Skip to the indexed answers
Digital Marketing Index

usdigitalmarketingindex.com SEO

How can a redesign of content hierarchy reduce organic traffic even when all URLs stay unchanged?

Because a URL is only an address, and search engines evaluate what is at the address and how the rest of the site connects to it. A redesign that keeps every URL can still change how many clicks a page sits from the home page, which pages link to it and with what words, what headings and text the page itself contains, and whether its content appears in the HTML at all. Each of those is something search engines read. No redirect is needed, no error appears, and the page can lose visibility all the same. “We did not change any URLs” rules out one cause of a post-redesign decline. It leaves the others standing.

Internal link depth changes

A new hierarchy moves pages relative to the home page.

A before-and-after site tree with identical destination URLs:

BEFORE                                   AFTER

Home                                     Home
|-- Services  (main menu)                |-- What we do  (main menu)
|   |-- /boiler-repair/                  |   |-- Heating
|   |-- /boiler-service/                 |   |   |-- Repairs and servicing
|   |-- /power-flush/                    |   |   |   |-- /boiler-repair/
|   |-- /radiator-install/               |   |   |   |-- /boiler-service/
|-- Guides  (main menu)                  |   |   |-- Upgrades
|   |-- /guides/boiler-pressure/         |   |       |-- /power-flush/
|   |-- /guides/bleed-radiator/          |   |       |-- /radiator-install/
|-- Footer: all services listed          |-- Resources  (footer only)
                                         |   |-- /guides/boiler-pressure/
                                         |   |-- /guides/bleed-radiator/
                                         |-- Footer: top-level sections only

What changed, though no URL did:

Page Clicks from home, before Clicks from home, after Linked from
/boiler-repair/ 2, and 1 through the footer 4 Fewer pages
/power-flush/ 2, and 1 through the footer 4 Fewer pages
/guides/boiler-pressure/ 2 2, through the footer only No main-menu path

Consequences of that kind of change:

  • A page linked from every page of the site, through a menu or footer, becomes a page linked from a handful.
  • A page that was two clicks deep becomes four.
  • Some pages lose every internal link and become orphans, still live at their URL.

Google’s documentation says it uses links to find pages to crawl and as a signal when determining the relevancy of pages (Google’s guidance on crawlable links and anchor text). A page that the site itself now treats as minor is presented to search engines as minor.

Navigation labels and topical context

The words in links changed too.

Before After
Menu link text: "Boiler repair" Reached under "Repairs and servicing"; the page name appears only on the fourth-level listing
Section: "Services" "What we do"
Section: "Guides" "Resources"
Breadcrumb: Home > Services > Boiler repair Home > What we do > Heating > Repairs and servicing > Boiler repair

The same guidance says good anchor text tells people and Google something about the page being linked to. Replacing descriptive labels with brand-voice ones, or removing links that named the page’s subject, removes that description.

Grouping matters as well. Pages that used to sit together under a clear parent may now be spread across sections, and a hub page that linked to all of them may be gone. Breadcrumb markup, if generated from the new tree, now describes a different position.

Template and content alterations

A hierarchy redesign can bring new templates with it, and then the content of pages changes without anyone editing an article.

Change Effect
Headings rewritten or demoted The page's stated topic changes; an H1 may now be a slogan
Introductory text shortened or removed Less content for the query the page answered
Sections moved into tabs or accordions loaded on click Content absent from the initial HTML, or absent from rendering if it needs interaction
Body copy replaced by images or video Text removed
Related-links modules removed Internal links lost
Titles regenerated from a new pattern Every title changes at once
Category pages lose their listings or descriptions Hubs become thin
Content now rendered by JavaScript Depends on rendering succeeding
Structured data dropped or changed by the new template Rich result eligibility changes
Canonical, robots, or hreflang tags altered by the template Indexing signals change
Slower pages Performance measurements change
Mobile layout hides sections With mobile-first indexing, the mobile version is what counts

None of these alters a URL. Several alter what Google indexes at it.

Investigating historical differences

The investigation compares the site before with the site after.

Sources of the “before”:

Source Gives
A crawl saved before launch Link depth, inlinks, titles, headings, word counts, canonicals for every URL
The old site on a staging or backup server Full pages
Web archive snapshots Old HTML for sampled pages
Search Console performance data Clicks, impressions, and queries by page, before and after the date
Old sitemaps and navigation files The old structure

Procedure:

  1. Fix the launch date and annotate it.
  2. In Search Console, compare the periods before and after, by page. List the pages with the largest loss.
  3. Check first for technical regressions on those pages: status codes, noindex, canonical changes, robots rules, rendering.
  4. Crawl the new site. For the losing pages, compare with the old crawl: click depth, number of internal links pointing in, anchor text of those links.
  5. Compare page content: title, H1, headings, body text length, presence of key sections in the served and rendered HTML.
  6. Compare queries: which queries did each page lose? Do they correspond to text or headings that were removed?
  7. Check whether another page on the site now receives those queries.
  8. Rule out outside causes: seasonality, by comparing with the same period a year earlier, and known search updates around the date.

A comparison sheet for one page:

Attribute Before After Changed
URL /power-flush/ /power-flush/ No
Status, indexability 200, indexable 200, indexable No
Click depth 2 4 Yes
Internal links in 212 9 Yes
Anchor text in "Power flush" "Learn more" Yes
Title Power flushing for central heating Upgrades Yes
H1 Power flush Make your heating work harder Yes
Body words in served HTML 640 180 Yes
FAQ section In HTML Loaded on click Yes
Structured data Service, FAQ None Yes

The values are invented to show the form. A sheet like this turns “traffic fell after the redesign” into a list of specific differences, each of which can be reversed or kept by decision.

Recovery follows from the list: restore descriptive links to important pages from the navigation or hub pages, restore titles and headings that state the subject, put removed text back in the HTML, and retest. Changes take effect as pages are recrawled, and results are read in the same reports over the following weeks.

The lesson for the next redesign is to save a full crawl beforehand and to treat internal linking, headings, and on-page text as things that need sign-off, as URLs do.

One specific case, content present on desktop and missing from the mobile layout, has its own method of detection, covered in the guide to how to identify important content missing from the mobile version of a website.

The summary: unchanged URLs protect against broken addresses, and nothing else; depth, links, labels, and content can all change beneath them.

Link trail

Reader comments on this answer

Add your comment to this answer

Your email address will not be published. Required fields are marked *