Important SEO Elements and Attributes: A Practical Audit Checklist
Published:
Last Updated:
Category: SEO & Content
Authors: Shusaku Yosa
An SEO audit covers several different things: document elements, link and image attributes, canonical signals, and indexing directives. Treating them all as interchangeable ranking factors leads to poor priorities. Group each setting by its purpose and compare CMS input with the actual page and response.
Group elements and attributes by what they control
Element, attribute, or setting | Purpose | Typical check |
|---|---|---|
title element | Identify the page’s subject | Unique wording matching the body |
description meta | Summarize content | No promises absent from the page |
H1–H3 elements | Express section hierarchy | Meaning rather than decoration |
Anchor href | Specify a destination | Correct language and live URL |
Image alt | Provide an appropriate alternative | Meaning matching the image’s purpose |
Canonical link | Signal the preferred duplicate URL | Correct equivalent destination |
Robots meta and headers | Control indexing or presentation | No unintended noindex |
Viewport | Support responsive presentation | Actual mobile readability |
Open Graph metadata | Describe shared previews | Current image and description |
Meta keywords | Keyword metadata | Unused in Google Web ranking |
Alt and href are attributes; title and H2 are elements. Name the actual setting when requesting a fix. Completing the table is not a score that guarantees rankings.
Audit titles and descriptions without character formulas
A title should identify the specific page, while a description summarizes its scope. Avoid fixed display-length formulas. Google’s title-link documentation explains generation from several sources and truncation according to available space.
Snippets can come from page content or the description, as explained in the snippet documentation. The exact CMS text is not guaranteed to display. A description is not a direct Google ranking factor; Google’s metadata explanation distinguishes this use from ranking.
Check headings, links, and image alternatives
Headings communicate sections; links help readers reach their next useful destination. Check meaningful wording, hierarchy, and live links rather than counting headings or strong elements. The related heading guide covers writing decisions in more detail.
Alt text should match an image’s function. Explain essential information in a meaningful image; do not burden readers with unnecessary descriptions of purely decorative images. Google uses image recognition and surrounding content as well as alt, according to its image guidance. Keyword stuffing is not a substitute for an appropriate alternative.
Separate canonical signals from robots rules
Canonicalization signals a preferred URL among duplicates; it is not an instruction Google must always follow. Robots.txt generally controls crawling, while noindex controls indexing. They are not substitutes.
Setting | Inspect | Avoid assuming |
|---|---|---|
Canonical | Correct equivalent URL | The duplicate URL disappears automatically |
Robots.txt | Intended crawl policy | Blocked URLs must vanish from Search |
Noindex | Intent and retrievability | Google can read it behind a crawl block |
Use the separate documentation for canonical consolidation, robots.txt, and noindex. Check HTTP headers as well as HTML.
Prioritize viewport, social metadata, and unused keywords
Viewport configuration is part of mobile presentation, not proof that text sizes, controls, or overflow work. Open Graph metadata supports shared previews; keep it consistent without presenting it as a direct Google ranking lever.
Google does not use meta keywords for Web rankings. An available CMS field is not necessarily a required optimization. Its AI features guidance also says no special AI tags or files are required. Baseline eligibility does not guarantee inclusion in an AI feature.
Compare CMS fields with actual HTML and response headers
Fictional audit | CMS input | Actual output | Search presentation |
|---|---|---|---|
Page A title | Specific pricing and eligibility | Old shared title remains | May differ again |
Page B indexing | Intended public page | Header retains noindex | Absence alone is not a diagnosis |
A published status does not establish that every field reached the response. Confirm the page, language, release, HTML, and headers. Keep input, output, and generated search presentation separate.
In these examples, investigate why the public page B retains noindex before refining wording. Page A also needs review, but confirm template scope and intended configuration before changing it. A stale response or version mismatch should not automatically be labeled a failed save.
Use a release checklist and verify the resulting page
Release checklist | Evidence |
|---|---|
Content and metadata | Topic, language, title, description agreement |
Canonical and directives | Intended URLs and HTML/header rules |
Navigation and images | Destinations, heading levels, appropriate alt |
Ownership | Version, reviewer, date, change scope |
After correction | Actual response, tests, subsequent processing |
Start by asking what each setting controls and for whom. Prioritize a match between publishing intent and the real page, rather than universal character counts or tag quotas.
