’s such a lack of understanding within luxury brands of how all this tech works right now, so giving them these ready-made products through platforms like Shopify to plug and play works well.”
Although brands can expect these platforms to introduce new advertising models as they seek to monetize their tech, Finkelstein believes that the way the technology behind AI agents works means pay-to-play advertising could become less relevant than with e-commerce as we know it.
“The idea that agentic AI applications could be the best personal shoppers ever comes from the fact that they’re unbiased. All of this is merit-based,” he says. Google, OpenAI, Perplexity and others all pitch their in-chat shopping features’ potential to solve online shopping’s discovery and personalization problems via users’ ability to opt in to saved chat history so their AI models remember their online shopping interactions and preferences. This, Finkelstein argues, could present smaller DTC fashion brands without large advertising budgets with a massive new customer acquisition opportunity.
“I think this era of agentic AI will usher in merit-based retail, where it’s not about who pays the most money to show up, but it’ll determine who deserves to be shown to the consumer based on all the online context. I think that will mean way more business from way more of these DTC modern brands.”
Marketing experts say brands can up their chances of discovery somewhat by updating their websites with early AIO best practices like longer, more detailed product description pages that describe use case scenarios, FAQ content, and richer product imagery that can be parsed by the AI models’ “multimodal” search.
But beyond these housekeeping duties, digital strategists say it’s more important than ever for advertising and creative teams to figure out a consistent, authentic brand story to communicate across their marketing, rather than traditional high-impact short-term campaigns.
“It’ll all be about brand POV and narratives that expose it,” says Mirek Nisenbaum, partner and digital director at branding studio BaseNYC. “The brands that will stand out won’t be the loudest or most clever, they’ll be the ones with the most coherent narratives. So they need to focus on clarity and consistency in their storytelling to be surfaced by the machine.”
While the original e-commerce era led to broad consumer access to thousands more brands than in shopping’s offline past, Nisenbaum points out that this noise became the death knell for many smaller brands, drowned out by larger brands’ advertising budgets. But LLMs, he argues, are looking for “authentic signals” rather than the loudest ones, which could benefit niche brands.
“It opens the door to audiences that are truly theirs,” Nisenbaum says. “The center of gravity will likely shift from “how do we get in front of qualified audiences” to “how do we get recommended to the right people”, so brand POV and its expression will matter a lot.”
At the same time, consumers are spending more time on ad-free platforms like Reddit during their shopping discovery journey, instead, in pursuit of human peer reviews of products over paid-for ads. Nowhere is this more relevant than for luxury brands’ higher-priced items. Experts say this places a renewed onus on brands’ customer service strategies and quality, creating a holistic picture for the AI to crawl.
“Ratings and reviews now serve as the primary trust signals for both shoppers and the AI models interpreting their intent,” says Doug Straton, CMO of UGC platform Bazaarvoice. “Generative systems rely heavily on real customer reviews to understand whether something is worth buying.”
And given that luxury is a high-consideration category and ripe for deep engagement, Nisenbaum says brands are wise to continue investing in high-quality creative campaigns that drive up overall human awareness of their brand, too.
“Storytelling is becoming more important and more fragmented at the same time,” Nisenbaum says. “Stories will have to be so clear and so strong to be able to survive through this fragmentation and retelling by the ‘machine’.”
