Ask Google AI Mode for a quiet espresso machine for a small kitchen and the answer can now include sponsored product cards. The retailer never wrote a headline for that placement or built a campaign for it. Google picked the product, then pulled the image, price and details from the retailer's Merchant Center feed. With the newest formats, Gemini also writes the line explaining why that product suits the question.

Google has been heading here for more than a year. In May 2025 Google said ads "may appear below and integrated into AI Mode responses" in the US. In February 2026 its head of ads and commerce, Vidhya Srinivasan, confirmed Google was testing a new AI Mode format "to showcase retailers that offer those products, clearly marked as sponsored". At Google Marketing Live in May 2026 it previewed AI-powered Shopping ads that write a custom explainer for each query. For a Shopify brand, that means your feed is the creative in these placements.

Below: where these ads run by market, how Google assembles the unit, which feed fields carry the most weight, how to write for conversational queries, what reporting still hides, and a 30-minute audit you can run this week.

1B+
Monthly AI Mode users, per Sundar Pichai at Google I/O 2026
12
Countries where ads can show inside AI Overviews, in English
+5%
Average lift in conversions or value Google reports for AI Max for Shopping

Where are Shopping ads showing inside AI Mode and AI Overviews?

There are two AI surfaces in Google Search and they follow different rules. AI Overviews is the summary at the top of a normal results page. Google says ads can show above or below it in all 200+ markets where AI Overviews appear. Ads inside the overview itself are narrower: English only, on mobile and desktop, in Australia, Canada, India, Indonesia, Kenya, Malaysia, New Zealand, Nigeria, Pakistan, the Philippines, Singapore and the US. Text and Shopping ads from existing Search, Shopping and Performance Max campaigns are eligible.

AI Mode is the separate conversational tab. Google has described ads there as a test in the US, in English, and its main help page on AI ad placements still covers AI Overviews only. Eligibility was set out in May 2025: advertisers using Performance Max, Shopping, and Search campaigns with broad match, including AI Max for Search. In September 2026 Google's Ads Liaison, Ginny Marvin, confirmed a "small experiment" letting exact and phrase match text ads serve in AI Mode where intent is explicit. Search trackers also reported Shopping carousels appearing inside AI Mode answers, rather than below them, from late August 2026.

Two big markets are missing. The UK has had AI Mode since July 2025 but is missing from the list of markets with ads inside AI Overviews. Japan has had AI Mode in Japanese since September 2025 and is also missing.

MarketAI Mode for shoppersAds inside AI OverviewsAds inside AI ModeMerchant Center AI performance insights
United StatesYesYes, EnglishYes, testing in EnglishYes
United KingdomYes, since July 2025No, above or below onlyNot announcedNo
CanadaYesYes, EnglishNot announcedYes
AustraliaYesYes, EnglishNot announcedYes
SingaporeYesYes, EnglishNot announcedNo
IndiaYes, incl. HindiYes, EnglishNot announcedYes
JapanYes, Japanese since Sept 2025No, above or below onlyNot announcedNo
Indonesia, Malaysia, PhilippinesYesYes, EnglishNot announcedNo
Availability by market as of 8 October 2026Not announced means Google has published no rollout for that market. Sources: Google Ads Help, Google Merchant Center Help, The Keyword.

There are two AI surfaces in Google Search and they follow different rules.

How does Google build a Shopping ad inside an AI answer?

A Shopping ad has never had a headline field. The card is assembled from your product data: title, main image, price, sale price, shipping, merchant name and, where available, ratings. What changes inside AI answers is the matching. Google uses broad match, keywordless targeting, Performance Max and Shopping campaigns to read a long, conversational question, and for ads inside AI Overviews the ad must also be relevant to the content of the overview itself.

Two newer products move even more of the copy into Google's hands. AI Max for Shopping, announced on 30 April 2026 as a closed beta in all markets and languages, adds text customisation that writes ad copy for Shopping ads from your feed. Google's own example talks about reading feed details like "fabric softness, material durability and fit". Then at Google Marketing Live, Google previewed AI-powered Shopping ads for Performance Max and AI Max for Shopping, which pick the most relevant products and write an explainer of why each one fits the shopper's question. Google says they will "soon be available".

Gemini now writes more of the words a shopper reads about your product, working from your feed and your product page. Give it thin source material and you get a thin explanation, or none.

