Why Is Google Rewriting My Shopping Ads? Here Is What the AI Description Test Means for Your Campaigns

Google is testing AI-generated descriptions inside Shopping ads. Advertisers who have spent years perfecting their product feed copy now have to think about what that means for their click-through rates and how Google is deciding what to say about their products.

On July 30, 2026, PPC analyst Brodie Clark spotted something new in Google Shopping results: product ads with AI-written descriptions appearing beneath the standard title and price. According to Search Engine Roundtable and Search Engine Land, this is an extension of a test Google had already confirmed earlier in July for standard text search ads. The statement Google gave at the time was brief: “This is a small experiment to see if adding AI-generated context to Search ads helps people make more informed decisions.”

That quote is doing a lot of work. It sounds user-friendly. But for anyone running Google Shopping campaigns, it raises real questions. If Google is generating its own copy for your product ads, what is it pulling from? Can you influence it? And what happens to your CTR if the description Google writes is generic, off-tone, or flat-out wrong?

What Is Google’s AI Description Test?

The test adds an AI-generated text snippet underneath Shopping and Product ad listings in Search results. Think of it like a mini blurb sitting between the product title and the usual price or rating display. Google is generating this text algorithmically, not pulling it directly from your product description field in Merchant Center.

Earlier in July, a similar test appeared on standard sponsored search results, where an AI-generated summary appeared below the headline and display URL. That test reportedly upset advertisers because the AI text sometimes felt redundant, incomplete, or misrepresented the product being advertised. Now the same experiment is moving into Shopping format, which carries higher commercial intent traffic.

Google has not confirmed a wider rollout or given advertisers any opt-out mechanism. It has framed this as a limited test.

Why Does This Matter for Ecommerce Brands?

Shopping advertisers invest real time in their product feed. Titles are structured to front-load brand, product type, and key attributes. Descriptions are written to handle objections, highlight differentiators, and move intent buyers toward a click. Images are tested. Prices are competitive. The entire listing is tuned.

AI-generated descriptions introduce a layer outside that control. The concern is not theoretical. If Google writes “A durable, versatile jacket for everyday use” over a product listing for a technical mountaineering shell that retails at $380, the AI copy is doing no favors. It is competing with your feed work rather than complementing it, and it could pull a searcher’s attention toward a vague summary instead of the specific attributes they were scanning for.

There is also a brand tone issue. Advertisers who have built a consistent voice across their ads, landing pages, and emails may find that Google’s AI-generated text sounds nothing like them. For premium or positioning-sensitive brands, this is not a small problem.

CTR is the core metric to watch. If AI descriptions reduce the information density of a Shopping listing or soften the urgency and specificity that earns clicks, conversion volume follows CTR down. Budget does not adjust automatically.

What Data Is Google Pulling From?

Google has not published documentation on exactly how it generates Shopping ad descriptions. But based on the signals Google uses across its product ecosystem, the most likely inputs are:

Your Merchant Center Product Data

Product title, description, product type, and custom labels are the most obvious sources. If your descriptions in Merchant Center are thin, vague, or keyword-stuffed without actual information, the AI has less good material to draw from and may produce output that is correspondingly vague.

Your Landing Page Content

Google crawls destination URLs. If your product page has strong, specific copy covering materials, sizing, use cases, or technical specs, that content can inform what the AI says. A product page that buries the details in a tab that Google cannot easily parse gives the AI less to work with.

Third-Party Product Reviews and Aggregate Signals

Google has access to review data across Shopping surfaces. For well-reviewed products, it is possible that sentiment or attribute summaries drawn from review data shape the AI description. This is not confirmed, but Google already uses this type of signal in AI Overviews for informational queries.

How to Optimize Your Feed for AI Descriptions

You cannot yet opt out of this test. But you can make the inputs better, which gives the AI more useful raw material and reduces the chance it produces something generic or off-brand.

Write Specific, Attribute-Rich Product Descriptions

If your current Merchant Center descriptions are short or keyword-padded, rewrite them. Lead with the product’s primary use case, its most important technical attribute, and one or two reasons a buyer would choose this over a cheaper alternative. Think two to four sentences. The AI is more likely to produce a useful output when the description it is working from already answers “why this product” clearly.

Front-Load Your Product Titles Properly

Your title is likely the highest-weighted input for Shopping ranking and for any AI-generated copy. Keep the Google-recommended structure: Brand, Product Type, Key Attribute (size, color, material, compatibility). A well-structured title gives the AI a skeleton to build around even if other inputs are thin.

Audit Your Landing Page Copy

Make sure the product page your Shopping ad links to has clearly structured, crawlable copy. Avoid JavaScript-rendered descriptions that Googlebot may not execute. Structured product schema markup (using schema.org/Product) helps Google understand product attributes precisely, which may influence what AI-generated descriptions it surfaces.

Use Custom Labels to Signal Priority Products

If you run a large catalog, not every product merits the same feed investment. Use custom labels to segment your highest-margin or highest-conversion SKUs and give those product listings the most complete feed data. That is where AI description quality matters most.

Common Mistakes to Avoid

Waiting for Google to make it official

This test may or may not roll out broadly. But feed quality improvements have benefits beyond this specific test. They affect Shopping ranking, Quality Score, and the information available for AI Overviews to reference. Waiting for a formal announcement before improving feed data is leaving upside on the table now.

Assuming AI descriptions only affect impression share

The risk is at the CTR level. A Shopping ad with a weak or inaccurate AI description may actually reduce click-through from buyers who had high intent, because the AI copy either confuses them or fails to confirm the product match they were scanning for. Do not assume this is neutral. Set up a CTR alert in your Shopping campaigns for any unusual drops starting now.

Treating all SKUs equally

For a large catalog, trying to rewrite every product description is not realistic. Segment by revenue contribution or margin, and start at the top. Twenty to thirty percent of your SKUs likely drive eighty percent of your Shopping revenue. That is where feed investment has the highest return.

What This Means Going Forward

Google’s pattern over the last 18 months has been consistent: introduce a test on a limited basis, confirm when asked, wait for advertiser feedback, and then either scale it or quietly drop it. AI summaries in text search ads followed this path starting mid-July. Shopping is next.

The broader theme is that Google is slowly inserting its own AI-generated layer between advertisers and the users clicking on their ads. This started with automated headlines in Responsive Search Ads. It continued with AI Max’s query expansion and automated ad copy generation. AI-generated Shopping descriptions are another step in the same direction.

Agencies and in-house teams that built their value around writing great ad copy now have a strong incentive to shift that skill set toward feed management, landing page structure, and product data quality. That is where the leverage is moving. The brands that win in this environment will be the ones whose product data is so thorough and well-structured that even Google’s AI cannot easily generate something vague about them.

At Incisive Growth, we have been watching this shift across both text search and Shopping formats, and our recommendation to clients is not to panic, but to treat it as a prompt to audit product feeds now rather than after a wider rollout forces the issue.

If you want a concrete starting point, our Google Ads management service includes a full Merchant Center and product feed audit as part of onboarding. Getting your product data in the best possible shape before this test scales is the highest-ROI action you can take today.

Frequently Asked Questions