Aizii Learn · AI Search
← Back to Learn

Do product videos help with AI search

A product video helps with AI search when it answers a question a shopper actually asks, in words and in pictures, on a page a system can read. It does not help because it is a video. Google says that creating valuable, non-commodity content will likely influence a site's presence in generative AI search more than any other suggestion in its guide, and that high quality images and video are part of that content. It also says that no special markup is required for generative AI search, and that structured data is not required. A video that shows how a product works, how it fits, or how it is used is valuable content. A video that repeats the product name and the price is not.

No vendor publishes a formula for how much a video helps an AI recommendation, and no vendor promises that a video will get a brand cited. Treat a video as content, not as a lever, and judge it by whether it answers a question.

Where it goes wrong

Brands make videos that look good and say nothing, and then they wonder why the video changes nothing.

  • The video repeats the product page. A video that reads the product name and the price aloud adds nothing a shopper could not read in ten seconds, and an engine that reads the page will not learn anything new from it.
  • The video has no words a system can read. A video with no title, no description, no transcript, and no markup is a black box. A system that cannot read the words in the video cannot quote them, and a shopper who cannot hear the video cannot follow it.
  • The video is on a page where it cannot be watched. Google says VideoObject structured data must be added to a page where users can watch the video, and that it is a bad user experience to point users to a page where they cannot watch it. A video that is linked but not playable is a dead end.
  • The video is marked up, but the markup disagrees with the page. A VideoObject block that names a different video, or states a different duration, or points to a thumbnail that does not load, teaches a system not to trust the next block on the page.
  • The video is made for the algorithm, not for the shopper. A video that is cut to a length, or filled with keywords, or made only to rank, will be skipped by the shopper, and an engine that watches what shoppers do will learn to skip it too.

The steps

Step 1. Pick a question the video can answer. The marketer picks one question per video: how does this product fit, how is it used, what does it look like in use, or what goes wrong if it is used the wrong way? A video that answers one question well is worth more than a video that tries to answer five.

Step 2. Show the answer, and say it in words. The marketer writes the video so the answer is visible and spoken. Show the shoe on a wet path, show the coffee maker in a small kitchen, show the jacket in rain. Say the answer in the narration, and write the same answer in the description on the page, in plain words.

Step 3. Put the video on a watch page. The developer puts the video on a page whose main purpose is to show that video, with a title and a description that match what the video shows. Google says a video must sit on a watch page, a page whose main purpose is to show one video, and that each watch page needs its own title and description.

Step 4. Add VideoObject markup, with the three required properties. The developer adds a VideoObject block with name, thumbnailUrl, and uploadDate, which Google lists as required for video structured data eligibility. Add contentUrl or embedUrl, and a description, where the setup allows it. The name should be unique for each video, the thumbnail must be the video's own image, and the upload date should be in ISO 8601 format.

Step 5. Add key moments where they help. The developer adds Clip markup where a video has clear segments, such as fitting, use, and care, with a name, a start time, and a URL that deep links to that point. Google says the video must have the ability to deep link into some point other than the start, and that the total video duration must be a minimum of 30 seconds for key moments. Do not add key moments to a video that has no clear segments.

Step 6. Check that the video can be fetched and watched. The developer checks that the video file can be fetched, that the thumbnail loads, and that the video plays on the page. Google says it finds videos through the HTML tags video, embed, iframe, and object, and that a thumbnail must be accessible to Googlebot. A video that cannot be fetched cannot be shown, however good its markup is.

The machine-readable piece

This is the shape of a VideoObject block, using the schema.org vocabulary. Replace the placeholders with your own video facts, and only use values that match the visible page. The upload date is a placeholder shape, not a real date, and must be replaced with the real upload date of the video.

<script type="application/ld+json">
{
  "@context": "https://schema.org",
  "@type": "VideoObject",
  "name": "How the Example Trail Shoe Fits on a Wet Path",
  "description": "A short video showing how the Example Trail Shoe fits, how the sole grips a wet path, and how to lace it for a wide foot.",
  "thumbnailUrl": "https://[yourdomain.com]/images/trail-shoe-video-thumb.jpg",
  "uploadDate": "YYYY-MM-DD",
  "duration": "PT1M30S",
  "contentUrl": "https://[yourdomain.com]/videos/trail-shoe-fit.mp4",
  "embedUrl": "https://[yourdomain.com]/videos/trail-shoe-fit"
}
</script>

Check the block after you publish it. Open the page, view the source, and confirm the name, the thumbnail, and the upload date match the video on the page. Then run the page through the Rich Results Test and fix any error it reports. A video with a clean block is eligible for a video result, which Google says is not promised, but a video with a block that disagrees with the page will not be trusted.

What it looks like when it works

Say you run a skincare brand, selling a face cream, a cleanser, and a serum, with a guide on how to use them together. The marketer makes one video that answers a single question: how much serum do you use, and in what order? The video shows a hand applying the serum, says the amount in plain words, and the page under the video states the same amount in text.

The developer puts the video on its own watch page, adds a VideoObject block with the name, the thumbnail, and the upload date, and adds two Clip segments, one for the amount and one for the order. A shopper asks an assistant how to use the serum, and the assistant can find the page, read the text, and cite the brand with the steps as written. The video did not make the sale on its own. It answered the question, in words and in pictures, on a page a system could read.

The fundamentals

These mechanics are shared across every engine and covered in full in their own guides:

What this guide does not cover

This guide covers whether product videos help with AI search, and how to make a video a system can read. It does not cover how to shoot or edit a video, the full technical audit that checks every crawler and every page, or the transaction layer that lets an agent complete a purchase. It does not cover how to run a video platform, or how to measure video views in full. Those are separate jobs, and this guide is the map, not the whole trip.

Check your own store now

Run the free scan. One click, no signup, and it shows whether AI crawlers can read your store today: aizii.ai/aeo-scan

Run the free scan

Sources

Note: links marked specific point to the page that supports the claim in this guide. Links marked general docs point to the vendor's general docs section for that topic, not to a page that proves every step. Check the vendor docs for the current path before relying on a menu name or click path in this guide.