Answer in brief
Give each important video its own watch page, where the video is the main content, and describe it with VideoObject structured data; Google requires a name, a thumbnail address and an upload date. Keep the thumbnail and the video file at stable addresses and publish a transcript under the player. Mark key moments, for which Google asks for a video of at least 30 seconds, then open the video indexing report in Search Console to see whether the video was indexed and, if not, why.
Search engines index pages, and a video rides on one
A video file contains no words a search engine can read on its own. What gets indexed is a page, and the video travels with it. Google’s documentation sets the conditions out as a list: the watch page must be indexed, the video must be embedded on that page and not hidden behind other elements, and it must have a valid thumbnail available at a stable address.
Two further lines in the same list are easy to miss. The indexed page must be performing well in Search before its video is considered for indexing, and an indexed page does not mean an indexed video. In other words, video indexing in Google depends on the quality of the page built around the file.
Google names four places where a video can then appear: the main results page, Video mode, Google Images and Discover. The rest of this article follows the order of work: the page, the markup, the transcript, the poster, the chapters and the measurement. It applies whether the file sits on your own server or on a video platform.
A watch page is a page built around a single video
Google defines a watch page as one whose main purpose is to show users a single video. Its examples include a video landing page, a sports highlight page and a news video page. Its counter-examples are just as useful: a blog post that reviews an embedded video, a product page with a 360-degree video, and a category page listing several videos of equal prominence.
In practice, each video you want found needs an address of its own: the service explainer, the case film, the tutorial. Put the player at the top, at its real size, and let it load without a click. Search Console lists “Cannot determine video position and size” among the reasons a video is not indexed, typically when an image has to be clicked before the player appears. Google also says not to rely on user actions such as swiping, clicking or typing to load a video.
Give the page a title and description that are unique to that video, as Google asks. The same film may still be embedded in an article or on a product page; Google says that is fine and that such a page can appear as an ordinary text result. Remember one limit from the Search Console help: Google indexes only one video per page, so a gallery of ten films surfaces one at most.
VideoObject markup: three required properties and one rule
Structured data is a description of the video written for machines. Google says it tries to understand a video automatically, but that VideoObject markup lets you influence the description, thumbnail, upload date and duration shown in results, and can make the video easier to find. Three properties are required: name, thumbnailUrl and uploadDate. Description, duration, contentUrl and embedUrl are recommended.
The two address properties are often confused. The first, contentUrl, points to the bytes of the video file itself, and Google calls it the most effective way for it to fetch the file. The second, embedUrl, points to the player. Neither should be the address of the page. Dates and durations are written in ISO 8601 format; Google’s own example, PT00H30M5S, stands for thirty minutes and five seconds.
The rule is consistency. Google asks that the markup match the actual video and that the name, description and thumbnail address be unique for every video on the site. The Search Console help adds that the title, thumbnail address and video address should be the same in every source that describes the video: sitemap, HTML tags, meta tags and structured data. Markup makes a video eligible; Google states that it does not guarantee any particular feature in results.
The transcript belongs on the page, under the player
The W3C’s Web Accessibility Initiative defines a basic transcript as a text version of the speech and non-speech audio information needed to understand the content. A descriptive transcript adds the visual information as well. According to the W3C, transcripts are used by people who are Deaf or hard of hearing, by people who have difficulty processing auditory information, and by others; in everyday terms that includes anyone who cannot turn the sound on.
The same guidance says where to put it: for media on your own website, usually on the same page, with the transcript or a link to it right under the video. It also says how to make it useful. Publish it as ordinary HTML, group the text into logical paragraphs and sections, identify the speakers, add any text that appears on screen but is not spoken, and use timestamps only where they help. The W3C notes that adding headings and links also helps with SEO.
Be clear about what a transcript does not do. Google’s video documentation does not list it among the indexing requirements, and its structured-data page does not include it among the required or recommended properties. A transcript earns its place by serving readers and by putting the film’s words on the page. There is one tension to watch: Search Console’s help describes a watch page as one whose content is mainly a single video. Keep the player first, and after publishing check how Search Console’s video indexing report classifies the page. If it no longer counts as a watch page, use the W3C’s other option: a link to the transcript directly under the video, with the full text on a page of its own.
One poster, one address: why the thumbnail has to stay put
The poster is the still image shown before playback; Google’s documentation calls it the thumbnail and will not show a video in its video features without a valid one. The same page carries a warning that explains many silent failures: some content delivery networks use quickly expiring addresses, and if the thumbnail’s address changes too often Google may not be able to index the video. The instruction is to use a single, unique and stable thumbnail address for each video.
The technical requirements are short. Google accepts BMP, GIF, JPEG, PNG, WebP, SVG and AVIF, asks for at least 60 by 30 pixels and prefers larger. The file must be reachable by Googlebot, so it cannot sit behind a login or a robots.txt block, and at least 80% of its pixels must have an alpha value above 250, which rules out transparent images. The poster can be supplied through the poster attribute of the video element, a video sitemap, the thumbnailUrl property or an Open Graph tag, and Google asks for the same address in all of them.
