Discussion
  • Enterprise SEO

A global template change dropped traffic in some directories only. How do you isolate the cause across different content types?

Published 1 reply

Discussion starter: an enterprise 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

    When a global template change hurts only some directories, the cause is almost always an interaction between the change and something specific to those directories: a content type, a component that behaves differently, a data dependency, a rendering path or a linking pattern. The way to isolate it is to compare affected and unaffected directories systematically, before and after, on the things the template controls. Treat it as a difference-finding exercise: the template is the same everywhere, so the question is what is different about where it hurt.

    Start by confirming the shape of the loss

    1. Timing. Line up the deployment with Search Console data by directory. Hourly data from the Search Analytics API can show whether the drop started within hours of the release or days later.
    2. Scope. List which directories dropped, by how much, and which stayed flat or rose. Include directories that share the template but did not lose traffic; they are your control group.
    3. Metric. Did impressions fall, or clicks with stable impressions? Falling impressions point to crawling, indexing or ranking; falling clicks with stable impressions point to how results display.
    4. Queries. Within affected directories, did head terms or long-tail queries fall? A long-tail collapse often means content or links disappeared; a head-term drop can mean relevance signals changed.

    Compare affected and unaffected directories on what the template controls

    Element What to compare Typical culprit
    Rendered main content Text length and headings in rendered HTML, before and after Long content moved into tabs or accordions that load on interaction; truncation on certain content types
    Internal links Number and targets of links per page Related-content or breadcrumb modules removed or rendered client-side for some types
    Structured data Presence and validity by content type Markup generated from fields that only some content types have
    Canonicals, robots and hreflang Values on sample URLs Logic that behaves differently for paginated, filtered or localised pages
    Titles and headings Changes in how they are generated A title pattern that suits products but breaks articles
    Page weight and speed Response size, render time Heavy components only on some types; HTML over size limits on long pages

    Crawl a sample from each affected and unaffected directory on the new template and on the old version (from staging, an archive or a rollback environment). Differences that appear only in affected directories are your candidates.

    Use logs and Search Console to confirm

    • Server logs by directory: did Googlebot’s request volume, status codes or response sizes change in affected directories after the release?
    • Page indexing report, ideally with a separate sitemap per directory so you can see indexing changes by section.
    • URL Inspection on a sample from affected directories: compare the rendered HTML Google sees with what you expect.
    • Search appearance filters in the Performance report: did rich results disappear for one content type?

    Rule out coincidences

    Before concluding the template is responsible, check the alternatives. Did a core update roll out at the same time, and do the affected directories contain the kind of content that update targeted? Is there seasonality specific to those content types? Did a competitor or a new SERP feature appear for those queries? A template regression usually shows up quickly and aligns closely with the release; an update or seasonal effect usually does not align as neatly and affects comparable content on competitors’ sites too.

    Hypothetical example:

    A publisher rolls out a new article template across news, guides, recipes and reviews. Guides and recipes lose a large share of impressions within a week; news and reviews are stable. A crawl comparison shows that the new template collapses any section longer than a set length into an accordion that loads its content with JavaScript on click. News and reviews are short, so nothing changed for them. Guides and recipes are long, and most of their content now depends on interaction. URL Inspection confirms the rendered HTML Google sees is missing most of each guide. The team changes the component to include all content in the HTML with the accordion as a visual treatment only, and impressions recover over the following weeks.

    Organise the investigation so it finishes

    These investigations drag on when they are run as open-ended exploration. A simple matrix helps: directories as rows, template elements as columns, and in each cell whether the element changed for that directory and whether the metrics moved. The pattern usually jumps out. Assign one person to own the matrix, and set a short deadline for a first hypothesis, with engineering on standby to confirm it.

    Decide how to fix it

    Once you have a mechanism, the fix is usually specific to a component: change how it renders for long content, restore a linking module for certain types, fix structured data generation for a content type. If a fix will take time and the loss is large, roll back that component for the affected directories while the fix is built. Rolling back the whole template is rarely necessary once the cause is known.

    Watch for this:

    Global templates are often tested on a few sample pages that represent the most common content type. Build a test set that includes the longest, shortest and most unusual pages of each type, such as long guides, pages with many tables, localised pages and paginated archives, and check them before the next global release.

    Confirm timing and scope, compare affected and unaffected directories on what the template controls, confirm with logs and Search Console, rule out updates and seasonality, and fix or roll back the specific component rather than the whole release.

    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.