AI Mode knows your customers. Does it know your products?

Most e-commerce stores are still tuning their product feeds for traditional Google Shopping: title, price, image, done. However, today’s customers are increasingly asking an AI assistant rather than a search engine. Google AI Mode crossed the 1-billion monthly active users threshold less than a year after launching.

And this exact AI Mode is now becoming even more personal—and at the same time, an ad space. Google has started testing ads in AI Mode even for users who have connected their apps via Personal Intelligence, abandoning its original ad-free approach.

So, the question isn’t whether your products will appear in personal AI answers. The question is: does the AI understand your products well enough to select them?

What actually changed?

Personal Intelligence in a nutshell

Personal Intelligence is a mode where a user connects search with their Google apps, such as Gmail or Photos. AI Mode can then use limited information from these apps—like summaries and insights—to better tailor answers to the specific individual.

Real-world example: A user asks, “What should I pack for next week?” and the AI knows from their emails that they are flying to Norway. The response is no longer a generic checklist, but a specific recommendation for that exact trip.

March promise vs. October reality

Back in March, Google claimed that ads wouldn’t be shown in AI Mode for people with connected apps and that this wasn’t changing anytime soon – while simultaneously hinting that ads would eventually work similarly here, too.

That future has arrived. Google updated its documentation, stating that those same summaries and derived information from connected apps can now also be used to serve ads.

Aha moment: Ads in AI Mode no longer just respond to what the customer typed. They respond to who is writing it and what situation they are in.

What advertisers will see (and what they won’t)

It is important to distinguish what this change means and what it doesn’t. Google lists several safeguard rules:

Opt-in only: Users must connect their apps themselves and can disconnect them at any time.
No direct targeting on Gmail or Photos content: Ads will not be directly based on inbox or photo library contents.
No access for advertisers: Advertisers will not have access to users’ connected apps.
Sensitive data off-limits: Sensitive information will not be used for ad personalization.

And one more essential point: Google has not yet announced any new targeting options, new advertiser controls, or a broader test rollout.

Key takeaway: You won’t get new targeting or new customer data. The only lever in your hands is your product data.

Why e-commerce owners should care more than it seems

Personalization is no longer a bonus; it’s an expectation

Customers want personalization. According to McKinsey, 71% of consumers expect it, and 76% get frustrated when it’s missing.

AI Mode with Personal Intelligence is the direct response to this expectation. However, AI can only offer what it knows about a product. If all it knows about your product is “XYZ Shoes, black, €89,” it has no way to connect it to a customer running their first half-marathon in the rain two weeks from now.

AI selects products based on how well you describe them

In classic search, matching a keyword to a title was enough. In personal AI answers, something else happens:

AI understands the customer’s situation (trip, event, problem, preferences).
It looks for products whose data answers this situation.
It pulls those it understands best into the answer (and into the ad).

If usage scenarios, precise attributes, materials, or compatibility are missing from your feed, your product simply gets left behind at this step. Not because it’s worse, but because the AI doesn’t know who it’s for.

Aha moment: The more personal the customer’s question, the more detailed your data must be. A bare-bones feed won’t cut it in a personal AI response.

What to do today? A checklist for e-commerce stores

You don’t have to wait for Google to expand the test. You can prepare your products for personal AI responses right now:

Describe scenarios, not just parameters. Who is the product for, when will the customer use it, and what problem does it solve? This is the exact information AI pairs with the user’s situation.
Add exact attributes. Material, size, compatibility, gender, age, season. Every missing attribute is one personal question your product cannot answer.
Clean up product titles. Put the most important information first, use natural customer language, and eliminate spam and exclamation marks.
Categorize correctly. Google categories and product_type help the AI place the product into the right context.
Think about long, specific queries. Prompts in AI Mode are full sentences with context. Descriptions should also cover fringe scenarios, not just the primary keyword.
Start with your top products. The ones generating most of your revenue. Feed optimization on these will yield the fastest return.

Key takeaway: Personal Intelligence knows the customer. Your task is to make sure the AI knows your product just as well.

How we can help you with this

When dealing with a few dozen products, titles and attributes can be fine-tuned manually. However, with hundreds or thousands of SKUs, this becomes unsustainable: text remains brief, attributes incomplete, and the feed falls behind what the AI requires. That is precisely the problem we at BlueWinston solved by developing the AI Product Data Suite, enabling you to automatically prepare your feed so that AI can understand it.

AI selects products based on how well you describe them. We at BlueWinston will prepare your feed so that AI understands it.

Our tool helps you with:

AI Product Titles: We create titles matching how customers actually search.
AI Product Descriptions: We write descriptions with usage scenarios that provide AI with context for personal answers.
AI Product Attributes: We fill in missing attributes that AI uses to match products with the customer’s situation.
AI Google Categories and Product Types: We place products into the correct categories.
AI Product Long-tail KWs: We cover highly specific, long-tail queries typical for AI Mode.

Once your data is in order, our P.Max Labelizer ensures your budget flows into products that actually sell.

Conclusion

Google has opened ad space in the most personal version of search yet, but it hasn’t given advertisers new targeting tools – only new rules of the game. E-commerce stores with product data rich and accurate enough for AI to understand and recommend will win. Start optimizing your feed today while the competition is still waiting.

Frequently Asked Questions (FAQ)

What is Personal Intelligence in Google AI Mode? It is an opt-in mode where a user connects search with their Google apps, such as Gmail or Photos. AI Mode then uses summaries from these apps to tailor responses to the specific person.

Will advertisers see the content of users’ Gmail or Photos? No. Google states that advertisers have no access to connected apps and ads are not directly based on inbox or photo content. Sensitive data is not used for ad personalization.

Can I target users with Personal Intelligence in Google Ads? Not currently. This is a test phase, and Google hasn’t announced new targeting options or controls. You can influence whether your product appears primarily through the quality of your product data.

Which product data is most important for personal AI answers? Descriptions containing usage scenarios, complete attributes (material, size, compatibility, target audience), accurate titles, and proper categorization. This information enables AI to match a product to a specific customer’s situation.

Is feed optimization worth doing if the feature is still just a test? Yes. Enriched titles, descriptions, and attributes improve performance in classic Shopping and P.Max campaigns today. When the test expands, you will already be ahead of competitors who waited to get started.