Skip to main content
0 likes, 0 dislikes

This is a real case study, with real Search Console figures, for a real Coventry business: Mirage Body Shop, an accident repair and paint shop on Bodmin Road, CV2. We built their website in spring 2025, the site then ran for most of a year with no SEO work at all, and in 2026 the owner asked us back to fix it. The numbers below are read straight from Google Search Console for miragebodyshop.co.uk on 23 August 2026. Where a figure is an estimate or has a caveat, it says so.

The timeline, honestly

  • April to May 2025 — we built the original site: a custom PHP MVC build, bespoke theme, a database shaped around a body shop's workflow. No template, no page builder.
  • June 2025 to January 2026 — no maintenance and no SEO, by the customer's own decision. Nothing was published, nothing was optimised, nothing was measured.
  • February 2026 — the owner got back in touch. Search Console for that month shows 17 clicks from 655 impressions. For a business that has been online for nine months, that is close to invisible.
  • February to April 2026 — code first. We upgraded the platform (the MVC core, theme system, caching, the SEO head and schema layer) so that every later change would land on a site that could carry it. Search work done on a broken platform is wasted work.
  • From April 2026 — on-page SEO: metadata, headings, internal links, service pages, a blog programme, local signals, and the corrections listed below.
  • August 2026 onwards — off-page: directory citations, followed links from real business directories, social profiles, and the Google Business Profile. This part is still running as we write.

The result so far, month by month

Google Search Console, property https://miragebodyshop.co.uk/, all queries, all devices, UK. August is 1 to 21 August only.

Month (2026)ClicksImpressionsAverage position
February1765516.1
March533,10014.1
April734,91014.8
May14311,30019.5
June20718,40012.6
July27729,50013.5
August (to the 21st)25934,80013.4

From February to July that is sixteen times the clicks and forty-five times the impressions. Over the last sixteen months the property shows 1,020 clicks and 101,000 impressions in total; almost all of that happened after April. Average position has stayed in the 12 to 14 range while impressions multiplied, which is the pattern you want: the site is appearing for far more searches, and holding its place rather than sliding down as the query set widens. The May dip to 19.5 is what happens when a batch of new pages starts appearing for queries it has not earned yet.

Two caveats. Clicks are the metric that matters and they are still modest; a local body shop does not need thousands of visitors, it needs the right few hundred. And click-through rate has fallen from 2.6% to 0.7% as impressions grew, because a site that now shows for a thousand different queries is shown for many it will never win. That is normal and we say it out loud rather than hide it.

What the queries say

In the 28 days to 13 August, the queries bringing clicks were the ones a body shop actually wants: car body shop near me (average position 9.5), bodyshop near me (7.1), car body shop (3.3), body shop repairs near me (2.0), plus the brand itself. A single blog post, on the difference between a light scratch and a deep one, is the site's second-biggest page by impressions at around 3,400 in its period, ranking in the top ten. That post exists because the question is one customers ask at the counter, not because a keyword tool suggested it.

One more figure worth having: Search Console's generative AI report showed 6,180 impressions in three months for the site by mid-August, and when we tested the AI engines on 21 August the business was named first for street-level questions such as "car body repair Wyken Coventry" and "bodyshop Coventry Bodmin Road". It is still absent on the city-level head terms. That is the next fight, and it is a links fight, not a content one.

What we actually did

Not a list of services; the specific work, in the order it happened.

  1. Platform first. The site moved onto the current version of our hand-built PHP MVC platform (live on 13 August 2026 after months of staged work), with a theme system, proper caching, and an SEO head that every theme shares. Until that landed, three stacked defects meant every theme's head silently dropped the Google tag, Open Graph tags and the custom head code. We found it by fetching live pages and counting: zero og tags, zero gtag. Fixed across all eleven themes.
  2. Headings that are headings. Thirty-six theme templates were rendering the content's H2 to H6 as paragraphs with a class. To a reader it looked fine; to Google the page had one heading. Fixed in every template.
  3. Correct business facts. The live site said the shop was "based in Binley (CV3)". It is on Star Industrial Park, Bodmin Road, CV2 5DB. Wrong address in the metadata, the intro copy and five blog posts — corrected everywhere, and matched to the Google Business Profile. Schema changed from the wrong AutoRepair type to AutoBodyShop with the address split properly, opening hours, service area and a service catalogue.
  4. A fabricated rating, removed. A previous developer had left an invented 4.8-from-127 aggregate rating in the structured data. The real Google listing was 5.0 from 44. We removed the fake one and did not replace it with anything; reviews live on Google, not in our markup.
  5. Service pages with an H1. Eighteen service URLs had no H1 at all. Every one now has one, and the six area "hub" pages, which were six near-identical pages carrying 584 impressions and zero clicks between them, were rewritten to be genuinely different, ring by ring outward from the shop.
  6. The estimator page. It was the sixth-largest page by impressions (1,762) with a title that contained none of the words people searched, 154 words of content, four clicks and a 0.23% click-through rate. Retitled Auto Body Repair Estimate Calculator, grown to 455 words with five internal links, and eight panel prices corrected against real quotes. We do not yet have an "after" reading for that page; it will be in the next update of this article.
  7. Content at scale, but not slop. 212 of 218 blog posts were rewritten against a scoring tool in one supervised run, every post given a brief, internal links and a real question to answer. The six that failed timed out at the CDN, not on content. We will not pretend that makes 218 great articles; it makes 218 articles that are no longer embarrassing, which is the floor you need before any of them can rank.
  8. Off-page, measured first. Before touching links we measured: Ahrefs DR 22, 17 live referring domains, of which 13 were flagged as spam and two were our own sites — so three or four legitimate third-party links in total. A disavow file went to Google on 21 April. The rival body shops' "500 referring domains" turned out to be mostly one free newspaper-directory network, which changed the plan from "build hundreds of links" to "win about twenty real citations". Three followed links landed on 21 August from directories with DR 81, 75 and 73, each verified by fetching the public page and reading the link's attributes. Google reviews went from 42 to 45 over the same period without a review campaign.

What we will not claim

  • We do not have a map-pack position to quote. We have not measured it in a way we would stand behind, so it is not here.
  • The "before" figures start in February 2026, the earliest month the Search Console property holds data for. We cannot show you the 2025 numbers because nobody was measuring them, which is itself part of the story.
  • None of this is a ranking guarantee. A body shop in Coventry competes with businesses that have five to ten times its review count. What we can show is the direction and the rate, with the dates attached.

What this tells a business owner

A well-built website that nobody maintains does not hold its value; it sits still while everyone around it moves. Mirage's site was technically sound for the whole of its quiet year and it earned 17 clicks in the month the owner came back. The recovery did not come from a trick. It came from fixing the platform, telling Google the truth about the business, giving every page one job, and then measuring every month. If you want the same process applied to your own site, start with our SEO service, read the full Mirage story, or, if you are local, see what we do for businesses in Coventry.

Figures last updated 23 August 2026. We will refresh this article as the off-page work lands and the August numbers close.

Comments (0)

No comments yet. Be the first to comment!

Leave your thought

Your comment will be moderated before being published.