Visual SERP Indexation: ImageObject, VideoObject & Lazy-Loading Fixes
JavaScript intersection-observer lazy loaders frequently leave data-src empty during initial HTML fetches, stripping product visuals from image search.
Dmitry K., Technical SEO Lead · SEO Space
Indexation benchmark targets
- Native lazy attr
- src + loading="lazy"
- Hero LCP image
- fetchpriority="high"
- Video JSON-LD
- VideoObject + Clip
Preserving Crawlability on Lazy-Loaded Media
Always keep a valid high-resolution URL inside the standard src or srcset attribute — avoid empty placeholder data URIs coupled with data-src attributes. Above-the-fold hero visuals must never have loading="lazy", using fetchpriority="high" instead.
Visual search surfaces—Google Images, Video carousels, and visual Shopping packs—drive massive non-brand discovery when high-resolution assets are served directly via crawlable `<img srcset>` and descriptive filenames rather than CSS sprites or lazy JS blobs.
- Avoid `data-src` hacks; use native `loading="lazy"` strictly below the fold with a clean `src` fallback.
- Include contextual `alt` descriptions, surrounding captions, and structured `ImageObject` / `VideoObject` nodes.
- Whitelist `Googlebot-Image` and `Googlebot-Video` across CDN hotlink protection and WAF rules.
<img src="/media/audit-diagram.webp" alt="Crawl pipeline diagram" width="960" height="540" fetchpriority="high" /> Image & Video Sitemaps Plus Clean Format Negotiation
Serve modern AVIF/WebP formats cleanly via `<picture>` or CDN `Accept` headers while keeping canonical URL extensions consistent inside `sitemap-images.xml`. Avoid blocking `/media/` or CDN buckets in `robots.txt`.
| Visual Asset | Minimum Resolution | Key Discovery Signal |
|---|---|---|
| Hero / Editorial Image | >= 1200px wide (16:9 / 4:3) | max-image-preview:large + descriptive alt |
| Product Gallery Photo | >= 1000px square/portrait | Clean background + ImageObject in Product |
| Demo / Guide Video | >= 1280x720 poster | VideoObject + HTML transcript + watch URL |
Embedding Transcripts & Captions for Multimodal AI Retrieval
Search and AI models cannot cheaply parse spoken details inside raw MP4 streams unless backed by crisp VTT caption tracks and accessible HTML transcripts rendered right beneath the player.
<figure>
<img src="/media/crawl-budget-diagram.avif" alt="Server log status code breakdown showing 94% HTTP 200 share" width="1200" height="675" loading="lazy" />
<figcaption>Production Googlebot log distribution after facet parameter blocking.</figcaption>
</figure> Need this indexation setup audited on your domain?
Send your website URL. We check raw HTML parity, log crawl hits, canonical discipline, and sitemap architecture manually within 24 hours.