Skip to main content

AI for shopping: what it is, why it matters, and examples from leading brands

Explore Klaviyo AISee what's new
Set as Preferred Source

More people are turning to AI for shopping to guide their next purchase, just as they would with Google and TikTok.

As a B2C brand, you can likely assume that some of your customers are discovering your products through AI agents like ChatGPT and Gemini. In fact, 39% of consumers have purchased a product recommended by AI in the past 6 months, according to Klaviyo’s 2026 AI Consumer Trends Report.

The research, which grouped 8,000 consumer respondents into 4 personas based on their AI usage and trust, found that 85% of AI Enthusiasts (26% of total respondents) have used AI while shopping over that same timeframe.

What’s especially interesting is that AI for shopping is happening across channels that brands own and those they don’t. On the one hand, consumers use AI agents like ChatGPT, Google Gemini, and Claude to research, compare, and click through to purchase. On the other hand, brands are integrating AI throughout the customer experience.

Today, B2C brands are using AI for shopping to:

  • Scale 1:1 personalisation across marketing and customer service.
  • Guide shoppers toward purchase with AI customer service agents.
  • Support customers faster and more comprehensively with AI and self-serve options for customer service.

Here’s what AI for shopping looks like in practice, and how real-life B2C brands are embracing it to increase sales and deepen customer loyalty.

What is AI for shopping?

AI for shopping is the process through which consumers use AI to discover, compare, and buy products. The AI shopping experience has some distinct qualities compared to traditional search:

  • Shopping is conversational and contextual. Rather than searching for a product and hopping through websites to conduct research, consumers are having conversations with AI agents like they would with a personal shopper. They’re getting recommendations that align with previous searches and answering follow-up questions from the AI agent to find exactly what they need.
  • Consumers use more descriptive, emotional language with LLMs. According to our AI Consumer Trends Report, 30% of AI searches contain 8+ words, and 78% of consumers include emotional or personal context at least some of the time. For example, someone might search, ‘I’m looking for accessories for my office that feel warm and inviting. My office is too cold, and I’d rather my space feel cosy so I like sitting down to work’.
  • Product discovery, comparison, and purchasing can happen in one place. With agentic commerce, consumers can find new products, compare options, and even check out in one conversation with an AI agent, rather than searching across websites and marketplaces.

What do consumers want from AI shopping experiences?

AI can accomplish more for product discovery and comparison in fewer steps. But when customers use AI for shopping, they want more than just speed. Consumers look to AI shopping experiences for:

  • Personalisation: 55% of all consumers have had at least one ‘aha’ moment when AI impressed them with accuracy or personalisation, according to the research behind our AI Consumer Trends Report. AI agents can offer more in-depth, contextual personalisation in a way that traditional search and social media algorithms can’t.
  • Simplification: Consumers’ most common ‘aha’ moment happened when AI clearly explained a complex concept in simple terms, according to our research. AI agents can give shoppers valuable context about new or expensive purchases by consolidating a lot of information to a single, quick response.
  • Efficiency: 62% of consumers would prefer that AI remember their preferences over re-explaining them to a human customer agent, according to the research behind Klaviyo’s 2025 Online Shopping Report. To save customers time and effort, brands can train their own AI agents to reference customer data and, when necessary, hand off conversations to humans with complete context preserved.

How do consumers use AI to shop?

43% of consumers have used AI when shopping for non-essential products in the past 6 months, according to our AI consumer trends research. Most often, they use AI to search for products in electronics or technology, travel or hotels, and health or beauty.

Our research also found that 70% of consumers have used AI multiple times to find the best product or deal, and 69% have used it multiple times for general product recommendations.

But even as many shoppers experiment with AI, not everyone is ready to hand off the entire buying process to it. 85% of consumers have at least some trust in AI to provide accurate and personalised shopping recommendations, but only 27% completely trust AI to do so.

Where do AI shopping experiences break down?

Consumers know inadequate AI when they see it. And when they see it in their shopping experience, that translates to missed revenue for your brand. Here are consumers’ biggest red flags when using AI for shopping:

  • Delayed answers: According to Klaviyo’s 2025 AI shopping index research, 75% of consumers have abandoned a purchase because they couldn't get instant answers. Structuring your product catalogue for LLM discoverability and training your AI customer agent to answer FAQs are the new requirements for converting customers.
  • Inadequate personalisation: After receiving poorly personalised content, 21% of consumers become less likely to open messages from a brand, according to our AI consumer trends research. This underscores the importance of AI product recommendations that pull from real-time customer data, so your messages reflect real context on the customer’s end.
  • Detached support: Less than half of consumers (43%) say AI has improved customer service quality, and 42% of people feel neutral about it, according to our AI consumer trends research. The No. 1 consumer concern about AI in brand experiences is customer service interactions that make them feel like they’re talking to a robot. Train your AI customer agent on data and brand context so support doesn’t feel generic.
  • Misinformation: 27% of consumers are most uncomfortable with AI when it shares inaccurate, misleading, or low-quality responses, according to our AI consumer trends research. Inaccurate answers from AI agents about your products can damage your brand’s reputation.