In practice, store each poster at a permanent path and exclude it from any build step that renames files on every release. Do not share one poster between films: the Search Console help warns against using the same thumbnail for different videos, because Google may then treat them as one and mix their details. If you want previews and key moments, Google asks that the video file itself be available at a stable address too.
Key moments are chapters you can mark yourself
Key moments let a searcher jump to a part of a video, in Google’s words like chapters in a book. Google tries to detect the segments automatically, but says it will give priority to the ones you set. For a video on your own page there are two methods. Clip markup states the start and end time and a label for each segment. SeekToAction markup tells Google where timestamps go in your address structure so that it can identify the moments itself.
Google attaches conditions. The video address must be able to open at a point other than the start, for example with a time parameter; the total duration must be at least 30 seconds; and no two clips on the same video may share a start time. For SeekToAction, Google must also be able to fetch the video file. Clip markup is supported in every language in which Google Search is available, while SeekToAction is limited to a published list of languages, so check that list before choosing.
Write the labels for the viewer. “Pricing explained” or “What we need from you” tells a searcher where to jump; “Part 2” tells them nothing. The same labels make good headings inside the transcript, so one piece of editorial work serves both. For a film hosted on YouTube, Google reads timestamps and labels placed in the video description.
Embedding from a platform without giving the visit away
Many businesses host on YouTube or Vimeo and embed the player on their site. Google says that in this case it may index the video both on your page and on the platform’s own page, and that both versions may appear in video features. Your page therefore competes with the platform’s page for the same film, and a visitor who lands on the platform is surrounded by other people’s videos.
Google still recommends structured data on your own watch page when the player comes from a third party, and allows such pages in a video sitemap. For further features it advises checking that the host lets Google fetch the video file. Previews and key moments depend on that access.
Self-hosting gives full control over file and poster addresses, but Google notes that both the site hosting the page and the server streaming the file need enough capacity to be crawled. A platform removes that burden and adds its own audience. Decide film by film where you want the viewer to end up. When the next step is an enquiry, the watch page on your own site is the one worth building properly.
Three Search Console reports show what actually happened
Google points to three reports. The video indexing report shows how many indexed watch pages contain an indexed video and why other videos were not indexed. The video rich result report lists problems in the VideoObject markup. The performance report, filtered by the Videos search appearance, shows impressions and clicks for results displayed as videos.
In the indexing report every page has one of two statuses, “Video indexed” or “No video indexed”, and only indexed pages are included. The second status comes with a reason: “Video isn’t on a watch page”, “No thumbnail URL provided”, “Thumbnail could not be reached” and others. For a single address, the URL Inspection tool adds a video indexing section whenever a video was found on the page. This is where video indexing in Google stops being guesswork.
Fix the cause, then ask for validation. The help page says validation typically takes up to about two weeks and sometimes much longer, so plan in weeks. Judge the outcome by search visits to the watch page and by what viewers do next, not by the number of films uploaded. Nothing guarantees a video result. What this work secures is that every important video is eligible and every failure has a named reason.
Practical checklist
- Create a dedicated page for each video you want found, with the player at the top, at full size and loading without a click.
- Add VideoObject markup with a unique name, description, thumbnail address and upload date, and test it before publishing.
- Publish the transcript under the player as ordinary text, with short paragraphs, speaker names and a few headings.
- Move every video poster to a permanent address and use that same address in the markup, the sitemap and the page.
- Open the video indexing report a few weeks after publishing and work through each listed reason until none remains.
Questions and answers
Why does Google index my page but not the video on it?
Because the two are judged separately. Google states that an indexed watch page does not mean the video will be indexed as well. The usual causes are a page that is not built around one video, a thumbnail that is missing, unreachable or changing address, and a player that is not present when the page loads. The video indexing report in Search Console names the reason for each address.
Do I need a transcript if the video already has captions?
They do different jobs. Captions are written to be read while watching; the W3C says a transcript should also carry the important visual information for people who are not seeing the video. By the W3C’s own summary of the WCAG standard, captions for prerecorded video with audio are Level A and transcripts are Level AAA. Google does not require a transcript for indexing, but it gives every visitor a readable version of the film.
Is it better to host a video myself or embed it from YouTube?
Either can be indexed. With an embed, Google may index both your page and the platform’s page, so the two can compete. Hosting the file yourself keeps its address and the poster’s under your control at the price of server capacity; a platform is simpler and brings its own viewers. Choose by where you want the visitor to land, and in both cases build a proper watch page with markup.
Can several videos on one page all appear in search?
The page itself can rank, but only one of its videos will be indexed; the Search Console help is explicit on that point. A gallery in which several films carry equal weight is not a watch page in Google’s terms either. Use it for navigation and give each film an address of its own if you want it to be found separately.
How long does it take for a video to show up in search results?
No fixed period is published. The page has to be crawled and indexed first, and then the video is processed; the report lists a reason, “Video not processed yet”, with the advice to check back in a few days. After a correction, the same help page puts the usual validation time at around two weeks, sometimes far more. No step guarantees that a video feature will be shown.