TL;DR
- Existing Search, Shopping and Performance Max campaigns already qualify for AI Mode ads, and there's no opt-out.
- AI Mode showed about 4.3 products per shopping answer, versus 22.5 on a regular results page.
- Merchant Center's 6 conversational attributes give Gemini the product answers it uses to pick and describe items.
- Enrich your top 20–50 SKUs first, then compare them against an untouched set for 30 days.
What are Google AI Mode ads?
Google AI Mode ads are Sponsored Shopping and text ads that show inside or next to Gemini's written answers in AI Mode. As of November 2026 they run in the US and 10 more countries, and they can also appear on the follow-up questions a shopper asks next.
Can you opt out of ads in Google AI Mode?
No. Search, Shopping and Performance Max campaigns enter AI Mode through the normal auction, and Google offers no setting to target or exclude it on its own. The October 28, 2026 expansion to 10 more countries kept the same rule. Results are blended into your standard campaign reports.
How do I get my products shown in Google AI Mode?
Run Shopping or Performance Max on key products, since both qualify for direct and contextual matching. Then add Merchant Center conversational attributes through a supplemental source, starting with question_and_answer and related_product. Each product can carry up to 30 Q&A pairs, and the fields don't affect product approval.
💬 Quick reality check: If you run Shopping or Performance Max, your ads are probably already showing inside Google AI Mode. You just can't see it in your reports, and you can't turn it off.
Do you know what Google shows when a shopper types "best waterproof hiking boots for wide feet under $200" into AI Mode? Not ten blue links. Not a grid of 20 products. A written answer, a handful of products, and a small "Sponsored" label on some of them.
That's Google AI Mode ads.
Google says AI Mode passed 1 billion monthly users in May 2026, and ads now run inside its answers, including on the follow-up questions shoppers ask next.
The brands that win here won't be the ones with the biggest budgets. They'll be the ones whose product data can actually answer a shopper's questions. Most feeds can't do that yet. Here's how AI Mode ads work, how Google decides which products to show, and the Merchant Center changes that help yours get picked.
What Are Google AI Mode Ads?
AI Mode is Google's conversational search experience, powered by Gemini. Instead of typing two keywords, shoppers ask full questions and keep asking follow-ups.
Google AI Mode ads are Sponsored text and Shopping ads that appear inside or next to those AI-generated answers. Here's where each format stands:
| Format | Status (as of Nov 2026) | What it does |
|---|---|---|
| Shopping ads in AI Mode | Live: US + 10 more countries | Product image, price and merchant inside the AI answer |
| Text ads in AI Mode | Live: US + 10 more countries | Standard Search ads within or below the answer |
| Multiturn ads | Live in the US | Ads can show on follow-up questions, not just the first one |
| Conversational Discovery ads | Testing (US) | Gemini tailors the ad creative to the shopper's exact question |
| Highlighted Answers | Testing (US) | Ads appear inside AI-generated recommendation lists |
| Direct Offers | Pilot | Personalized discounts shown when Google detects high purchase intent |
| AI-powered Shopping ads | Announced for the US (a Search format) | Gemini writes a custom explainer for why your product fits the query |
Sources: Google Ads & Commerce Blog, GML 2026; PPC Land, Nov 1, 2026.
Four things every eCommerce brand needs to know:
- No new campaign needed. Existing Search, Shopping and Performance Max campaigns are automatically eligible through the normal auction.
- No opt-out. You can't target AI Mode on its own, and you can't exclude it separately from Search.
- No separate reporting. AI Mode impressions, clicks and conversions are blended into your standard campaign reports.
- You don't write all the copy. Google can add an AI explainer next to your ad. You can't write, edit or approve it. Gemini builds it from the query, the AI answer and public sources like your website.
That last point is the one that should change how you think about this.
How Google Matches Ads in AI Mode
Google uses two matching methods:
- Direct query matching: the shopper's question already has clear buying intent ("buy wide-fit hiking boots"). The ad matches the query.
- Contextual matching: the question is longer or exploratory, and buying intent shows up in the AI's answer or suggested next steps. The ad matches the query and the AI response.