What does the consumer AI shopping experience look like?

The consumer AI shopping experience varies across AI surfaces and brand websites. Let’s look at a few examples of the AI shopping experience today:

Perplexity

Here’s an example of a consumer using Perplexity AI to shop for an eyelash curler.

The user searches, ‘find me an eyelash curler that has good reviews and is under $20’. In the Answer tab, Perplexity finds a few options, creates a comparison table, and links out to where the consumer can make a purchase.

Screenshot of the Perplexity search engine that shows a table of product recommendations for an eyelash curler with good reviews under $20, based on a one-line conversational prompt. The table is white on a grey-ish green background with simple black text and blue links that lead the user to websites where they can buy the recommended eyelash curlers.

In the Shopping tab, Perplexity displays more in-depth product cards for top picks. These product cards include star ratings, summaries of customer reviews, and multiple links for where to buy the item. Each product card also cites specific review sources, which often include the brand’s website, Reddit threads, TikTok posts, and third-party review sites.

Screenshot of the Shopping tab in Perplexity, showing product listings for specific eyelash curlers with good reviews under $20. In a 3-column format, there are 3 images of eyelash curlers: first, there is a thin, black eyelash curler with red handles that look like scissor handles. Second, there is a bulkier eyelash curler in black and grey. Third, there is a shiny, pink, and sleek eyelash curler that looks like a more traditional eyelash curler. Underneath are smaller images of two additional eyelash curlers, one inside packaging and the other with black comfort pads on the handles.

In 2025, Perplexity partnered with PayPal and Venmo to launch Instant Buy, an agentic check-out option for US users.

Google

In this Google AI Search example, the shopper asks about a pair of Reeboks that are stylish and good for running and weightlifting. Google provides a comparison table, then asks a follow-up question to help narrow down available options.

Screenshot of a Google AI Search showing product recommendations for Reeboks that are stylish and good for running and weightlifting, based on a conversational prompt. There are 3 small images of running shoes in a 3-column format, the first being a black shoe, the second a neon-yellow shoe with bright orange and black zig zags, and the third being a grey and black shoe. Beside the 3 images, there is a large, zoomed in image of the bright orange shoe from the second thumbnail, with more small thumbnails of the product underneath.
Screenshot of Google AI Search showing product recommendations for Reeboks that are stylish and good for running and weightlifting, with a follow-up question for the user: ‘Would you like to see specific colourways for one of these models, or are you looking for a particular price range?’ Above the question are 4 images of shoes, each of a medium size, in mostly black and white, neutral colours.

In 2025, Google launched agentic check-out in Google AI Search and Gemini with select retailers, so consumers can find and eventually purchase what they’re looking for within the same interface.

Brand-side AI customer agents

Brand-side AI customer agents are AI agents customers can interact with on a brand website. While many brands primarily use AI agents for customer service, they can also assist shoppers during the research phase the same way ChatGPT or Claude do.

Here’s an example of an brand-side AI customer agent from beauty brand Half Magic. The conversational chat experience invites consumers to ask for product recommendations, check order statuses, share product usage tips, and more.

Screenshot of Half Magic’s AI agent on their branded homepage. There is a hero image of a close-up of a person with dark hair and a wide smile with diamonds embedded in their white teeth. The person is closing their eyes, and you can see blue and purple glitter eyeshadow on their eyelids. On the right side of the screen there is a white chat window with a grey text bubble where the shopper can choose what they would like help with from the brand’s AI assistant.

In this exchange, the shopper prompts the assistant to find face makeup options that work well for sensitive skin. The AI agent follows up in a fun, friendly tone with an initial recommendation that links to a product listing.

Screenshot of Half Magic’s AI agent sharing a product listing based on a conversational prompt asking for face products good for sensitive skin. In a black text box at the top of the page, there is a question: ‘What are the best face makeup options for sensitive skin?’ Underneath is an answer from the AI agent, showing a bold headline for a product: ‘Beam Trap Illuminating Balm Light Beam/Deep Beam’, with an image of a product in neon yellow packaging. Underneath are additional selection buttons that say, ‘Track Order’, ‘Subscription Frequency’, and ‘Skip Next Order’. There is a response text box at the bottom of the screen.

5 examples of real brands using AI shopping to get results

B2C marketers’ top technology stack priority for 2026 was to expand their AI adoption, according to Klaviyo’s 2026 future of consumer marketing research. If that’s you, here’s a little inspiration from 5 real-life brands that are already delighting customers with AI shopping experiences:

1. Naked Wardrobe creates a concierge DTC experience with their AI customer agent

Luxury apparel brand Naked Wardrobe wanted to create the kind of concierge experience that was formerly only available at brick-and-mortar stores, where an attentive, expert team could anticipate and respond to each customer’s needs in real time.

After launching an AI customer agent trained on the brand’s customer data and product catalogue, online shoppers can get prompt, personalised recommendations from a simple web chat.

‘Now, AI is styling our customer, answering her questions, and selling her the right piece, all at two in the morning when our team is asleep’, says James Thorngren, Naked Wardrobe’s director of marketing.

