A single Monday morning after launch, a page that used to sit third for its money keyword slides to page four, and the traffic report for the quarter starts bleeding out in real time. The redirects are clean. The Core Web Vitals are better than they were. Nothing was deindexed.

And yet the phone stops ringing. When a redesign quietly costs rankings, the culprit is almost never the technical migration everyone spent weeks stress-testing. The culprit is the copy someone rewrote along the way, which shifted what the page appears to be about in the eyes of a search engine sorting for intent.

Search intent mismatch is the failure mode nobody catches in QA because the page still looks great. It answers a slightly different question than the one people are typing in. For a diagnostic walkthrough of the same pattern, listen to SEO.co on why good content fails to rank, which covers the ground clearly.

The Problem Nobody Catches Until Traffic Drops

A page ranks because a search engine has decided, over time, that it satisfies a specific query well enough to earn the click. That decision hangs on signals baked into the copy: the headings, the first two hundred words, the questions the page answers explicitly, the modifiers around the target term, the internal links pointing in.

Rewriting a page for a redesign changes most of those signals in an afternoon.

The rewrite is usually well-intentioned. The old service page was long and clunky. Somebody trimmed it. A designer wanted parity across templates, so the FAQ block came off.

The new brand voice called for cleaner headings, so the H2 that used to read "emergency water damage restoration in a residential basement" became "we're here when you need us." All defensible edits. Every one of them narrows what the page tells a crawler it's for.

The Obvious Fix Falls Short

The instinct when rankings drop after a launch is to audit the technical migration: check the redirects, re-crawl the site, diff the sitemaps, make sure nothing shipped with a noindex tag. This is the right first pass. It solves the problem some of the time, and misses it the rest.

A permanent server-side redirect passes ranking signals between an old URL and a new one, and Google's own redirect guidance is clear that a clean 301 is the recommended path when URLs change. What that guidance can't do is make the destination page mean the same thing to a query as the old one did. If the page at the end of the redirect is now a shorter, more brand-forward, less specific version of the original, the redirect worked exactly as designed and the ranking still slipped.

How the Mismatch Actually Shows Up

A few patterns account for most post-redesign intent drift. If you're diagnosing your own site, look for these before touching anything else:

  • The commercial page turned informational. A service page that used to open with pricing, service areas, and a booking form gets rewritten to open with a brand story. The query is transactional. The page now reads like a magazine feature.
  • The specific page turned generic. Headings that once named the exact service, problem, and deliverable get flattened into elegant one-liners. What replaced them is prettier and less legible to a crawler trying to match intent.
  • The FAQ block disappeared without anyone flagging it. Featured snippets and People Also Ask placements are often earned by explicit question-and-answer content the designer viewed as clutter. When that block goes, so does the snippet.
  • The template changed for a whole page type. If the drop is limited to one kind of page, the rewrite happened inside the template, not the individual content. Look at that template for removed descriptions, changed headings, or a stray noindex before anything else.
  • The refresh created near-duplicates. A content refresh sometimes spins up a second page covering the same territory in slightly different words. The two now split the signals the original page used to hold alone, and both underperform.

What Actually Works

Fixing intent mismatch is less about writing more and more about writing to the query. Here is the workflow that reliably moves pages back:

  1. Read the SERP before you read the page. Open an incognito window and search the target query. Look at the top ten results. Are they buying guides, product pages, calculators, or definitions? Your page has to belong in that set.
  2. Compare the old page to the new one, line by line. Pull the pre-redesign version from the Wayback Machine. Note every heading that changed, every section that was removed, and every modifier that got softened.
  3. Restore the specific language, not the old design. Put the intent-carrying language back into the new template: the concrete headings, the question-and-answer content, the service and product specifics, the modifiers that told a crawler what job the page did.
  4. Consolidate before you create. If two pages now compete for the same query, pick the stronger URL, merge the useful content from the other into it, and redirect the loser.
  5. Recheck the internal links pointing in. If every internal link used to point at your service page with the words that named the service, and now they all say "learn more," you've stripped a signal the page depended on. Restore descriptive anchors where they read naturally.

None of this requires new backlinks or a fresh authority push. It requires putting back the parts of the page that told a search engine, unambiguously, what query it was the right answer to.