Large Language Models (LLMs), such as those that power modern AI assistants, are beginning to cross paths with online shopping in ways that will redefine the market.
In the early stages of generative AI, the first AI agents (including ChatGPT) were based on older, indexed content from a selection of websites. This naturally limited their utility for information retrieval.
However, expectations for artificial intelligence have significantly changed with the rollout of Google’s countermeasure to ChatGPT: Gemini.
Google’s AI had the advantage that Google already indexed content on the web regularly. Google thus pushed the expectations for up-to-date information in the responses from AIs.
An Innovation in Holiday booking
Google was thus also the first to integrate artificial intelligence with flight and hotel searches.
This innovation made it easier than ever to search for, for example, flights or hotels. You now only had to write a simple prompt such as:
“Find the cheapest flight to Malaga in july.”
Due to Google’s competitive advantage, they had a setup where they could easily add AI functionality to their existing services using AI and thus develop a unique value proposition within search.
But the development naturally didn’t stop there…
ChatGPT cannot compete with Google’s access to up-to-date information. Instead, they have developed the so-called “agent mode,” where their AI has a general understanding of where to look for information. When the user writes a prompt that requires up-to-date information, it checks relevant websites in real-time.
This naturally means that it takes longer to give the user an answer, which would normally be a disadvantage. However, ChatGPT turned this to their advantage by allowing the search to continue in the background, making it possible to make more complex queries and simply notifying the user when the search is complete.
In this way, one can now perform very complex searches such as:
“Find the cheapest combinations of flights and hotel for a Mediterranean destination for two adults and two children for 14 days in July. The hotel must be located less than 1 km from both the beach and the city center. The hotel must have a pool.”
The purchase is just one click away
Combined with this new way of scanning the market, ChatGPT has introduced “Instant Checkout,” which means the user no longer needs to click through to a series of different websites to gather and complete their purchase.
The AI agent functions as an intermediary that can execute purchases across platforms. When it has found the ideal flight and hotel, it presents a total price and a single “Buy Now” button. With the user’s permission and stored payment information, the agent handles the bookings with both the airline and the hotel portal in the background. The complex trip is booked with one click.
But this move by OpenAI doesn’t leave Google on the sidelines. On the contrary. Where ChatGPT has to build agents that interact with the open internet, Google already owns the entire underlying infrastructure for e-commerce: “Google Merchant Center” and their own payment solution, Google Pay.
For years, companies have fed their complete product catalogs – prices, stock status, sizes, colors, and specifications – directly into Google’s system to be able to advertise via Google Shopping.
When Google truly connects the Gemini model’s conversational intelligence with the Merchant Center database, a trading platform emerges that no one else can match.
Bruger: “I need a new pair of running shoes for asphalt, max 1000 kr.”
Gemini: “Okay. Based on your previous purchases (via Google Pay) and your searches, you prefer [Brand X]. I can see in the Google Merchant Center that [Sports Store A] has last year’s model on sale for 799 kr. [Webshop B] has the newest model for 999 kr. with free shipping.
I can also see that [Local Store C] (which I know from Google Maps is close to your work) has the newest model in stock in your size. Should I order them online from [Webshop B] or reserve them for pickup at [Local Store C] this afternoon?”
From SEO to AIO and PFO
In the aforementioned scenario, the user never visits a webshop. The entire buying journey – from inspiration and research to comparison and purchase – takes place in the conversation window with Gemini. Payment is handled instantly via Google Pay.
For businesses and marketing experts, this signals a fundamental change in the terms of digital marketing. For approximately twenty years, the focus has been on SEO (Search Engine Optimization) to achieve a high ranking on the search results page.
The future belongs to AIO (AI Optimization) and PFO (Product Feed Optimization).
It will become irrelevant how beautiful your website is if your product data is not 100% accurate, up-to-date, and detailed in the Google Merchant Center. If your stock status is incorrect or you are missing specifications, Gemini simply won’t recommend your product.
The fight for the customer shifts from winning a click on a search page to becoming the preferred choice for an AI assistant that already has the customer’s trust and payment card ready.