Over 90 days, the AI customer agent resolved 94% of product recommendation queries autonomously.

2. Harney & Sons’ AI customer agent resolves product recommendation queries

Family-owned tea brand Harney & Sons uses an AI customer agent alongside a self-serve customer portal and a shared helpdesk where human and AI agents can handle customer requests across channels.

The AI customer agent works 24/7 to answer support and recommendation queries autonomously. If the conversation needs a human, the AI agent passes it on to the human customer service team’s helpdesk with full context preserved. But in 60 days, the AI agent resolved 77% of product recommendation queries on its own.

3. Folk Clothing’s AI customer agent takes over simple support

For years, the team at London-based fashion brand Folk Clothing struggled with customer service logistics that relied on a basic email inbox and a phone line that was easily overwhelmed. So, the brand implemented an AI customer agent that could provide instant answers to common customer questions, without any input from the team.

In 90 days, Folk Clothing’s AI customer agent resolved 53% of support conversations. ‘It’s already made a noticeable difference to my workload’, says Hayley Scott, ecommerce coordinator at Folk Clothing. ‘I was relieved to see that even our most loyal clients, who are used to high-touch service, haven’t noticed a drop in quality.’

4. NANUK’s AI customer agent frees up their human reps to focus on high-impact touchpoints

NANUK makes waterproof cases that safeguard valuables anywhere. Their customer service team is staffed with human reps who’ve been with the brand for over a decade. The team creates the most value when they’re navigating novel situations, explaining case technology, and managing social relationships with customers.

The problem was they didn’t have time or bandwidth to focus on any of that. The human customer service team at NANUK was responding to customer queries one by one via email and phone, and the question more often than not was a frustratingly simple one: ‘Will it fit in the case?’ Answering it slowed down responses for all queries.

Now that NANUK has launched an AI customer agent trained on their brand context and customer data, they’ve automated answers to those kinds of repetitive technical FAQs. In 90 days, the AI agent resolved 100% of product recommendation queries and return and exchange queries, and 84% of all chat queries.

‘The AI agent does not replace human agents’, says Sophie Morin, director of digital marketing and ecommerce. ‘It’s another layer of service that gives us a faster, more accurate way to answer simple questions. It allows us to bring our customer service to another level.’

5. K9 Ballistics’ AI customer agent respects the brand policy fine print

K9 Ballistics, a brand that makes durable, high-quality pet supply products backed by real guarantees, was using an AI customer agent that was too generous in its interpretation of the brand’s warranty. It often promised out-of-policy product replacements and issued resolutions whenever customers demanded them.

‘It was very much a people pleaser’, recalls Sarah Rutten, customer experience and AI specialist at K9 Ballistics. It also focused exclusively on post-purchase questions, without the ability to answer pre-purchase questions or provide personalised product recommendations to shoppers.

Klaviyo Customer Agent solved those problems. When Rutten ran quality assurance on early Customer Agent responses, she found that it could tell customers ‘no’ when necessary. Unlike the brand’s prior AI agent, Customer Agent sticks tightly to their warranty policy. No unauthorised product replacements or other policy exceptions.

‘When customers push the boundaries, Customer Agent doesn’t overpromise’, Rutten says. ‘It has guardrails built in that prevent it from just bending to the customer.’ In 90 days, it resolved 91% of discount-related queries and 88% of product recommendation queries autonomously.

How Klaviyo positions your brand to succeed with AI for shopping

AI shopping isn’t the future. It’s already here. If you haven’t already, now is the time to plan for how you’ll embed AI across your customers’ shopping experience.

Klaviyo is the AI agent-first operating system for B2C. Customer profiles, marketing channels, and agentic AI all run on the same real-time customer data, so Composer, your AI marketing agent, and Customer Agent, your AI customer service agent, both work from what your customers actually bought, browsed, and asked about.

With built-in and agentic AI across marketing and customer service, Klaviyo can help you take advantage of AI shopping opportunities at each phase of the customer journey:

  • Discovery: Optimise your content for LLM visibility with well-structured product catalogues and a robust customer review strategy. Launch campaigns faster with Composer, Klaviyo’s AI marketing agent, and drive better engagement with more precise targeting and fresh content.
  • Nurture: Use predictive AI to develop more 1:1 customer relationships. With Klaviyo, you can send messages at each shopper’s preferred time, on the channels where they’re most likely to engage next. Create voice and tone guardrails for all AI-generated content, so every interaction feels on-brand and human.
  • Conversion: Customer Agent, Klaviyo’s AI customer service agent, can answer pre-purchase questions, share personalised recommendations, and help shoppers surface relevant offers.
  • Retention: Grow customer loyalty with retention messaging rooted in when customers buy, what they’re interested in, and their lifetime value. Use Klaviyo’s predictive analytics to identify at-risk customers before they churn and trigger personalised win-back flows.
  • Analytics: Use the Klaviyo app for ChatGPT or connector with Claude to gain access to your analytics through simple text prompts. Lean on your AI assistants to capture zero-party data that can enrich customer profiles and improve future personalised messaging that helps shoppers find exactly what they’re looking for.

Put Customer Agent to work for your brand.

Get a demo