atlassian.com

On-page SEO is the portion of search-engine optimisation you directly control — the markup metadata and content structure that determines how Googlebot understands and ranks the page.

com the way Googlebot does: following the same HTTP sequence rendering JavaScript where needed and parsing the fully-resolved DOM rather than the raw source HTML that many auditors check and which misses dynamically injected metadata. We measure the element length because Google rewrites titles that fall outside the optimal range and a rewritten title is a title you've lost control of.

The meta description is absent or empty — Google will generate a snippet from body text which is almost always less persuasive than a crafted description. We count elements because a single H1 matching the primary keyword is a direct ranking signal. No canonical tag is present meaning Google will choose its own preferred URL from any variants it discovers — often not the one you would pick.

We measure the rendered word count because thin-content pages (under 300 words) are disproportionately likely to be classified as low-value by Google's Helpful Content systems. No structured data (JSON-LD or Microdata) was detected — structured data is required to qualify for rich results in Google search.

These signals together explain the bulk of Google's per-page quality assessment and they're all within the control of the site owner.

The element remains the highest-leverage on-page signal: it's the line Googlebot uses to determine relevance the text displayed in the SERP the label in the browser tab and the anchor text when a user shares the page.

Google rewrites titles for a large fraction of queries — this happens most often when the title is too long (over 60 characters truncating in the SERP) too short (under 20 characters giving Google insufficient topical signal) keyword-stuffed or duplicated across multiple pages. Staying between 50 and 60 characters with the primary keyword in the first 40 keeps you in control of what searchers see.

Meta descriptions don't directly affect rankings but they heavily influence the click-through rate on your SERP listing — and CTR is a ranking signal once Google has enough query-level data for the page. A well-crafted 120–160 character description that answers the searcher's intent and includes a clear reason to click consistently outperforms the auto-generated snippet Google constructs from body text.

Alt text on images serves a dual purpose: Google's image-understanding model uses it when visual recognition is ambiguous and screen readers depend on it for accessibility. Canonical tags prevent the parameterised www vs. apex HTTP vs. HTTPS and trailing-slash vs. com's canonical tag is set.

Word count below 300 is one of the strongest predictors of a thin content classification — not because length equals quality but because pages with very little text have very little for a language model to assess.

An A-grade on-page SEO audit in 2026 covers eight specific signals each independently verifiable and each improving a distinct ranking factor. A unique keyword-matched under 60 characters with the primary target keyword in the first 40 characters is the single non-negotiable starting point. ) rather than keyword density maintains control of the SERP snippet.

Exactly one per page matching the primary intent of the target query gives Googlebot a clear topical anchor. A self-referencing canonical tag set to the absolute HTTPS URL eliminates URL variant competition across the entire parameter space. Every content image with descriptive alt text (decorative images using empty alt= ) satisfies both the accessibility and image-search ranking requirements.

At least 600 words of original useful copy on the homepage clears the Helpful Content threshold on all but the most competitive queries. A JSON-LD block matching the page type — Organization or SoftwareApplication for company homepages Article for blog posts FAQPage for Q&A content Product for e-commerce — qualifies the page for rich results in Google and Bing.

Consistent og:title and og:description tags that mirror the on-page title and description tie the on-page and social layers together preventing the confusion that arises when a searcher shares a page and the preview looks like a different site.

On the open web the median homepage gets the basics wrong in remarkably consistent ways: titles drift past 60 characters so Google rewrites them for every searcher meta descriptions are either missing or duplicated site-wide across hundreds of pages structured data is absent so the domain qualifies for no rich results and homepage word counts sit under 300 words.

Pages that rank in the top cluster for competitive terms share a different measurable profile: a tight intent-matched title under 60 characters a click-earning description a single H1 aligned to the primary query at least 600 words of substantive content and at least one valid JSON-LD block that earns a rich result in the SERP.

The gap between these two groups is almost never about budget — it's about applying the same discipline consistently across every page template and catching regressions before they accumulate. 1 AA compliance.

Treat this report as a recurring check rather than a one-off audit: template changes CMS upgrades A/B tests and platform migrations all silently break on-page signals and the teams that catch regressions within a week consistently outperform the ones that discover them in their quarterly audit.