Great Reviews, Zero Rankings: How We Rebuilt a Removals Firm's Website

Brewood Removals does the one thing every local business is told will win on Google: it collects reviews. A 5.0 average from 74 Google reviews, 4.8 on Trustpilot, more again on the trade review sites. By every trust signal an owner is told to chase, it had already won. And it ranked nowhere. Not page two, nowhere: zero page-one positions for any meaningful "removals company [town]" search, after more than three years online. This is the full account of why that happened, what we found when we took the old site apart, and the rebuild we launched on 12 June 2026. It is also a case study with a missing chapter, because the ranking verdict is still months away, and we would rather show you the work than invent the result.

The Paradox: A 5.0 Rating That Ranked Nowhere

Sean Hamilton runs Brewood Removals out of Brewood in South Staffordshire, covering the West Midlands. The reviews are genuinely earned: real house and office moves, real customers, 73 five-star and a single four-star that, as you will see, we deliberately left in place. Trustpilot sits at 4.8. The business itself is sound and properly run. Sean answers every enquiry personally, day or night, and the moves are carried out by a crew led by James, who has more than fifteen years on the vans.

So this was never a quality problem. It was a discoverability problem, and a stark one. Reviews power the Google Map pack and they build trust with a customer who has already found you. What they do not do is rescue a website that Google has quietly decided is thin or duplicated. The reviews were doing their job. The website was undoing it.

One more thing made this an easy brief to accept: Sean is a repeat client. We launched his locksmith business, Lockerfella, which hit number one on Google, ChatGPT and Gemini in 21 days. He had watched a clean build work once already, and asked the obvious question. Why was the removals site, with years more history and far better reviews, invisible?

The Diagnosis: 247 URLs and 160 Doorway Pages

We inventoried the live site on 12 June 2026, before touching anything. There were 247 URLs in the sitemap and index, and the breakdown told the whole story:

• Around 160 near-identical area pages, one for practically every suburb of Birmingham, Coventry, Wolverhampton and Walsall. Same layout, same paragraphs, only the place name changed.
• 69 thin blog posts, most from 2023, with titles in the mould of "The No 1 Best Man and Van".
• Only about ten pages with any genuine differentiation between them.

This is a textbook doorway-page build: the pre-2024 local-SEO theory that more pages targeting more place names means more rankings. It stopped working years ago, and Google's spam policies name the pattern outright. The trap is that it looks productive. A site with 250 pages feels like more SEO than a site with 25. Google counts it the other way.

"Doorway abuse is when sites or pages are created to rank for specific, similar search queries. They lead users to intermediate pages that are not as useful as the final destination."

Google Search Central, spam policies for Google web search

The Search Console numbers, which we documented in full back in April, were brutal in the quiet way these things always are. Over twelve months: 180 clicks against 111,000 impressions, a 0.2% click-through rate, average position 40. And the coverage report showed 220 of 272 known URLs not indexed, an 81% rejection rate. Google was being shown the pages and declining to keep most of them.

The clincher: across all of it, we could find exactly one page that ever surfaced in search, a Cannock URL on a low-competition compound query. For a business this well-reviewed, in a region this size, that is a rounding error. The reviews had built a reputation the website was incapable of cashing in.

Why We Rebuilt Instead of Patching

The tempting option, and the one a lot of agencies would sell, is remediation: keep the 247 pages, rewrite the worst, bolt on schema, freshen the thin posts. We turned it down for a single reason. The doorway pattern itself is the liability, and you cannot improve your way out of 160 near-duplicate pages. Polish a doorway page and it is still a doorway page. The only honest move on duplication at that scale is removal.

So we proposed total replacement. The 247 old URLs become about 25 new ones: ten real area pages, around eight service pages, and the core. Fewer pages is not the goal in itself; fewer deeper pages is. Each area page is grounded in real, dated jobs from Sean's own diary, with the surrounding villages rolled up into the town that actually ranks for them, instead of getting a thin page of their own. This is the lesson of Google's helpful-content work spelled out in a sitemap: ten pages that could only have been written by this business beat 160 that could have been written for anyone.

"Are you producing lots of content on many different topics in hopes that some of it might perform well in search results?"

Google Search Central, a self-assessment question from Creating helpful, reliable, people-first content

410 Gone, Not 301: The Redirect Decision

This is the part most rebuilds get wrong. The reflex when you delete pages is to 301-redirect every one of them to the nearest survivor, to "keep the SEO value". With doorway pages, that reflex is actively harmful.

A 301 from a page Google distrusts does not hand your new page clean authority. It hands it the association. Point 160 pages Google considers spam at your shiny new Wolverhampton page and you risk importing the exact signal you are trying to escape. So we mapped every old URL deliberately rather than spraying redirects:

• 225 of the 247 URLs return a 410 Gone. That status tells Google the page is intentionally and permanently removed, with no replacement implied. It is the cleanest possible way to say "this was a mistake, forget it ever existed".
• 18 get a 301 redirect, used only where a genuine equivalent page exists on the new site.
• 3 keep their URLs outright.

The full disposition map lives in the site's .htaccess, so every one of those calls is auditable. This route is slower and less comfortable than a wall of redirects, and it is the right call. You do not launder a doorway page by pointing it somewhere nicer.

What We Built: 25 Deep Pages, Schema, a Moving Planner

One disclosure first, held to the same standard as our editorial standards: the new site is hand-coded PHP, not WordPress. We chose that to hit aggressive Core Web Vitals targets on a budget. Every other pattern here, the content model, the schema, the AI Discovery Files, applies directly to the WordPress builds we ship every week. Where a client needs a CMS, we build WordPress and engineer the same disciplines into it. Brewood did not need one.

