62%
Share of Googlebot visits spent on filter URLs. Colour, shape, brand, price and sort combined into 41,000 crawlable addresses, most of them the same twelve frames in a different order.
Finding 1 of 4
A fictional business. The figures are a modelled scenario, not a client result.
Wychelm Opticians · 14-practice optician chain · Birmingham
14 of 14 practices indexed, +71% eye tests booked
10 months · UK & Ireland · Health & wellness · Signal plan
6/6 is how UK optometrists write normal vision: you read at six metres what a typical eye reads at six metres.
01The challenge
Cloned branch pages, booking on an uncrawlable subdomain and frame-filter URLs eating the crawl budget pushed 9 of 14 practices out of Google's index.
Wychelm grew one practice at a time, and every new branch page was cloned from the last: the same introduction, the same list of services, a different address. Google treated nine of them as duplicates and folded them into the other five. Booking lived on a separate subdomain that rendered in the browser and shut crawlers out, so no practice had a booking page Google could reach. Meanwhile the frames catalogue spun out tens of thousands of filter combinations, and Googlebot spent most of its visits there.
The practices, before · one letter per practice
9
9 of 14 practice pages missing from Google
02The examination
62%
Share of Googlebot visits spent on filter URLs. Colour, shape, brand, price and sort combined into 41,000 crawlable addresses, most of them the same twelve frames in a different order.
Finding 1 of 4
88%
Text similarity between branch pages. Google kept five practices as canonical and folded the other nine into them, so a search for Erdington or Kings Heath landed on Moseley.
Finding 2 of 4
0
Booking pages Google could crawl. The booking app ran on its own subdomain, rendered entirely in the browser and blocked by its own robots.txt; the Book buttons were scripts, not links.
Finding 3 of 4
5 clicks
From the home page to a typical practice page, reachable only through a store-locator widget. Each practice had, on average, a single internal link pointing at it.
Finding 4 of 4
03The prescription
Built on our Signal plan: the technical engagement for sites where crawling, rendering and architecture decide who gets found.
robots.txt rules for sort, session and multi-select filter parameters, while single-facet pages such as round or tortoiseshell frames stay crawlable and linked. Sitemaps trimmed to canonical URLs; server logs checked every week.
Filter URLs canonicalised to their category. Every practice page rewritten from the practice itself: its optometrists, equipment, parking, opening hours and the neighbourhoods it serves, with a self-referencing canonical and its own Optician schema.
A practice hub at /opticians/, every practice linked from the main navigation, plus links between neighbouring practices and from service pages such as children's eye tests and contact lens fitting.
Booking moved off the subdomain onto server-rendered /book/ pages on the main domain, linked with real anchors, so each practice page and its booking page reinforce each other.
04The duochrome crawl test
In an eye test, the optometrist shows you letters on red and on green. If one side looks sharper, the prescription is off; when both read the same, it is right. Here the red half is the frame-filter URLs and the green half is the practice pages. Adjust the three settings until neither half stands out.
Crawl acuity
6/60
Both halves are blurred: Googlebot is wading through filter URLs and cannot see the practices. Filter URLs: 62% of visits. Practice pages: 8% of visits, 5 of 14 indexed.
Crawl budget is how many pages Googlebot is willing to fetch from a site in a given period. Every visit spent on a filter combination is a visit not spent re-reading a practice page.
A modelled illustration for a fictional site: real crawl budgets depend on a site's size, speed and popularity.
05Month by month
Month 1
A full crawl, twelve weeks of server logs, similarity scores for all fourteen branch pages and a map of every route into booking.
Months 2–3
robots.txt rules and filter canonicals shipped; sitemaps cut from 44,000 URLs to 1,900. Filter URLs fell from 62% to 34% of Googlebot visits.
Months 4–5
Every practice page rewritten with its own team, equipment, parking and neighbourhood, and checked for similarity before it went live.
Months 6–7
Booking rebuilt as server-rendered /book/ pages on the main domain; the old subdomain now redirects there, one URL at a time.
Months 8–10
Practice hub, neighbour links and service-page links. Practice pages recrawled every four days, and the last dropped practice back in the index in month 10.
06Results
Modelled scenario for a fictional business — not a client result. Rankings and traffic are never guaranteed; see our Terms.
| Search | Landing page | Before | Month 10 |
|---|---|---|---|
| opticians moseley | /opticians/moseley/ | #6 | #2 |
| eye test harborne | /opticians/harborne/ | #11 | #1 |
| opticians erdington | /opticians/erdington/ | not in the top 50 | #3 |
| kids eye test kings heath | /opticians/kings-heath/ | not in the top 50 | #2 |
| contact lens fitting solihull | /opticians/solihull/ | #9 | #2 |
| opticians jewellery quarter | /opticians/jewellery-quarter/ | not in the top 50 | #1 |
| independent opticians birmingham | /opticians/ | #18 | #4 |
Modelled Google UK positions for each query as typed, before the work and in month 10.
Branch pages slipping out of the index, a booking system Google cannot see, or a catalogue eating the crawl? Start with the Signal plan, or tell us what is blurred.