Parameter URLs
14,200
Sort orders, session IDs and filter combinations on every listing, each one a crawlable, indexable URL with its own self-referencing canonical. Most were near-copies of about 30 real pages.
Finding 1 of 4
A fictional business. The figures are a modelled scenario, not a client result.
BRODOREMONT VARNA · VARNA
Index cut by 98%, service-page impressions +600%
01THE CHALLENGE
An old CMS had flooded the index with 25,000 parameter and print URLs, while the yard's English and Russian service pages sat orphaned and rarely crawled.
Brodoremont's two dry docks take tankers, bulk carriers and Black Sea ferries for hull blasting, steelwork, propeller and shaft repairs. Most enquiries come from ship managers abroad who search in English or Russian. But a fifteen-year-old CMS had turned every sort order, session and print button into a URL of its own, and Google spent its visits on those instead of on the 22 service pages that win the work.
410 real pages out of 25,410: the sliver on the right.
Bulgarian · English · Russian
02SURVEY
Parameter URLs
14,200
Sort orders, session IDs and filter combinations on every listing, each one a crawlable, indexable URL with its own self-referencing canonical. Most were near-copies of about 30 real pages.
Finding 1 of 4
Print versions
6,800
A /print/ copy of every page and news item, linked from every template. For 212 queries Google ranked the print copy instead of the original: no menu, no contact details, no quote button.
Finding 2 of 4
Orphaned service pages
38 of 44
The English and Russian service pages existed, but nothing linked to them: the language switcher led to the home page and the menus were Bulgarian only. Google found them slowly, through an XML sitemap last regenerated in 2019.
Finding 3 of 4
Crawl budget
98.4%
Ninety days of server logs: of roughly 2,800 Googlebot requests a day, almost all went to junk. An English or Russian service page was crawled about once a month, so new prices and dock dates took weeks to show.
Finding 4 of 4
Where Googlebot went · share of requests
03THE REPAIR
Built on our Signal plan: the technical engagement for sites where crawling, indexing and architecture decide what gets found.
Print versions got a noindex first, so Google could see it, and a 410 two months later; the print button became a print stylesheet. 2,900 tag pages were merged into 18 topic hubs or retired, and archive pagination was noindexed but kept crawlable until it dropped out.
Every template now declares one canonical URL. Session IDs moved to cookies and sort and filter parameters were stripped server-side with a 301. Only once they had left the index were they blocked in robots.txt: any earlier, and Google could never have seen the redirects.
A service hub per language, breadcrumbs, a related-services block and a language switcher that leads to the same page in the other language. All 38 orphaned pages now sit two clicks from the home page.
Each service exists as a bg, en and ru trio with reciprocal hreflang and an x-default to English, listed in per-language XML sitemaps that hold only canonical, indexable 200 URLs.
04LOAD THE SHIP
Each container on deck is a thousand junk URLs. Pile them on and the hull sinks below her load-line mark; take them off and she rises, and more of Google's visits reach the service pages that bring in the work.
Below her marks: 25,410 URLs in the index, 22,910 over the load line.
Where Googlebot's requests go
Junk URLs in the index
?sort=, ?sid= and filter combinations
/print/ and ?print=1 copies
one tag, one post, no content
date archives and deep pagination
Internal links to the English and Russian service pages
Modelled: crawl and impressions follow a simple model of this fictional site, not measured data.
05MONTH BY MONTH
Back above her marks in month 9
Months 1–2
A full crawl, 90 days of server logs and Search Console exports; every URL sorted into keep, merge, redirect or remove.
Months 3–5
Canonicals on every template, parameters stripped with 301s, print versions set to noindex. The index starts to drain.
Months 6–7
Service hubs in all three languages, breadcrumbs and a working language switcher: the 38 orphaned pages join the site.
Months 8–10
Hreflang trios and clean sitemaps per language; print URLs return 410 and tag pages fold into hubs. Logs checked every week.
Months 11–12
Service pages rewritten with dock dimensions, lead times and class approvals, and a quote form on every one.
06RESULTS
Indexed URLs
25,410 → 410
25,410 → 410 indexed URLs
Service-page impressions / month
+600%
4,200 → 29,400 a month
Service-page recrawl interval
32 → 1.2 days
Average across the 66 service pages
Organic quote requests / month
7 → 24
From the quote form, organic sessions
Modelled scenario for a fictional business — not a client result. Rankings and traffic are never guaranteed; see our Terms.
| Query | Language | Month 1 | Month 12 |
|---|---|---|---|
| ship repair varna | EN | #38 | #4 |
| dry dock black sea | EN | #52 | #7 |
| propeller repair bulgaria | EN | #47 | #5 |
| судоремонт варна | RU | #29 | #3 |
| ремонт судов болгария | RU | #44 | #6 |
| докование судов черное море | RU | #61 | #9 |
| кораборемонт варна | BG | #8 | #2 |
| сух док варна | BG | #11 | #3 |
Modelled average positions on Google. Organic quote requests: 7 → 24 a month.
Is an old CMS filling your index with copies while the pages that sell sit unvisited? Start with the Signal plan, or tell us what your site is carrying.