Which product feed fields matter most for AI Mode ads?

Every attribute helps. Some do far more work when the query reads like a sentence, and those are the ones we check first on Shopify feeds.

On Shopify, most stores sync products to Merchant Center through the Google and YouTube channel or a feed app. That sync usually maps your product title and body copy straight across and leaves the richer attributes blank. Highlights, details, material and pattern rarely come through unless you store them in metafields and map them on purpose, or add them in a supplemental feed. Two stores selling near-identical products can look very different to Google as a result: one sends a name and a paragraph, the other sends a name, a paragraph, eight highlights and twenty specs.

AttributeRule or limitWhy it matters in AI answers
titleUp to 150 charactersMain matching signal. Lead with product type and the attribute shoppers ask about.
descriptionUp to 5,000 charactersRaw material for Gemini explainers. Cover who it is for and when to use it.
product_highlight2 to 100 highlights, 150 characters eachReady-made reasons a product fits a question.
product_detailUp to 100 name and value pairsSpecs that settle comparison questions: dimensions, capacity, wattage, fabric weight.
gtin and brand14 digits per GTIN, brand up to 70 charactersLinks your offer to the right product in Google's catalogue so it can be compared.
color, size, material, pattern, gender, age_groupRequired to separate variantsAnswers follow-ups like 'in navy' or 'for wide feet'.
image_link and additional_image_link500 x 500 minimum from 31 January 2027, up to 10 extra imagesThe visual card. Lifestyle angles help in carousels.
price, sale_price, shippingMust match the landing pageAI answers compare offers on price and delivery. Mismatches trigger disapprovals.
google_product_category and product_typeOne category, product_type up to 750 charactersCategory context for matching broad questions.
Feed attributes ranked by impact on AI answer placementsLimits from the Google Merchant Center product data specification. In January 2026 Google also announced dozens of new conversational commerce attributes, such as answers to common product questions and compatible accessories or substitutes, starting with a small group of retailers.

How should you write product titles and descriptions for conversational queries?

Shopify product names are often brand names. "The Aspen" works on your site because the photography and context do the explaining. In a feed it tells Google almost nothing. Shoppers in AI Mode ask for things like a warm jumper that will not itch for a cold commute, so your title and description need to contain the facts that answer that sentence.

Keep on-site naming as it is and fix the feed with a supplemental feed or a feed app. A title formula that works for most categories is brand, product type, key attribute, material, audience, then variant. "The Aspen" becomes "Aspen Merino Wool Crew Jumper, Men's, Midweight, Navy".

  • Put the product type near the start, because cards truncate long titles.
  • Open the description with two plain sentences: who it is for, and what problem it solves.
  • Name real situations: small apartments, sensitive skin, long-haul flights, toddlers. Conversational queries are full of them.
  • Write specs as text as well as in product_detail: "holds 1.5 litres", "weighs 280 grams".
  • Skip promotional claims, ALL CAPS and "best" language. They breach Shopping policy and add nothing to an explainer.
  • Keep feed and product page consistent. AI Max and Performance Max read your landing page too, and contradictions weaken both.
  • If AI writes your titles, Google's spec has structured_title and structured_description attributes to declare it.
Read: Shopify products in ChatGPT and agentic storefronts

What can you see in reporting for AI Mode and AI Overviews ads?

Not much. Google's help page states that "Google Ads currently doesn't offer segmented reporting when ads show within Search AI Overviews." We have not seen a separate AI Mode segment either. Those impressions and clicks are blended into your normal Search, Shopping and Performance Max numbers.

The one new window is on the organic side. Merchant Center's AI performance insights report shows share of voice, a competitor average, query frequency and how many of your products show for conversational shopping queries in AI Mode and AI Overviews. It is available for English queries in the US, Canada, Australia, India and New Zealand. Google says it covers organic traffic only: "Paid Ads traffic isn't included in this report." It also shows no click data.

QuestionVisible?Where
Paid impressions inside AI OverviewsNo separate segmentBlended into Google Ads totals
Paid impressions inside AI ModeNo separate segmentBlended into Google Ads totals
Organic product visibility in AI answersYes, as share of voiceMerchant Center, 5 markets
Clicks from AI answers to a productNoNot reported
Which long queries triggered your adsPartly, with no surface labelSearch terms reports
What you can and cannot measure todaySources: Google Ads Help on ads and AI Overviews; Google Merchant Center Help on AI performance insights.
Read: Google Merchant Center now shows how your products perform in AI Mode and AI Overviews

Until Google adds a segment, use proxies. Track the share of search terms with six or more words over time. Enrich the feed for one group of products and leave a comparable group alone, then compare impressions and conversion value over four to six weeks. It will not isolate AI Mode, but it tells you whether better feed data is paying off.

