Google AI Mode & Shopping Ads: What Actually Changed
The DSA deadline slips to Feb 2027, but the real shift is Universal Cart keeping checkout inside Google. Why feed quality is now the Shopping control factor.
Welcome back to the front row.
This week, Google gave advertisers a gift. Most of them missed what it cost.
This Week
The DSA migration deadline moved. September 2026 just became February 2027. Google Ads Liaison Ginny Marvin confirmed the extension on LinkedIn - advertiser feedback, Q4 protection, more time to test. The announcement landed and the paid search community exhaled.
That exhale is the problem.
MAIN STORY
Your deadline moved. Your window didn't.
Google didn't slow down last week. It split the road.
On one side: Dynamic Search Ads get six more months. The automatic migration to AI Max, originally set for September, now begins February 2027. DSA campaign creation — which had already been restricted — returns on June 15. You have a runway. Use it.
On the other side: Google AI Mode crossed one billion monthly users. Four new Gemini-powered ad formats went live inside AI Mode responses. And somewhere between those two announcements, the relationship between your Google Ads account and your website changed in a way that no deadline extension will reverse.
Here's the number that doesn't move: 92 to 94 percent of AI Mode sessions end without a click to an external site. Intent is resolved inside Google before the user ever reaches your page. Not sometimes. Not in edge cases. The dominant behavior in AI Mode is already zero-click.
That's not a search update. That's a structural shift in what Google is.
The mechanism is called Universal Cart. Announced at Google I/O 2026, it's a checkout layer that works across Search, Gemini, YouTube, and Gmail simultaneously. A user finds your product, adds it, pays — without visiting your website. Your Merchant Centre feed is the inventory. Google Pay is the transaction layer. Your website becomes the fulfilment backend, not the discovery surface.
This isn't coming. Nike, Sephora, Target, Ulta Beauty, Walmart, and Shopify merchants including Fenty and Steve Madden are live now. The UCP checkout is already inside Shopping ads on YouTube and Direct Offers Demand Gen campaigns as of Google Marketing Live 2026.
The traffic model that has anchored digital marketing for two decades — Google sends users to you, you convert them on your property — is being restructured from the bottom up. Google is keeping the transaction. You're keeping the fulfilment.
Zoom out and the pattern is clear. AI Mode is the answer surface. The AI Brief gives you prompts-based control over how AI Max behaves inside that surface — messaging guidelines, matching guidelines, audience direction. Your feed quality determines whether your products exist in the AI layer at all. Your landing page, which AI Max now selects automatically via Final URL Expansion, is still relevant — but as a signal to the AI, not as a destination you control.
The keyword was the old unit of control. The feed is the new one. The prompt — your AI Brief input — is how you steer what you can't manually specify anymore.
If your Merchant Centre feed is incomplete, your products aren't invisible in AI Mode. They simply don't exist there.
That's where most accounts are right now. Running DSA campaigns they haven't touched in eighteen months, with product feeds built for the 2019 Shopping tab, waiting on the deadline that just moved.
The extension is real. Use it to test AI Max properly — enable features one at a time, monitor search term reports, set your AI Brief guidelines before they're mandatory. But don't use it to wait.
Because while the search migration runway just got longer, the commerce layer didn't extend anything. The storefront is already inside Google. The question is whether your products are on the shelves.
Audit your Merchant Centre feed this week. Not next quarter. Title completeness, description depth, GTIN accuracy, image quality. These are now distribution criteria, not listing criteria. Then enable AI Max on one campaign — the one with your strongest conversion history — and run it for thirty days with AI Brief messaging guidelines set from day one.
The deadline is February 2027. The window is now.
PROMPT OF THE WEEK
Use this to stress-test your product feed against AI Mode readability before your products compete inside it.
"You are a Gemini AI model answering a shopping query inside Google AI Mode. A user has asked: [insert your target query]. You have access to the following product data: [paste your product title, description, and key attributes]. Based only on this data, write the AI-generated product recommendation you would surface. Then identify every gap in the product data that prevented you from giving a more specific, confident answer."
What this gives you: a first-person view of how Gemini reads your feed — and a prioritized list of exactly what's missing before you're competing in AI Mode at scale.
THIS WEEK
AI Brief rolls out for AI Max Search campaigns. Prompt-based control over how AI Max targets and messages — three guidelines: what ads should say, what searches to match, and who to reach. English-language rollout first, Performance Max and Shopping to follow. If you're in a regulated industry and have avoided Final URL Expansion, this changes the calculation.
DSA creation restored June 15 — the operational implication. Google reversed the earlier restriction. You can now create and edit DSA campaigns through January 2027. This is useful for accounts that need to build migration test data. It is not a reason to delay the migration itself.
AI Mode crossed one billion monthly users. Google confirmed the figure at I/O 2026. Queries are more than doubling every quarter. The surface where your ads will increasingly serve is already at scale. The reporting to isolate AI Mode performance from standard search still doesn't exist. You're flying with one instrument.
TOOL OF THE WEEK
Google Merchant Centre — Feed Diagnostics Free. Inside your existing account.
Most Google Ads practitioners haven't opened their Merchant Centre feed diagnostics in months. That's the problem. The feed diagnostic panel now surfaces AI readiness signals — title length relative to query patterns, attribute completeness scores, and image quality flags — that directly affect whether your products surface in AI-powered Shopping placements.
How to use it right now: Go to Merchant Centre → Products → Diagnostics. Filter by "Item issues." Sort by affected items, not by issue type. Fix the top ten by volume this week. These aren't cosmetic errors anymore — they're exclusion criteria for the AI layer.
Why it matters this week specifically: Universal Cart and UCP checkout are live for major retailers. Feed quality is now distribution quality. The same data that gets your product into a standard Shopping ad is the data that determines whether you exist inside AI Mode commerce at all.
MY TAKE
Everyone celebrated the deadline extension like it was a concession. It wasn't. It was Google protecting Q4 revenue — theirs and yours — while continuing to build the infrastructure that makes the deadline almost irrelevant.
The search migration can wait until February. The feed can't.
AI Mode is already a commerce surface. The products on it right now belong to brands whose Merchant Centre data is clean, complete, and structured for AI retrieval. If yours isn't, you're not late to AI Max. You're absent from the new shelf entirely.
Six months is enough time to get this right. It is not enough time to keep waiting.
That's Issue #9 of NVM. If this gave you one thing to act on this week, share it with someone who needs it. Reply to this if you have a question — I read every one. See you next week. — Mostafa
NVM · Front row seats to AI and digital marketing
- Subscribe → nvm.marketing/newsletter
- Join the community → nvm.marketing/community
- Write for NVM → nvm.marketing/blog
Get the next issue first
Weekly intelligence on AI, search, and performance marketing, straight to your inbox. No noise.
Prefer to discuss on LinkedIn? Join the conversation on the original post.
Join the discussion on LinkedInComments
Sign in to join the conversation
Sign in