28th November 2025

Navigating the Future of eCommerce: ChatGPT’s New Shopping Research Tool & What It Means for Your Business

Jordan Boyne, SEO Director at Rush Ventures

Over the past 18 months, AI assistants have moved from experimental toys to serious shopping infrastructure.

OpenAI’s new Shopping Research feature in ChatGPT is the latest step: a built-in product discovery experience that builds personalised buyer guides, compares options and links to merchants directly from the chat interface.

Crucially, this is not happening in isolation. Google is rolling out its own agentic shopping stack including conversational shopping in Search and Gemini, plus agentic checkout that can track prices and complete purchases on participating merchants using Google Pay. Perplexity has launched “Shop like a Pro”, a native AI shopping assistant with one-click checkout (“Buy with Pro”) and visual search (“Snap to Shop”) for U.S. Pro users.

Together, these moves signal a broader shift we’d describe as Generative Engine Optimisation (GEO) which means optimising for AI agents (ChatGPT, Gemini, Perplexity, etc.) that increasingly mediate search, discovery and purchase decisions.

What Is ChatGPT’s Shopping Research Tool?

Shopping Research in ChatGPT allows users to discover products via conversation, rather than filters and category trees. Users describe what they’re looking for, budget, context, constraints and then ChatGPT can:

ask follow-up questions to clarify intent

assemble a curated buyer’s guide with shortlists, pros/cons and comparison points

surface product cards with images, key specs, review summaries, price indications and links to merchants

In parallel, OpenAI is rolling out Instant Checkout, allowing users in supported regions to purchase certain products directly within ChatGPT via the Agentic Commerce Protocol. This has been initially launched for U.S. Etsy sellers, with Shopify merchants and PayPal wallet support being layered in.

1. The scale of LLM-driven shopping

This isn’t a niche behaviour anymore:

OpenAI reports that 700M+ people now use ChatGPT each week for everyday tasks, including finding products.

Adobe’s 2025 research on U.S. consumers found that 39% have already used generative AI for online shopping, with over half planning to do so in 2025; 55% use generative AI for research and 47% for product recommendations.

A Bain & Company study found that around 68% of LLM users rely on AI tools for researching and summarising information, and 42% already ask them for shopping recommendations.

Capgemini’s retail research shows 71% of global shoppers want generative AI integrated into their shopping journeys, especially younger demographics who expect highly personalised, digitally streamlined experiences.

For many consumers, “searching” is increasingly asking an AI engine, and that AI is now capable of taking them all the way from idea to purchase.

2. Behavioural drivers: convenience as the core motivator

A lot of this can be explained through simple behavioural science, as we know that convenience is one of the strongest motivators in digital behaviour.

Google’s own framing of agentic shopping is explicitly about removing “endless scrolling, tracking prices and jumping between tabs”. ChatGPT, Perplexity and Gemini are doing the same thing by collapsing multi-step, high-friction journeys (search → compare → check reviews → find best price → checkout) into a single guided conversation.

For consumers, that means:

fewer decisions about where to search

fewer interfaces to learn

less cognitive load in comparing long lists of similar products

From a behavioural perspective, AI-driven shopping is just the logical next step in the long-running trend of platforms removing friction wherever they can.

Who Is Most Affected?

The impact is broad, but it’s not evenly distributed.

Most exposed / biggest upside

D2C and multi-brand eCommerce in “high-consideration” categories
Electronics, home & garden, sports/outdoors, beauty, wellness and premium apparel where buyers want guidance, comparisons and reassurance are the categories most often highlighted in AI shopping examples and demos.

Brands with complex, spec-heavy products
Think smart home, fitness equipment, cameras, audio, and skincare. These are ideal for buyer guides and agentic assistance, because users benefit when the AI can translate specs into simple “for you, this is the best trade-off” narrative.

Retailers heavily reliant on organic discovery
Merchants whose growth has historically come from Google Search, Google Shopping, marketplaces and comparison sites now need to consider AI assistants as a parallel discovery layer, not just an afterthought.

Less impacted (for now)

Purely impulse-driven, low-consideration commodities where price and availability trump education.

Service-first local businesses (e.g., plumbers, hairdressers) where the interaction is more about booking than comparing product specs, though local intent in LLMs is growing fast.

How to Prepare: Your Strategy for AI-First Shopping

1. Optimise your product data for AI discovery

AI agents can only recommend what they can understand. That makes clean, structured data non-negotiable.

Implement and maintain schema.org Product, Offer and Review markup with accurate names, specs, pricing, availability and review data.

Ensure product feeds (whether via Google Merchant Center, Shopify, custom APIs, etc.) are complete and up to date. AI shopping experiences increasingly sit on top of these feeds (or similar schemas), whether in Gemini’s Shopping Graph, ChatGPT’s Shopping Research or Perplexity’s Merchant Program.