Here's the catch: exact and phrase match keywords only qualify for direct matching. Broad match, AI Max for Search, Shopping and Performance Max qualify for both. Since AI Mode questions are mostly the long, exploratory kind, tightly built keyword accounts reach only part of this inventory. Shopping and PMax reach all of it.
"In AI Mode, queries are, on average, three times longer than traditional searches."
Shashi Thakur, VP & GM of Google Search Ads, Google Ads & Commerce Blog, Oct 28, 2026
Why AI Mode Shows Fewer Products (And Why That Raises the Stakes)
Google claims click-through rates are similar between sponsored and organic product listings inside AI Mode. Let's be honest about that one: it's based on 15 days of internal US data (Aug 20 to Sep 3, 2026), and Google published no actual rates. Treat it as directional.
The number that matters more comes from outside Google. A Productrise study of 100,000+ US and UK queries (July 2026) found AI Mode showed products on about 23% of shopping queries, versus about 88% in standard search. When it did show products, it averaged 4.3 listings, versus 22.5 on a regular results page.
This is not a minor gap. It's structural. In AI Mode you're not fighting for a spot in a 22-product grid. You're fighting to be one of four products Gemini decides to recommend.
Why Your Product Feed Is Now Your Ad Copy
Do you know how a traditional Shopping ad gets picked? Title keywords, price, bid. AI Mode doesn't work that way.
A shopper in AI Mode asks things like "Will this sofa fit through a 30-inch doorway?" or "Does this case work with my phone model?" Gemini picks products, and writes the explainer next to your ad, based on the product data it can find. Vague data gets you vague (or zero) recommendations.
Your standard feed was built to answer four questions: what is it, what does it cost, what does it look like, where do I buy it. It was never built to answer "is this right for me?"
That's exactly the gap Google built conversational attributes to fill.
The 6 Merchant Center Conversational Attributes, Explained
Google launched conversational attributes at Google Marketing Live in May 2026. They're six optional Merchant Center fields that help AI Mode, Gemini and Business Agent understand the nuances of your products. They're available in all countries, they don't affect product approval, and Google recommends adding them through a supplemental data source rather than your main feed. (Google Merchant Center Help)
| Attribute | What it tells the AI | Example | Priority |
|---|---|---|---|
question_and_answer | Answers to real shopper questions | "Is it machine washable?" → "Yes, cold wash, hang dry." | 🔴 Start here |
related_product | Accessories, required parts, substitutes, often bought with | often_bought_with:id:SKU123 | 🔴 Start here (AOV lever) |
item_group_title | One clean name for a whole variant family | "Organic cotton men's T-shirt" | 🟡 If you have lots of variants |
variant_option | What makes each variant different | size:8,shoe width:narrow | 🟡 If you have lots of variants |
popularity_rank | How well the product sells vs. the rest of your catalog (percentile) | 95.5 | 🟡 Quick win from sales data |
document_link | PDFs like manuals, size charts, spec sheets | https://yoursite.com/size-guide.pdf | ⚪ Electronics, appliances, furniture |
Does it actually move the needle? Google says that in testing with lululemon, attributes the brand submitted were used in relevant AI Mode product recommendations about 50% of the time (reported by Nudge, Oct 2026). That's one brand and a Google-reported number, with no sales impact attached. But it's the clearest signal so far that this data gets used.
💡 TIP: Know the Q&A rules before you write 30 FAQs. Each product can have up to 30 Q&A pairs, each question and answer up to 1,000 characters, with a 10,000-character total per product. Google says: no prices or dates, no keyword lists, and don't repeat what's already in your title, description, product highlights or product details. (Google Merchant Center Help)
The related_product relationship types
This one is easy to underuse, so here's the full menu Google supports:
- often_bought_with: the phone case for the phone
- accessory: the webcam for the desktop
- required_part: the battery the lamp needs to work
- part_of_set: the chair that matches the table
- substitute: a comparable product that does the same job
- different_brand: the identical product under a cheaper house brand
Up to 30 per product, matched by your product ID or GTIN. often_bought_with and accessory are the ones that help Gemini build bundles, and bundles mean bigger carts.
