Identify whether the defect is in the content, published metadata or observed search result; correct the responsible layer and assess later search evidence separately.
You can publish a good page title and meta description without Google displaying those exact words. The website’s declared metadata, its visible heading and the search result are different layers. Diagnose the layer that is wrong before changing plugins or rewriting every page.
The practical goal is an accurate invitation to the right content: a reader should understand the page’s task and find what the result promised. This guide explains how to write that metadata, verify what the website serves and assess search appearance without treating a preview as a guarantee.
Identify the item you are trying to change
| Item | Where it belongs | What to inspect |
|---|---|---|
| Document title | The HTML title element | The delivered page head and browser tab |
| Main page heading | Visible page content | The actual article or landing-page heading |
| Meta description | A description entry in the HTML head | The delivered metadata, not only the editor field |
| Google title link and snippet | A particular search result | The result URL, query, date, device and language |
| Site name | Search attribution for the website | Site identity, distinct from one page’s title |
MDN documents the title element as the document title used in the browser’s title bar or tab. It is not a visible article heading. Its metadata reference separately defines the description name. A field labelled “SEO title” in an editor is an input to your publishing system; inspect its output before drawing conclusions.
Write a title that makes the page distinguishable
Start with the reader’s actual task and the page’s distinctive scope. A fictional article called “Email Guide” is difficult to distinguish from dozens of others. “Set up email on a new domain without losing incoming messages” communicates a more specific purpose, provided the article actually explains that process.
Google’s title-link guidance recommends descriptive, concise titles and identifies several sources it may use, including the document title, prominent headings and links. Avoid repeated keyword lists, obsolete years and a long branding slogan appended to every page. A mismatch between the title and the content is worth correcting even if you cannot control the final wording Google selects.
Use a restrained brand suffix if it helps identify the publisher. Read several titles together: can someone distinguish an introductory explanation, a troubleshooting guide and a comparison? Do not add “best,” “verified” or a current year unless the content supports those claims and has a maintenance plan.
Describe the answer, its scope and a useful limit
A meta description should tell a reader what this page helps them do. For the fictional email guide, an editorial example is: “Plan the mailbox move, preserve existing messages, check DNS and test incoming mail before changing the old service.” That summary describes actions instead of promising effortless migration or guaranteed delivery.
Google’s snippet documentation explains that snippets are primarily drawn from page content and can vary with the search. A meta description is used when it is a useful representation. There is no universal character length that guarantees full display: search results truncate to suit the presentation.
Put the important meaning early, remove repetition and write naturally. Use the page’s actual audience or constraint when it matters: a personal mailbox differs from a business-wide migration. An editor’s length indicator can help with composition, but a green indicator is not evidence of indexing, accuracy or a displayed result.
Make the visible page fulfil the same promise
Compare the title and description with the opening answer, main heading and section headings. If a title promises recovery steps but the page contains a sales introduction and no recovery procedure, changing metadata alone will not solve the reader’s problem.
The SEO Starter Guide stresses useful, organised, current content and explains that improvements do not guarantee ranking. Use that principle as an editorial check: can the reader reach a supported answer, understand limits and find the next relevant task?
Remove stale offers or dates from both visible copy and metadata. If a translated page exists, review its own title and summary rather than leaving another language’s text in place. For a broader content review, use the refresh and consolidation workflow.
Check who generates the published metadata
In WordPress, a theme, core behaviour, an SEO integration or custom code may contribute to the result. The document-title reference shows how WordPress assembles and filters a title. A per-page setting can also interact with a template that adds the site name or archive context.
Record the intended editor field and save through the site’s normal workflow. Then open the public URL as a logged-out visitor and inspect its head. Look for the expected title and description, accidental empty values and competing output. If the editor is correct but the delivered page is stale, investigate the responsible publishing/cache layer rather than repeatedly changing the wording.
Do not install a second SEO plugin to overwrite an unexplained first one. A maintainer should resolve competing ownership with a reversible, scoped change. Social sharing previews are another system; the Open Graph guide covers those separately.