Treat review quality and sentiment as a ranking factor; assistants lean heavily on aggregated reviews when summarising pros and cons.

2. Build content that matches conversational intent (and looks like a buyer’s guide)

ChatGPT’s Shopping Research and Google’s AI Mode both respond best when there is content that mirrors the way people actually ask questions:

Create buyer guides around “jobs to be done”:

“Best [category] for [use case] under [budget]”

“Which [product type] is right for [specific scenario]?”

Structure content in decision-friendly formats such as shortlists, comparison tables, “good/better/best” recommendations, and simple “if X, choose Y” guidance.

Use headings framed as questions, and language that reflects the way customers really talk. This helps LLMs understand and reuse your content in a way that feels natural in conversation.

3. GEO and SEO: continuity, not a reset

One important nuance to be aware of is that GEO doesn’t replace SEO, it compounds it.

Most of the work required to succeed in AI shopping overlaps strongly with already-good SEO practice:

Clear information architecture and crawlability

High-quality, intent-led content that actually answers questions

Comprehensive product data and structured markup

Strong review/profile pages that signal trust and authority

Where GEO adds new emphasis is in:

Treating AI answers / assistants as a primary surface, not just the blue links beneath them

Designing content and data for summarisation and recommendation, not only page-level ranking

So, an “agentic strategy” isn’t about throwing away the SEO playbook. It’s about doubling down on the most durable, under-prioritised parts of it (structured data, rich content, reviews, first-party data) and making sure they are consumable by AI agents as well as search crawlers.

4. Measure AI-driven discovery, not just SERP position

Traditional SEO metrics (rank, impressions, CTR) only tell part of the story. As AI layers thicken, consider measuring:

Assistant-driven referrals: traffic or conversions attributed to ChatGPT, Gemini, Perplexity or other AI sources where you can see them. Adobe’s data already shows that AI-referred visitors tend to be more engaged (more pages, lower bounce rates) than average.

Category-level visibility: are your hero categories prominently represented in AI-generated shopping guides when you test them qualitatively?

Share of voice in AI recommendations (where third-party tools exist): what percentage of suggested products in a given AI experience belong to your brand vs competitors?

We’re early in this measurement ecosystem, but the direction of travel is clear: “Do we show up in AI shopping journeys?” will become a core KPI.

5. Treat checkout and API integrations as “watch and prepare”, not “implement now”

Both ChatGPT and Google are starting to support agentic checkout, but the reality is still highly constrained by region and platform:

OpenAI’s Instant Checkout currently supports purchases by US ChatGPT users from US-based Etsy sellers, with Shopify merchants being added and other regions planned, but not yet live.

Google’s agentic checkout is “starting to roll out” with eligible US merchants using Google Pay, including select Shopify brands, and will expand over time.

A recent PayPalOpenAI partnership means PayPal wallet payments will be available for Instant Checkout from 2026, also starting with specific markets and merchant types.

In other words, this is not something every brand can turn on tomorrow even if they wanted to.

The pragmatic stance today is:

Audit your technical readiness: are you on a platform (e.g., Shopify, Stripe-based stack) that is likely to be early in these ecosystems?

Stay close to your platform and payments partners: many of the first-wave integrations will come via existing commerce platforms (Shopify, PayPal, Stripe, Google’s Shopping Graph) rather than bespoke builds.

Review developments regularly with your tech and agency teams. Treat agentic checkout as a roadmap item to revisit quarterly as eligibility and geo-coverage expand.

The strategic point to emphasise is to be ready and informed, not prematurely over-engineered.

Why Now Is the Time to Take GEO Seriously

Putting this information together, a few themes emerge that can anchor your approach to GEO and agentic commerce for next year and beyond:

AI shopping is already mainstream
Large shares of consumers are using LLMs for research and shopping, and platforms are racing to capture this intent with native shopping agents.

Discovery is shifting from lists to conversations
Instead of scrolling product grids and SERPs, users increasingly get a single, summarised answer with a short list of recommended products. If you’re not on that shortlist, you effectively don’t exist in that journey.

The fundamentals still matter more than ever
Good SEO practice (structured data, authoritative content, strong reviews) is now the input layer for GEO. The brands that have quietly invested here are best placed to win in AI-mediated shopping.

Platforms will diverge in their specifics
Google’s AI shopping is naturally linked to Merchant Center, Performance Max and Google Ads; ChatGPT is building an agentic commerce layer with partners like Etsy, Shopify, Stripe and PayPal; Perplexity is building its own merchant programme and Pro-led shopping experience.
The strategic foundations (data quality, content, reviews, experimentation) are shared, but tactics will become platform-specific.

Early movers get a compounding advantage
The brands that start now auditing data, strengthening buyer guides, aligning SEO and GEO will be best placed as AI assistants capture more of the discovery and consideration funnel.