Product Studio is a set of generative AI tools from Google that creates and edits product images for Shopping listings and ads. It can place a product photo in a new scene, remove or replace a background and increase the resolution of a small image, all from inside Google’s merchant tools.
How Product Studio works
You start from an existing product image in Merchant Center or upload a new one. You then describe the scene you want in plain words, such as “on an oak kitchen worktop in soft morning light”, or pick a suggested style, and the tool produces several versions with your product placed in that setting. Other functions tidy an image: removing a cluttered background, replacing it with plain white, or upscaling a low-resolution photo so it is sharp enough to use.
At the time of writing (October 2026), Google has rolled Product Studio features out in stages, by country and language. What you can use depends on your account and region, so check inside your own Merchant Center rather than relying on an announcement or a screenshot from another country.
Generated images can be added to products as additional images, which then become available to Shopping ads and free listings.
Why it matters
Images shape whether someone clicks a Shopping ad, and a small retailer rarely has the budget for a lifestyle shoot of every product. Product Studio lets a candle maker in Bath or a homeware shop in Glasgow test seasonal scenes for a new range in an afternoon. It also gives Performance Max more image variety to build ads from, which matters because that campaign type assembles ads from whatever assets you supply.
The limit is honesty. The product shown must be the product you sell. Under UK consumer protection law and the CAP Code, an image that makes an item look larger, a different colour or better finished than it is can be misleading, whether a photographer or an AI produced it.
Common mistakes
- Replacing a clear main image with a busy generated scene. The main image should show exactly what arrives in the box; scenes work better as additional images.
- Accepting the first output without checking the details. Labels, logos, textures and proportions are where generated images drift.
- Generating settings that look wrong to British shoppers, such as American-style kitchens or sockets that are clearly not UK three-pin.
- Assuming no disclosure is needed. Google’s image rules ask that AI-generated product images keep the metadata that marks them as generated, so do not strip it out in another editing app.
- Upscaling a poor photo instead of retaking it, then wondering why it still looks soft on a phone.
How to act on it
Start with your best-selling products, where better images have the most effect, and keep their main images as accurate studio shots. Use Product Studio for additional images and seasonal scenes, check every output against the real item and keep a note of which images were generated. Then compare click-through rate on your Shopping ads before and after over several weeks, because a single week of data is too noisy to judge.
Images are part of how I run Performance Max campaigns for retailers, but better pictures will not rescue a broken product feed, so I check the feed first.