A control can look like a link without containing a destination that a crawler can extract. A card may open when clicked because JavaScript listens for the event, while the underlying markup has no anchor or usable href.
Inspect the generated HTML when checking internal links. The visible design and the content editor’s link field do not establish what the final page prints.
Use an anchor with a destination
<a href="/guides/shoe-sizing/">Running shoe sizing guide</a>
The anchor gives the link structure, href identifies the destination, and the text explains what is there. Google recommends this pattern in its crawlable-link guidance.
An empty href, a JavaScript-only destination, or a click handler on a generic container needs a different investigation. A button can be appropriate for an action such as opening a menu; navigation to another page should expose the actual destination.
Check how relative paths resolve
Both relative and absolute URL references can work. A path starting with / resolves from the site’s root. A path without that leading slash resolves relative to the document’s base location.
<a href="/guides/">Guides</a>
<a href="https://example.com/guides/">Guides</a>
These can lead to the same address on the intended site. A relative guides/ reference on a deeply nested page may lead somewhere else. Inspect the resolved URL before blaming the server for returning a missing page.
Write anchor text for the destination
“Running shoe sizing guide” tells the reader more than “click here.” Use the words that describe the linked page in the current context. Repeating a long commercial phrase in every paragraph can make the article harder to read.
A heading may contain a link, and a navigation menu can contain a list of links. Those surrounding elements do not change the need for a working destination. Check image links too: the image and its alternative text contribute to the accessible link name.
The title attribute is not a replacement for clear visible text. Opening a new tab with target changes browsing behaviour; it does not make a broken destination work.
Review follow settings intentionally
Ordinary links do not need a rel="follow" attribute. A link-level nofollow value and page-level robots settings operate at different scopes. Nofollow is also not a security control or an absolute guarantee that a URL will never be fetched.
Follow a complete discovery path
Start at the homepage or an important category and follow the links to a detail page. Include pagination and related items. A URL listed in a sitemap may still have no useful route from the interface.
Use a crawler’s Inlinks view to find which pages print a destination. The broken pagination investigation shows how to connect a failed request with the source element and its shared template.
For generated navigation, my website architecture and programmatic SEO service covers route and template rules. Classifieds and marketplace SEO often needs the same checks across categories, locations, and frequently changing listings.