The build, in the order of what matters:

• Exactly ten area pages, a hard rule set against the old 160, covering the patches the business actually works: Wolverhampton, Walsall, Stafford, Cannock, Birmingham, Coventry, Dudley, Sandwell, Solihull and Shrewsbury. Each is built from real dated jobs, carries its own per-town Service schema, and absorbs its surrounding villages rather than spawning a thin page for each.
• A service cluster of around eight pages, each with real job cards, service-relevant reviews and its own FAQs: house removals, office removals, man and van, packing, storage, piano removals, student moves and house clearances.
• An advice cluster of long-form guides, each one passing the same E-E-A-T audit we hold our own posts to: a visible byline, published and "editorially reviewed by Sean Hamilton" dates, an author-bio box linking to a public editorial-standards page, a table of contents wired into the page's schema so the visible structure and the markup cannot drift, and attributed expert quotes with live source links.
• A reviews page rendering all 74 Google reviews verbatim, with Sean's replies and self-hosted avatars so the page makes no calls to Google at runtime. The lone four-star review stays. We never curate.

Performance was treated as a launch gate. The mobile PageSpeed Insights run on 14 June 2026 returns 100 for Performance, Accessibility, Best Practices and SEO, with First Contentful Paint at 1.1s, Largest Contentful Paint at 1.2s, Total Blocking Time at 0ms and Cumulative Layout Shift at 0. That last one mattered most on a content-heavy site: we cut layout shift from 0.44 to roughly 0.02 with a pre-paint class swap. The site runs on 365i's standard Linux hosting with the bundled CDN, and you can re-run the test yourself.

Underneath the visible site sits a single data layer. The words a human reads, the schema a search engine parses, and the nine AI Discovery Files an AI assistant reads are all rendered from the same source, so they physically cannot say different things. A full schema graph ships on every page type, from the per-town Service entities to the reviews.

And one interactive tool, the Moving Planner. Enter your move date and a few details and it builds a personalised, tickable, week-by-week countdown, saved privately in your own browser, printable, with an optional "email me my plan". It carries no prices anywhere, deliberately. It earns its place three ways at once: it captures leads, it hosts dozens of contextual internal links, and it surfaces real customer stories on the pages around it. One well-built tool doing the work of a dozen thin pages is the whole philosophy of the rebuild in miniature.

131 Real Customer Stories, Without Faking Reviews

This is the part I am most pleased with, because it is where most "E-E-A-T content at scale" projects quietly cross a line.

Sean handed us three banks of real material from years of work: 40 dated area jobs, 40 service jobs, and 51 testimonials from customers who had used the planner on the old site. That is 131 genuine, attributable cases, the kind of first-hand detail Google's guidelines openly reward. The temptation with material like that is enormous: mark it all up as Review schema with five-star ratings, and earn yourself rich review stars in the search results.

That would be review fraud. So the rule on this build is absolute. Customer stories appear as visible copy, attributed by family surname only, and never as Review or AggregateRating schema, never with a star rating attached. The only star ratings anywhere on the entire site come from the verified Google reviews cache. Three automated integrity checks enforce it at build time, including one that asserts the number of visible review cards on a page equals the number of Review schema nodes, so the two can never quietly diverge.

The distinction is simple and worth stating plainly, because a lot of agencies blur it: you can use real customer experience on every page you like. You cannot dress it up as something it is not. Stories are content. Stars are a claim, and a claim has to be verifiable.

What We Will Not Claim Yet: The Rankings

Here is the missing chapter, stated plainly. The new site went live at 21:00 on 12 June 2026. At the time of writing it is two days old. We have no ranking recovery to show you, because there is not one yet, and there could not be.

Google has to crawl and process 225 410-Gone responses, index the new 25-page set, and re-evaluate a domain that has spent three years training it to expect thin content. That takes weeks to months, and a core update landing somewhere in the middle would surprise nobody. Anyone who promises you a recovery curve two days after a rebuild is either guessing or selling.

What we can show you is real and verifiable today: the live site, the schema graph on every page, the nine AI Discovery endpoints, the PageSpeed scores, and the working planner. The strategy is sound and the build is done to standard. The proof of ranking impact is the one thing that can only be earned with time.

So we will update this page with Search Console figures when they are real: indexation of the new set, processing of the 410s, and movement on the ten target towns. If the strategy underperforms, that will be on this page too. A case study you can only half-prove today is worth more than one where you had to invent the ending.

Great Reviews, No Rankings? What to Check

If your business has glowing reviews and no rankings, the website is almost always the culprit, and it is usually one of two faults. Either it is too thin, with not enough genuine substance for Google to rank, or it is too duplicated, with pages so similar that Google indexes a handful and bins the rest. Doorway pages are the second fault at industrial scale, and they are common in trades because someone, somewhere, is still selling "a page for every town" as a package.

The fix is not glamorous. Fewer, deeper pages written from your real jobs. Honest removal of the junk with 410s, not redirect-spraying. Schema and performance enforced as budgets before launch, not patched after. Customer stories used as content, never faked as review markup. It is the same playbook that took Sean's other business to number one in three weeks, and the same discipline behind this very site's PageSpeed 100.

If you recognised your own site in the first sentence of this section, that is the conversation worth having. We will tell you honestly whether it is a content problem, a technical one, or both, before you spend a penny on a rebuild.

Frequently Asked Questions