Case study 01

Breeze Furnishings

E-commerce · Furniture & wallcoverings · WooCommerce

A noindex misconfiguration worked silently for over a year, deindexing 79 brand and product-tag pages. The work was finding it, unwinding the damage methodically — and the recovery signals are already showing: mobile average position up 8.7 places, Organic Shopping sessions up 29%, revenue per user up 26%. The underlying business never broke.

The problem

79

pages deindexed over 14 months — invisible to Google, working normally on site

The work

986

redirects mapped, noindex corrected, canonicals and structure fixed

The result

+26%

revenue per user — conversion rate unchanged throughout, position recovering

The verdict

"He truly excels at what he does"

— Breeze Furnishings

Background

To call this a migration recovery is an understatement. When Breeze Furnishings moved from Visualsoft to WooCommerce in July 2024, the redirects were completed but none of the content came across — like moving into a new house and the removal van not turning up. With no tracking in place post-migration and no content to support rankings, organic traffic took an immediate hit.

When I was approached, the priority was clear — not simply to restore missing content (that moment had passed) but to write updated copy reflecting current products and search intent, rebuild the blog with content focused on trends and styling, and build internal linking opportunities at the same time.

Further analysis in early 2026 then uncovered a second, more serious layer to the problem.

"Two separate issues compounded each other: a migration that launched without its content, followed by a noindex misconfiguration that quietly stripped the pages that had been rebuilt. Neither alone would have been catastrophic. Together, they meant the site spent the better part of two years working against itself."

The noindex issue

A Rank Math misconfiguration had set the site to noindex in January 2025. Over the following 14 months, Google progressively deindexed product and category pages. Because the decline was gradual, it didn't trigger an obvious single-event alert. By the time it was identified in late February 2026, 79 pages were confirmed deindexed and approximately 986 of 1,000 crawled URLs were returning 404 errors. The noindex directive was corrected immediately and re-indexation began within days.

Because the directive was applied in configuration rather than on-page, it left no visible trace. The pages continued to function normally. Traffic decline was gradual enough to be initially misread as seasonality. By the time the pattern became impossible to ignore, over a year of compounding loss had accumulated.

What the audit found

The crawl confirmed 79 pages carrying noindex directives that should not have been there — brand pages and product-tag taxonomies that were, in effect, invisible to Google despite being live and fully populated with products.

"The pages were working. Google just couldn't see them — and had spent a year being told not to look."

Beyond the noindex issue, the audit identified a cluster of structural problems that needed resolving to support a clean recovery: duplicate paginated page titles across taxonomy archives, cross-URL canonical conflicts, approximately 70 double H1s introduced via a Divi theme template, and over 600 pages missing meta descriptions entirely. None of these caused the crash — but all of them would slow recovery if left unaddressed.

GA4 analysis confirmed the picture: near-total collapse of the organic channel, with paid and direct largely intact. Crucially, revenue-per-session from the remaining organic traffic had held — indicating that the site's conversion quality was sound. This was a visibility problem, not a relevance or trust problem.

Search visibility impact

Source: Google Search Console — Q1 2025 vs Q1 2026, like-for-like 91-day periods.

Impressions nearly halved and organic clicks fell by more than a quarter. The branded/non-branded split tells the most important part of the story: brand demand was growing throughout this period, but the product and category pages needed to convert that intent were being systematically deindexed.

The average position improvements reflect the start of recovery — as deindexed pages were reinstated, those now visible are ranking significantly higher than before. Average position is a directional indicator rather than a precise measure given personalisation and device variance, but the scale of improvement across both mobile and desktop is meaningful.

GSC — Q1 2025 vs Q1 2026
Organic clicks−27%  (3,064 → 2,250)
Impressions−48%  (454,924 → 237,833)
Non-branded clicks−50%  (652 → 329)
Branded clicks+29%  (184 → 238) — brand growing while product pages deindexed
Average position+7.1 places  (17.7 → 10.6)
CTR+42%  (0.67% → 0.95%)
Mobile avg. position+8.7 places  (20.6 → 11.8)
Desktop avg. position+17.1 places  (34.2 → 17.1)
"Branded clicks grew 29% year-on-year while non-branded fell 50%. The brand was building — people were actively searching for Breeze Furnishings by name in greater numbers — but the pages they needed couldn't be found."

Traffic & revenue impact

Source: GA4 Traffic Acquisition by session channel group — Q1 2025 vs Q1 2026.

Sessions and revenue both fell, but the gap between them tells the important story. Revenue declined 19% on a 37% session drop — the business was extracting significantly more value from every visit that did come through. The conversion rate held exactly flat at 1.65% across both quarters. This was a discoverability problem, not a site quality problem.

GA4 — Q1 2025 vs Q1 2026
Sessions−37%  (10,754 → 6,790)
Revenue−19%  (£36,133 → £29,292)
Transactions−37%  (177 → 112)
Conversion rateUnchanged — 1.65% both quarters
Avg order value+28%  (£204 → £262)
Revenue per session+28%  (£3.36 → £4.31)
Engagement rate+5pp  (64.5% → 69.5%)

The channel breakdown reflects the selective nature of the damage. Organic Search took the hardest hit at −66% sessions. Direct proved most resilient — users who already knew the brand kept returning and spending at virtually the same rate throughout. Organic Shopping sessions grew 29% and returning users grew 39% — the feed re-indexation is beginning to compound, with average engagement time up 165% (61s → 162s) and events per session more than tripling.

"The site got harder to find. It didn't get harder to buy from."

The work

Recovery work was sequenced to address the root cause first, then the structural issues, then the migration-related redirect gaps that had accumulated separately from the core problem.

Work log — sequenced by priority
Root cause Rank Math noindex configuration corrected across all 79 affected brand and product-tag URLs. Fix applied February 2026.
Redirects .htaccess redirect file prepared mapping /brand/* to /product-tag/* to resolve URL structure conflicts and consolidate link equity. ~986 of 1,000 identified 404s resolved.
Pagination Duplicate paginated titles across taxonomy archives fixed via Rank Math %page% token implementation.
Canonicals Cross-URL canonical conflicts identified and resolved across affected product and taxonomy pages.
H1 structure ~70 double H1s traced to a Divi template-level issue. Template corrected to remove duplicate heading output.
Meta data 621 pages flagged as missing meta descriptions. Priority batch addressed; remainder added to ongoing content schedule.
Merchant feed Google Merchant Center feed migrated from CTX Feed to Google for WooCommerce for API-based sync. Salesfire connected via GMC integration.

Where things stand

This is a mid-recovery snapshot. The noindex fix was applied in February 2026 and the data above covers Q1 2026 — meaning most of the recovery period falls outside the comparison window. The position improvements in GSC and the Organic Shopping growth are the early recovery signals, not the completed picture.

The Breeze Furnishings case illustrates how SEO problems compound quietly before they become visible. A single configuration error, applied at the wrong level, had a larger and longer-lasting impact than any individual piece of content work could have caused or corrected. The most important finding required a crawl to surface — there was no visible trace on the site itself.

The data shows the underlying business was healthy throughout. Conversion rate held flat. Revenue per user grew. Direct users kept coming back and spending more. The work done on content, products and UX during the period of the issue was working — it just couldn't be seen through the noise of the visibility collapse. The task now is restoring the discoverability a site that earned its traffic once and can earn it again.

Recognise any of this?

Unexplained traffic decline, a gap between what the site should be doing and what the data shows — these are worth investigating properly. Get in touch and we'll start with what the crawl shows.