A meta tag analyzer shows what a published page actually returns, so you can check the title and description before blaming a CMS setting or a search result. Screpy reads one public HTML page, lists its head declarations and shows canonical and robots evidence separately. Use the report to find missing text, duplicate output or a value belonging to the wrong page, then correct the source that generates it.
How to use the Meta Tag Analyzer
- Enter the complete public URL of the page you edited, including its path. A homepage check does not inspect a product or article underneath it.
- Submit the URL and check the Final URL in the result. If the address redirected, the metadata belongs to the destination.
- Read the title, description and their character counts, then inspect the declaration table. Rows preserve source order and show a source line when available.
- Open the canonical and robots details before deciding which changes to make.
The table includes <title>, <meta> and <link> elements from the fetched head, so it can contain stylesheets and other resource links alongside SEO fields. A row is evidence of a declaration, not a verdict that its value is useful or that a search engine uses it.
Fix missing and duplicate meta tags
Warnings focus on title, description and viewport. A missing declaration or a blank first retained value needs review. More than one declaration with the same checked name triggers a duplicate warning, even when the values match. These are duplicates within the fetched page; the tool does not compare titles across a website.
This fictional head fragment has two descriptions:
<meta name="description" content="">
<meta name="description" content="Compare walking routes, distances and access information.">
The first description is empty. Screpy therefore reports the missing-value warning and the duplicate warning, and its illustrative preview uses that first value. It does not choose the later sentence for you or predict which text Google will display.
Find the source of both rows: a theme, SEO setting or page template may be adding a declaration independently. Keep one intended description in the delivered head rather than adding a third tag. Use the same approach for duplicate titles and viewport tags. If a value is visible but still flagged, compare its name with the checked spelling, such as description; differently written property values may remain in the table without matching that check.
Read title and description counts without chasing a limit
The counters describe the first retained title and description, using Unicode code points. They do not measure pixel width, score your writing or define a ranking threshold. An emoji or a character formed from several code points also makes character count different from visible text length.
Write a title that identifies this page's subject and a description that explains its specific value. Put important information early and remove repeated phrases or generic boilerplate. For a walking-route page, distance and access information are more useful than a list of loosely related travel keywords.
Google's title-link guidance describes several possible sources for a displayed title. Its snippet guidance explains that page content can supply the snippet and that different queries can produce different text. Google sets no fixed character limit for either field; available display space varies. The preview here illustrates observed source text, so a tidy preview cannot confirm the actual search result.
Check canonical and robots evidence separately
A green metadata status does not prove that the page is indexed or ready to appear in search. This analyzer's missing and duplicate warnings concern the three checked fields; canonical and robots declarations are displayed for separate inspection.
Compare each canonical's raw value and resolved destination with the URL you intend to prefer. A canonical is a preference signal, not proof of Google's selected URL. If several targets appear or one points to an unrelated page, use the Canonical Tag Checker for a focused review. Google's canonical guidance recommends consistent signals and absolute canonical URLs.
Read both HTML robots values and observed X-Robots-Tag response-header values, including the crawler they target. A valid title cannot override noindex. Google's robots guidance explains restrictive conflicts and the need for crawler access to read the rules. Use the Noindex Checker to investigate those declarations. Do not remove an intentional exclusion from a private or unwanted search page just to clear a warning.
Interpret viewport and other head declarations
The Meta Tag Analyzer lists more evidence than its automatic warnings evaluate. Review each field according to its purpose:
| Declaration | What to review |
|---|---|
viewport |
Check the intended width and scale. Presence alone does not prove a usable mobile layout. |
charset |
Inspect the declared encoding. A listed value is not a complete encoding audit. |
og:* and twitter:* |
Confirm that sharing copy belongs to this page. Use the Open Graph Checker for focused Open Graph checks. |
Stylesheet or other <link> rows |
Identify resources in the head; they are not extra descriptions to delete. |
keywords |
Do not add it merely to satisfy an imagined Google SEO requirement. |
For a typical responsive page, width=device-width, initial-scale=1 is a common viewport starting point. MDN's viewport reference explains those settings and warns against preventing user zoom. Check the actual layout separately. Google's supported-tag reference states that meta keywords have no effect on Google indexing or ranking; finding a tag does not establish its search value.
Recheck the published page and respect report limits
Change the responsible CMS field, template or response-header setting, publish the correction and check the same public URL again. Confirm the intended value and that unwanted duplicates disappeared. If the checker and your browser disagree, compare source HTML with the DOM after JavaScript: this tool does not run page scripts. A different language, redirected page or access screen may also explain unexpected text.
Up to 300 head declarations are retained, alongside separate limits of 50 canonical and 100 robots declarations. Partial coverage can indicate a retention limit or incomplete interpretation of robots evidence. A missing retained row then does not establish that the source lacks it. Long metadata values are shortened to 4,096 UTF-16 units; the preview and counts use the retained text, and this shortening alone does not necessarily mark the report partial.
An unavailable or unsupported response is a failed check, not a clean page. Verify that the URL returns accessible HTML and retry when appropriate. Download the JSON report when you need to share the observed evidence with the person fixing the template. For a wider review after metadata corrections, use the On-Page SEO Checker; this single-page metadata check neither edits your site nor confirms Google has recrawled it.