Your next step

Is your product page good source material?

Google's explainers read your product page as well as your feed. Our free teardown grades a Shopify product page on clarity, detail and trust signals and lists the fixes.

Should brands outside the US prepare now?

Yes. In Australia, Canada, Singapore, India and the English-language SEA markets, Shopping ads can already show inside AI Overviews, and the same feed quality decides which products get there. In the UK and Japan, ads currently sit above or below the overview, but AI Mode itself is live, so feed work done now carries straight over when Google extends ads there.

The order of work is the same everywhere. Fix disapprovals first. A product Google has not approved cannot appear anywhere. Then enrich the 20 to 50 products that drive most of your revenue, since those are the ones Google has the most signal on. Leave the long tail for a later pass or a rules-based fix in your feed tool. If you sell across several countries from one Shopify store, check each country's feed separately: currency, shipping and sale prices are where cross-border feeds most often drift out of line with the product page.

Feed work also pays off in places Google does not control. ChatGPT shopping, Perplexity and Microsoft Copilot all lean on structured product data. One clean, descriptive feed serves all of them, so we treat feed work as a priority whatever Google's ad rollout schedule looks like in your market.

A 30-minute product feed audit checklist

Open Merchant Center and your top 20 products by revenue. Work through this in order.

  • Minutes 0 to 5: check Diagnostics for disapproved and limited items, especially price and availability mismatches.
  • Minutes 5 to 10: read your top 20 titles. Does each lead with the product type and a key attribute, inside 150 characters?
  • Minutes 10 to 15: check GTIN and brand coverage. Any missing GTIN on a branded product is a gap.
  • Minutes 15 to 20: check whether product_highlight and product_detail are filled in on those 20 products. In the Shopify feeds we audit, both are usually empty.
  • Minutes 20 to 25: check images. Main image at least 500 x 500 ahead of the January 2027 enforcement, plus at least three extra angles including one in use.
  • Minutes 25 to 30: confirm price, sale price and shipping match the product page, then confirm a Performance Max or Shopping campaign is live in each market where AI Overview ads run.

Key takeaways

  • Ads inside AI Overviews run in English in 12 countries, including the US, Canada, Australia, India and Singapore. The UK and Japan only get ads above or below.
  • Ads inside AI Mode are still a US test, open to Performance Max, Shopping and broad match Search campaigns.
  • You cannot opt out, so your feed is the main lever you control.
  • Titles, descriptions, product_highlight and product_detail feed the AI explainers. Fill them for your top sellers first.
  • Google Ads does not segment AI placements. Merchant Center's AI performance insights covers organic visibility only, in five markets.
  1. 1.Google Ads Help: About ads and AI Overviews
  2. 2.The Keyword: New ways AI in Search helps your business (May 2025)
  3. 3.Google Ads Help: Google Ads highlights of 2025
  4. 4.The Keyword: What to expect in digital advertising and commerce in 2026
  5. 5.The Keyword: A new generation of ads for the AI era of Search (GML 2026)
  6. 6.Google Ads: AI-powered Shopping ads
  7. 7.The Keyword: AI Max for Shopping campaigns
  8. 8.Google Ads: AI Max for Shopping campaigns (performance figure)
  9. 9.Search Engine Journal: Google launches AI Max for Shopping and Travel campaigns
  10. 10.Search Engine Roundtable: Google Ads tests restrictive match types in AI Mode
  11. 11.Keywords Everywhere: Ads in AI Mode, confirmed steps and sightings
  12. 12.Google Merchant Center Help: About AI performance insights
  13. 13.Google Merchant Center Help: Product data specification
  14. 14.The Keyword: New tech and tools for retailers in an agentic shopping era (January 2026)
  15. 15.The Keyword: Sundar Pichai's I/O 2026 keynote
  16. 16.The Keyword: Introducing AI Mode in the UK
  17. 17.The Keyword: AI Mode expands to more languages (September 2025)