💸🧮 Planning spend for Shopping and PMax? Use our free Google Ads Budget Calculator to model your budget against your revenue goal.
How to Add Conversational Attributes: Step by Step
You don't need to enrich your whole catalog at once. Start with the products that drive your revenue, prove it works, then scale.
Step 1: Pick your top SKUs
Start with the 20–50 products that carry most of your revenue, plus anything you're actively pushing. The long tail can come later.
Step 2: Mine the questions people actually ask
You already have this content. It's just scattered:
- Customer support tickets and live chat logs
- Product reviews (look for questions buried in reviews)
- Existing FAQ sections on your product pages
- Your Google Ads search terms report: filter for long queries (8+ words) to see how people phrase questions
- Return reasons (sizing, fit and compatibility questions are gold)
Step 3: Write Q&As in a Google Sheet supplemental source
One row per product ID, with question_and_answer columns. Aim for 8–15 genuinely different questions per product (sizing, materials, care, compatibility, who it's for), not five rewrites of the same benefit. Keep answers factual. If it isn't true on the product page, don't put it in the feed.
Step 4: Map related products
Add often_bought_with and accessory links for every priority SKU. If you sell bundles, make sure the components point to each other.
Step 5: Add popularity rank and clean up variant families
Pull a sales-rank percentile from Shopify for popularity_rank. If your variants show up as 12 near-identical titles, add item_group_title and variant_option.
Step 6: QA, upload and keep it current
Check for formatting errors (commas and colons inside answers need escaping in Sheets), upload the supplemental source, and confirm there are no new issues in Merchant Center. Then set a review rhythm: new products, discontinued SKUs and new customer questions should flow into the sheet every month.
💡 TIP: Sales and promos do NOT go in the Q&A. Google's Q&A rules ban prices and dates. Run your offers through sale_price and Merchant Center promotions, where they belong. That also means your Q&A never goes stale when a sale ends.
How to Set Up Your Google Ads Campaigns for AI Mode
The good news: you don't need a new campaign type. The work is making sure your existing campaigns are eligible for the right kind of matching.
- Run Shopping or Performance Max on your key products. These are eligible for both direct and contextual matching.
- Check your Search keyword mix. If your account is 100% exact and phrase match, you're only eligible for the most obvious buying queries in AI Mode. Broad match or AI Max for Search opens up contextual matching.
- Keep your guardrails. Negative keywords and brand exclusions still apply in AI Mode.
- Fix your product pages, not just your feed. Gemini can use your website when it writes AI explainers, so PDPs that clearly answer sizing, compatibility and care questions help twice.
Our take: test structural changes like switching match types, turning on AI Max or trying AI Max for Shopping in a quiet period, not right before your biggest sale. The algorithm needs time to learn. Feed enrichment through a supplemental source is the low-risk move you can make any week, because it doesn't touch product approval.
How to Measure AI Mode Ads (When Google Won't Split It Out)
There is no AI Mode column in Google Ads (as of November 2026). Everything is blended into Search, Shopping and PMax reports. So you measure around it:
| What to track | Where | What it tells you |
|---|---|---|
| Long, question-style search terms | Google Ads search terms report | Whether you're matching conversational queries |
| AI share of voice + missing attributes | Merchant Center AI performance insights report (US, Canada, Australia, New Zealand, India) | How often your products show in AI Mode and AI Overviews vs. competitors |
| Enriched vs. non-enriched SKUs | Shopping / PMax product reports | Whether products with conversational attributes outperform the rest |
| Conversion rate on long-tail queries | Search terms + conversions | Whether AI Mode traffic is actually buying |
The cleanest test is simple: enrich your top 50 SKUs, leave a comparable set untouched, and compare impressions, CTR and conversion rate over 30 days. That gives you a real answer instead of a Google case study.
"Across our DTC accounts, median Meta link CTR is 1.57% and Google Ads CTR is 1.91%." — Top Growth Marketing
From where we sit, AI Mode is the first Google surface where product data quality decides visibility as much as bids do. The brands that treat their feed like a product page (answering the questions a great salesperson would) will get recommended. Everyone else will pay for clicks on a shrinking set of keyword-style searches.