Keep the site name separate from the page title
If the wrong text appears beside the favicon or as the source name, you may be looking at attribution rather than the article’s title link. Google’s site-name documentation describes an automated system using homepage signals and supported WebSite structured data to understand the preferred identity.
Keep your public brand consistent in the homepage and relevant settings. Do not rewrite every article heading to repair a site-name issue. Equally, a correct site name does not establish that each page’s metadata is correct.
Structured data needs to match the site’s real identity and content. Use the Article, organisation and breadcrumb checks for the separate markup workflow; adding more markup does not force a chosen search title.
Use snippet controls only for their intended scope
Google documents robots and snippet controls, including nosnippet, maximum snippet length and data-nosnippet for selected content. These are not commands to display your preferred meta description verbatim.
Before changing a template-wide rule, establish which URLs and text it affects. Accidentally suppressing useful summaries across an entire site is a different outcome from excluding one recurring piece of irrelevant text. Review the delivered rule after the change and retain a way to restore it.
Do not use snippet suppression as a privacy boundary. Information that must remain private needs appropriate access control; reducing its appearance in a search summary does not secure the page itself.
Distinguish today’s page from Google’s indexed observation
Use the URL Inspection documentation to distinguish indexed information from a live fetch. Record which URL was inspected and the observation date. A live accessible page does not prove that Google has already refreshed its stored result or will use your exact text.
- Save the intended metadata and verify the public HTML.
- Check that the page’s canonical and indexing settings match its intended role.
- Record the current search result and the query that produced it.
- Use the authorised Search Console property to inspect relevant evidence and, where appropriate, request a recrawl.
- Allow reprocessing before comparing again; avoid repeated speculative changes between observations.
For an inaccessible or excluded page, begin with the indexing diagnostic workflow instead of assuming the description caused the exclusion.
Measure outcomes without claiming that metadata caused everything
The Search Console Performance report provides clicks, impressions, CTR and position with page/query dimensions. Keep the before/after dates and compare an appropriate scope. Changes in query mix, position, seasonality or other page edits can affect the result too.
For a small new site, insufficient impressions are a limitation, not proof that the title failed. Maintain accurate metadata and useful content first. Accept the publishing work when the website serves the intended values and the page fulfils them; describe later search performance separately from that verified technical result.
Sources and verification
Editorial review 2026-10-13. Source check dates are recorded separately. Notes describe what each source supports and its limits.
View sources
- MDN: HTML title elementChecked The document title belongs in the head and supplies browser-tab/title information separately from visible headings.
Limits: No actual delivered document or browser tab was inspected. - MDN: Meta name attributeChecked A description metadata entry is distinct from an article heading and other named metadata.
Limits: A CMS field is treated as an input whose actual output still needs verification. - Google Search Central: Influencing title linksChecked Search title links are selected from multiple signals; concise descriptive titles and consistent prominent text help represent the page.
Limits: No exact wording, truncation length, update deadline or ranking result is guaranteed. - Google Search Central: Control snippetsChecked Snippets depend on page content and query; descriptions may be used and presentation can truncate them.
Limits: The fictional summary is editorial, and no fixed character count forces display. - Google Search Central: SEO Starter GuideChecked Useful organised current content and accurate representation are priorities; practices do not guarantee rankings.
Limits: No traffic or search demand has been measured for these candidate articles. - WordPress: wp_get_document_titleChecked WordPress assembles and filters document titles, with contextual components.
Limits: No SEO plugin is mandated or additional conflicting generator installed. - Google Search Central: Site namesChecked Automated site attribution differs from per-page title links and can use supported homepage identity signals.
Limits: WebSite markup is not claimed to force a particular attribution or page title. - Google Search Central: Robots and snippet controlsChecked nosnippet, maximum lengths and selective snippet exclusion have specific scopes.
Limits: These controls do not force the description verbatim and are not access protection. - Search Console: URL InspectionChecked Indexed observation and live-fetch output are separate evidence and current accessibility does not promise future display.
Limits: No authorised property or live/indexed report was operated by the worker. - Search Console: Performance reportChecked Clicks, impressions, CTR and position can be scoped by page/query and other dimensions.
Limits: A change in metrics is not attributed causally to metadata alone; no actual before/after result is claimed.