ubuntu.com

Title, description, canonical, social tags — what search engines read first.

Everything we check in this category passed.

Headings, copy depth, images, structured data, internal linking.

• Single H1 headinghigh impactlow effort

  2 h1 headings found: "A CTO's guide to real-time Linux", "Ubuntu 26.04 LTS Resolute Raccoon is available for download"

  Multiple h1s split the page's main-topic signal, leaving search engines to guess which headline actually defines the page. Keep the one h1 that best states the primary topic and demote the rest to h2 or h3, so the outline reads as a single subject with subtopics. A frequent culprit is a site logo or blog title wrapped in an h1 inside the shared header template — change that to a <div> or <p> and reserve h1 for the page's own content.

• Image alt textmedium impactlow effort

  50 of 63 images have alt text

  Alt text is how search engines understand what an image shows — it feeds image-search rankings and is what screen readers announce to blind visitors. Add a short, specific alt attribute to every meaningful image, describing the content rather than listing keywords: alt="Golden retriever puppy playing in snow" beats alt="dog puppy pet animal". Give purely decorative images an empty alt="" instead of omitting the attribute, so assistive technology knows to skip them rather than reading out the filename.

• Heading hierarchylow impactlow effort

  Heading level jumps from h1 ("Ubuntu 26.04 LTS Resolute Raccoon is available for download") to h4 ("私たちの日本のウェブサイトを試してみてください&nbsp;&rsaquo;")

  Headings form the page's outline; when levels jump (h2 straight to h4), search engines and screen readers get a broken table of contents and the relationship between sections turns ambiguous. Fix the jump by stepping down one level at a time — promote the deeper heading, or add the missing intermediate level. The usual cause is choosing heading tags for their default font size; set sizes in CSS instead and let the tags reflect actual document structure.

HTTPS, redirects, robots, sitemap, speed, crawlability.

• Fast time to first bytehigh impactmedium effort

  817 ms

  Time to first byte is the floor under every other speed metric — the browser can't parse, render, or fetch anything until the first byte arrives, so a slow TTFB drags down all Core Web Vitals and tests crawlers' patience. Add caching in front of the origin: full-page caching at a CDN (Cloudflare, Fastly) helps most, followed by server-side caches (Redis, object caching) and indexes on slow database queries. Measure from locations near your actual users — tuning only your fastest region leaves distant visitors just as slow.

• robots.txt present and permissivemedium impactlow effort

  no robots.txt found

  Without a robots.txt, crawlers assume everything is allowed — not fatal, but you lose the ability to keep them out of low-value areas (internal search results, carts, admin paths) and the standard place to advertise your sitemap. Create a plain-text robots.txt at the site root with at least a "User-agent: *" line and "Sitemap: https://www.example.com/sitemap.xml". Don't use it to hide sensitive URLs, though — the file is public, and disallowing a page doesn't remove it from the index; use noindex or authentication for that.

• Response compression enabledlow impactlow effort

  no content-encoding header

  This HTML is served uncompressed, so every visitor downloads far more bytes than necessary — text compresses extremely well, and the savings directly speed up first render on slow connections. Enable Brotli or gzip on your server or CDN: in nginx it's "gzip on;" (or the brotli module), in Apache it's mod_deflate via .htaccess, and on most CDNs it's a single toggle. Make sure compression covers all text types (HTML, CSS, JS, JSON, SVG), but skip already-compressed images — recompressing them wastes CPU for no gain.