Discussion
  • Crawling
  • Internal Linking
  • Technical SEO

A mobile template hides most internal links behind tabs and accordions. Has anyone measured whether that changes crawl paths compared with desktop?

Published 1 reply

Discussion starter: a technical SEO question, posted with an answer from Niraj Raut to open the thread. If you have dealt with this on a site, reply with what you saw, especially where it differs.

Disclosure: Niraj Raut, who posted this answer, runs the SEO consultancy linked at the end of it.

1 reply

  • admin
    Admin
    Original poster

    Collapsing links into tabs and accordions does not, by itself, change the paths Google can crawl. What changes them is whether the links exist in the mobile HTML at all. Google indexes the mobile version of a page, crawled with its smartphone agent, and it follows URLs found in the href of <a> elements. A link inside a collapsed panel that is already in the HTML is still a link. A link that only appears after someone taps a tab, opens a menu or presses “load more” does not exist for Googlebot, because Google’s crawlers do not click.

    That changes what is worth measuring. For Google, the desktop template is no longer the version that counts, so the useful comparison is a diff: which links does desktop expose that the mobile HTML does not? Those are the paths at risk. Whether a link inside a collapsed panel carries the same weight as a visible one is a separate and harder question. Google does not document it, and no public controlled test that I can point to isolates it. Discovery is well understood. Emphasis is not, and that is the part worth testing.

    What Google documents, and where it stops

    Google’s mobile-first indexing guidance is explicit on three points. First: “Google uses the mobile version of a site’s content, crawled with the smartphone agent, for indexing and ranking.” Second, a different mobile design is fine, for example “moving content into accordions or tabs”, as long as you “make sure that the content is equivalent to the desktop site.” Third: “Google won’t load content that requires user interactions (for example, swiping, clicking, or typing) to load.”

    Google’s pagination and incremental loading guidance adds the crawler’s side: “Google’s crawlers don’t ‘click’ buttons and generally don’t trigger JavaScript functions that require user actions to update the current page contents.” When looking for pages to index, Google “generally crawls URLs found in the href attribute of <a> elements.”

    Where the documentation stops: the mobile-first page talks about content equivalence and says nothing specific about internal links on mobile versus desktop, and neither page addresses whether a link inside a collapsed panel is weighted differently. Everything beyond that is inference, including what follows about emphasis.

    Hidden is not the same as absent

    Whether an accordion affects crawling comes down to how it was built. The same design can put links in the HTML or leave them out.

    Mobile pattern In the HTML Googlebot processes? Effect on crawl paths How to check
    Panels in the server HTML, collapsed with CSS, the hidden attribute or <details> Yes None for discovery View source and search for a URL you know sits in a closed panel
    Panels built by JavaScript at page load, then collapsed Yes, after rendering Found a step later, once the page is rendered Compare the raw HTML with the rendered HTML
    Panels whose content is created or fetched only when opened (some components unmount closed panels by default) No Links missing from Google’s link graph Rendered HTML from URL Inspection’s live test
    Menu built or fetched when the menu button is tapped No Navigation links missing on every page Same test, on any template
    “Load more” or infinite scroll with no numbered <a href> page links Only the first batch Deeper items lose their listing-page paths Look for page links such as ?page=2 in the HTML
    Separate mobile HTML (dynamic serving or separate mobile URLs) with fewer links Only what the mobile HTML contains Desktop-only links drop out of the version Google indexes Fetch with a smartphone user agent and compare
    Link-like items built as buttons or click handlers instead of <a href> No Not crawlable as links on either template Inspect the markup

    The second row is worth understanding because it is the subtle one. Google’s JavaScript SEO basics describe the sequence: Googlebot parses the fetched HTML for links, queues the page for rendering, then parses the rendered HTML for links again. Links in the server HTML can enter the crawl queue at the first step; links that exist only after JavaScript runs have to wait for rendering. For the fifth row, the same pagination guidance recommends linking each page to the next with <a href> tags and giving each page a unique URL, which a “load more” button can sit on top of without changing the design.

    Why desktop is the wrong baseline, and when it still matters

    Because Google indexes the mobile version, the desktop link graph is a reference, not a second chance. The Search Console Crawl Stats report breaks requests down by Googlebot type (Smartphone, Desktop, Image, Video, Page resource load, AdsBot, StoreBot and other agents), so check the split on your own property rather than assume it. Links present on desktop but missing from the mobile HTML are missing from the version Google indexes.

    Desktop still matters for two reasons. It is the reference for the diff. And other crawlers do not necessarily fetch the mobile template or run JavaScript. Crawlers that do not execute JavaScript see only the server HTML, so for them the second row in the table behaves like the third.

