Every time a user clicks a video result on Google, the traffic and citation often land on youtube.com, not your domain. This hidden cost accumulates silently over time. Key moments markup changes this dynamic by allowing your website to claim the specific timestamps that Google surfaces.
Instead of sending the viewer to a YouTube link with a chapter marker, this schema mechanism directs them to your own video page. The player starts at the relevant moment, keeping the user journey—and the resulting data—on your infrastructure. For brands tracking video SEO performance, this distinction matters because it reassigns the URL from a third-party platform to your property.
The following sections break down how this works. We start with the foundational video schema requirements and move to the specific SeekToAction property that makes the domain switch possible.
VideoObject schema: the prerequisite for Google video answers
Key moments markup is not a standalone toggle you can flip on in your CMS. It is a capability that activates only when the underlying video schema is correctly implemented. Think of the VideoObject as the foundation; without it, there is no structure for Google to recognize, index, or enhance.
For a video to become eligible for rich results, the metadata must be complete and valid. Google requires a specific set of properties to understand what the video is, where it is hosted, and how long it runs. If any of these core fields are missing or malformed, the video remains invisible to the Key Moments feature, regardless of how well-structured your chapter list is.
The essential video metadata properties include:
- name: The title of the video.
- description: A concise 150-300 character summary that contextually describes the content.
- thumbnailUrl: A link to an image accessible by Google’s crawler.
- uploadDate: The date the video was published, in ISO 8601 format.
- embedUrl: The URL to the embedded player. For YouTube, this typically follows the format
https://www.youtube.com/embed/VIDEO_ID. - duration: The length of the video, in ISO 8601 format (e.g.,
PT15M30Sfor 15 minutes and 30 seconds).
A valid VideoObject block
To anchor the technical explanation, here is a basic, valid JSON-LD structure. This is the minimum viable unit before you can consider adding actions for timestamped navigation.
{
"@context": "https://schema.org",
"@type": "VideoObject",
"name": "Product Tutorial: Advanced Settings",
"description": "A step-by-step guide on configuring the advanced settings of our platform, covering automation and data export features.",
"thumbnailUrl": "https://example.com/thumbs/tutorial.jpg",
"uploadDate": "2023-10-15",
"duration": "PT15M30S",
"embedUrl": "https://www.youtube.com/embed/dQw4w9WgXcQ"
}
If this block does not pass validation, subsequent layers of optimization, including key moments markup, will have no effect. The video will not appear in video carousels, and specific chapter links will not be generated. Accuracy here is the first gate to visibility.
SeekToAction: shifting the URL from YouTube to your domain
Standard video search results often direct users to a YouTube link, even when the content is hosted on your site. The SeekToAction property within your video schema changes this behavior. It instructs Google to surface your domain in search results, offering a direct link to a specific moment in your video rather than a generic chapter marker on a third-party platform.
The distinction matters for traffic quality. Without this action, a user clicking a result lands on youtube.com, where they are exposed to unrelated suggestions and ads. With SeekToAction, the user lands on your watch page, keeping the engagement within your ecosystem. This is a critical shift for any strategy focused on video SEO and long-term brand ownership.
The technical implementation
To enable this behavior, you must add a potentialAction block to your VideoObject schema. This block defines how the browser should respond when a user requests to jump to a specific second of the video. The key element is the target field, which uses a template to dynamically insert the timestamp.
Below is the structure you need to integrate into your JSON-LD. Note the {seek_to_second_number} parameter, which tells the browser to use the provided second value. The startOffset-input directive ensures the video starts playing at that exact point.
{
"@type": "VideoObject",
"name": "Video Title",
"embedUrl": "https://www.yourdomain.com/video-page",
"potentialAction": {
"@type": "SeekToAction",
"target": {
"@type": "EntryPoint",
"urlTemplate": "https://www.yourdomain.com/video-page?t={seek_to_second_number}",
"actionPlatform": ["web"]
}
}
}
Why this drives domain traffic
This configuration does more than just change the URL. It enables chapter-level navigation while retaining user attention on your property. When Google generates rich results for Google video answers, it uses these timestamps to create direct deep links. This means a user searching for a specific topic can jump straight to the relevant section of your video, bypassing the intro entirely.
This precision increases the likelihood of a user staying on your page, especially if the video is the primary content. By controlling the destination URL, you ensure that every click from search engines contributes to your domain’s metrics, not YouTube’s. It is a small code change with a significant impact on how video metadata is interpreted and served in search results.
Video chapter formatting: the missing link in schema markup
Adding SeekToAction to your JSON-LD does not magically create navigable chapters. The video file itself must contain a properly formatted chapter list on the hosting platform. Without these on-video timestamps, the schema has nothing to map to, and the seek_to_second_number parameter remains a dead link.
How the mapping works
When a user clicks a chapter in Google search results, the system looks for the specific timestamp defined in your video’s metadata. The startOffset-input directive in your schema tells Google to replace the variable with the exact second the user clicked. If the video’s actual chapter list does not align with these targets, the jump fails or lands on the wrong second. This synchronization is critical for the user experience; a misaligned jump breaks the promise of the rich result.
The two-schema conflict
A common mistake is applying standalone video schema to pages where the video is merely supporting content. If you embed a video in a blog post, do not add a separate VideoObject block. Instead, nest the video metadata as a property within the existing BlogPosting or Article schema. Conflicting signals confuse the index, potentially hurting both the article’s ranking and the video’s eligibility for rich results. The index needs a clear hierarchy: the text is the main content, the video is supplementary.
The core decision framework
Use SeekToAction and standalone VideoObject only when the video is the primary content of the page. This applies to watch pages or video landing pages where the user’s main intent is to view the video. In these cases, the schema accurately reflects the page’s purpose. For any other context, keep the schema nested and subordinate to the text. This distinction ensures that your video SEO efforts reinforce your page’s topical relevance rather than diluting it with contradictory structural data.
FAQ: key moments markup and video SEO essentials
Q: Does key moments markup require the video to be hosted on my own website?
No. The schema works with YouTube embeds, provided that the contentUrl and embedUrl properties point to a valid, accessible video file.
Q: How does Google use seek_to_second_number in video answers?
The search engine uses this parameter to generate deep links to specific timestamps. This allows users to start playing the video at a relevant chapter rather than from the beginning, which is a core function of Google video answers.
Q: Is video schema the same as key moments markup?
Not quite. Video schema (VideoObject) is the structural container. Key moments markup (SeekToAction) is a specific action within that container that enables timestamped navigation.
Q: Will adding this markup improve my video SEO on YouTube?
Not directly. The markup improves how the video appears in Google Search and rich results, potentially driving more external traffic to your domain, but it does not alter YouTube’s internal ranking algorithms.
As AI-generated answers reshape search, owning the user’s attention on your own domain becomes a strategic priority. Key moments markup is a low-effort adjustment that yields high-impact visibility in Google video answers. The shift from platform dependence to domain ownership in video search results is underway; how you structure your video metadata today will determine where users land tomorrow.