The TGM Take
Most advice on AI Mode ads is about bids and match types. We think the feed decides more. Gemini shows about four products per shopping answer and picks them from the product data it can read, so a well-funded account with thin listings loses to a smaller brand whose feed answers the shopper's question.
Our order of work: enrich the top 50 SKUs through a supplemental source first, since it can't hurt product approval, and leave match-type or AI Max changes until after peak season.
— Jack Paxton, Founder, Top Growth Marketing
Google AI Mode Ads: What's Changed (Timeline)
We update this section as Google rolls out changes. Newest first:
- Oct 28, 2026: Ads in AI Mode expand to Australia, Canada, India, Kenya, Malaysia, New Zealand, Nigeria, Pakistan, the Philippines and Singapore. Multiturn ads launch in the US. Google plans the UK, Brazil and Mexico for Q4 2026. (PPC Land)
- Aug 31, 2026: Shopping ads move into the body of AI Mode answers: a sponsored carousel between the AI text and the follow-up questions, plus a second sponsored block at the bottom.
- May 20, 2026: At Google Marketing Live, Google announces Conversational Discovery ads, Highlighted Answers and AI-powered Shopping ads, and launches Merchant Center conversational attributes.
- Apr 30, 2026: AI Max for Shopping campaigns launches.
- Feb 11, 2026: Google introduces Shopping ads in AI Mode in the US.
💼 Want your feed and Shopping campaigns ready for AI Mode? TGM manages Google Shopping and Performance Max for eCommerce brands, from feed optimization to scaling. See how we do it →
AI Mode is already deciding which four products a shopper sees. Want to make sure yours is one of them? Book a call with our Google Ads team and we'll audit your feed and Shopping setup for AI Mode.
💡 Further reading: Watching AI shopping outside Google too? Our breakdown of ChatGPT Shopping covers the other side of the same shift. Running Google Ads on a smaller budget? Start with our guide to Google Ads for small business. And if you sell on Shopify, here's our list of the best Google Ads agencies for Shopify.
| ✅ Do | ❌ Don't |
|---|---|
| Add conversational attributes through a supplemental Google Sheet source | Rework your main product feed to squeeze the new fields in |
| Write 8–15 different questions per priority product | Pad the Q&A with five rewrites of the same benefit |
| Run offers through sale_price and Merchant Center promotions | Put prices or dates inside Q&A answers |
| Test match-type changes or AI Max during a quiet period | Switch match types right before your biggest sale |
| Compare 50 enriched SKUs against an untouched set for 30 days | Judge AI Mode on Google's case studies alone |
Frequently Asked Questions
Do exact and phrase match keywords show in AI Mode?
Only for direct query matching, where the question already carries clear buying intent. Broad match, AI Max for Search, Shopping and Performance Max qualify for contextual matching as well. Most AI Mode questions are long and exploratory, so an account built only on exact and phrase match reaches part of this inventory.
Can I write or edit the AI explainer Google adds to my ad?
No. Gemini builds the explainer from the shopper's query, the AI answer and public sources such as your website. You can't write, edit or approve it, so product pages that answer sizing, compatibility and care questions do double duty.
What should I leave out of Merchant Center Q&A attributes?
Google's rules ban prices, dates and keyword lists, plus anything that repeats your title, description, product highlights or product details. Put sales in sale_price and Merchant Center promotions instead. Your Q&A then never goes stale when a promo ends.
Do AI Mode ads get the same click-through rate as organic listings?
Google says sponsored and organic product listings in AI Mode see similar click-through rates. That claim rests on 15 days of internal US data from August 20 to September 3, 2026, and Google published no actual rates. Treat it as directional until you have your own numbers.
How do I measure AI Mode ad performance without a separate report?
Watch long, question-style queries in your search terms report and check the Merchant Center AI performance insights report where it's available. For a cleaner read, enrich your top 50 SKUs, leave a comparable set alone, and compare impressions, CTR and conversion rate over 30 days.
Last updated:
0 Comments