    Scale changes the stakes. Google’s crawl budget guide is written for sites with a million or more unique pages changing about weekly, sites with 10,000 or more pages changing daily, and sites with many URLs in “Discovered - currently not indexed”. On a site of a few hundred pages, XML sitemaps, body links and breadcrumbs can cover a missing mobile path for discovery, so the real question there is emphasis. On a large catalogue, a missing path can leave deep URLs discovered late and recrawled rarely.

    Discovery versus emphasis

    Discovery is a yes-or-no question answered by the HTML. Emphasis is a matter of degree, and nothing documented answers it. Link-analysis patents have described weighting links by features such as position and prominence, but a patent does not show what runs today. On the other side, Google explicitly accepts accordions and tabs for content on mobile. That is an argument, not evidence, that links in the same containers are not heavily discounted. Neither side has public proof.

    The practical risk is narrower than the debate. It concentrates in URLs whose only internal links sit inside collapsed panels. A product linked from a category grid, a breadcrumb and a collapsed “related” panel does not need the panel. A guide linked only from a collapsed footer section does.

    How to measure it on your own site

    1. Diff the link graphs. Crawl the site twice with a JavaScript-rendering crawler, once with a smartphone user agent and viewport and once with a desktop one, and export unique inlinks and crawl depth per URL. URLs whose inlinks fall sharply on mobile, or that are reachable only on desktop, are the at-risk set. A third crawl, mobile without rendering, shows which links exist only after JavaScript runs.
    2. Confirm it from Google’s side. Run URL Inspection’s live test on each key template, open the tested page’s HTML, and search for URLs you know sit in closed panels or the mobile menu. If they are missing there too, the diff is not an artefact of your crawler settings.
    3. Check who is crawling and why. In Crawl Stats, group by Googlebot type for the smartphone and desktop split, and by purpose to track Discovery (“The URL requested was never crawled by Google before”) against Refresh (“A recrawl of a known page”). The report is available only for root-level properties, and its example URLs “are not comprehensive”.
    4. Use logs for behaviour, not paths. Server logs show which URLs Googlebot requested and when, not which link led it there. Segment URLs by where their internal links sit (visible, collapsed panel only, desktop only) and compare the time from publication to first Googlebot request, and the interval between recrawls.
    5. Run a discovery test if you need a causal answer. Publish two matched sets of new URLs: one linked only from inside a closed accordion on existing hub pages, the other linked only from visible body copy on comparable hubs, with neither set in the XML sitemap until the test ends. Measure time to first Googlebot request and time to indexing. Set the length from your logs: the test has to span several recrawls of every hub, so hubs recrawled daily need days to a couple of weeks, and hubs recrawled weekly need proportionally longer.
    6. Test emphasis separately. Split a large set of comparable hub pages, move the same priority links out of collapsed panels on half of them, and measure impressions and clicks on the target URLs (not the hubs) against targets linked from the control hubs. This needs many hubs, and targets that are not linked from both groups; otherwise the groups contaminate each other. Read it only after the hubs have been recrawled and the targets’ rankings have had time to stabilise.

    Hypothetical example:

    A retailer’s desktop category template has a mega menu with 60 category links in the HTML, a “related categories” block with 12 links, and 48 products per page with numbered page links. On mobile, the menu is fetched as JSON when the menu icon is tapped, “related categories” sits in an accordion whose component renders the panel only when opened, and a “load more” button replaces the page links after 24 products. A smartphone crawl finds 24 product links and a breadcrumb per category page, and nothing else.

    The accordion is not the problem. The render-on-open component, the fetched menu and the missing page links are. Subcategories that desktop reached through the mega menu now rely on XML sitemaps and body links, and products beyond the first 24 in each category have lost their category path. The fix keeps the design: render panels at load and hide them with CSS, server-render the menu, and put numbered <a href> links behind “load more”.

    What result would change the decision

    • The mobile HTML lacks links the desktop has: fix it now. No test is needed, restoring links to the HTML is low risk, and the design can stay.
    • Links are present, the discovery test shows no lag, and recrawl intervals match: keep the accordions. Their usability value is not costing you crawl paths.
    • Links are present but panel-only URLs lag or are recrawled less: add visible contextual links to the URLs that matter commercially, and keep the accordions for the long tail.
    • The emphasis test shows target URLs gaining when links are visible: move priority links out of collapsed panels on that template, then repeat the test on the next template rather than assuming the result transfers.

    The position I would take

    If this were my site, I would not start with a crawl-path study. I would start with the diff: a smartphone crawl with rendering against a desktop crawl, confirmed with URL Inspection on each template. If the mobile HTML is missing links, that is the finding, and it is an engineering ticket rather than a debate about accordions. If the links are all there, the remaining question is emphasis, which only a split test on your own templates can answer, and which is worth running only for links to pages that earn money.

    Need help with this on your own site?

    Niraj Raut works with businesses in Nepal, Australia, the UK, Europe and the US on technical SEO, ecommerce SEO, local SEO and AI search.

    See SEO services

Add a reply

Replies are open to approved members. Log in or apply to join.