Lincolnshire Chauffeur Services: A 23-Page Flagship Built on Honest Data

The new site for Lincolnshire Chauffeur Services went live on Tuesday 21 July 2026: 23 pages of executive airport-transfer content, a split-flap departures board, nine individually written town pages, and a mobile PageSpeed score of 99 with every Core Web Vital green. The day after launch we ran a keyword pass on the town pages and published this study. What makes the build worth writing up is not any of those headline numbers, though. It is the two honesty decisions at the heart of it: journey times published as labelled estimates rather than invented facts, and a deliberate choice to bank 99 rather than chase a vanity 100. And because the site has been live for one day, this case study claims precisely nothing about rankings. That chapter does not exist yet, and we would rather show you the build than invent the ending.

The Brief: A Chauffeur Firm Trading Since 2006

Bart Griemink runs Lincolnshire Chauffeur Services from Bicker, a village in the south of the county. The business does pre-booked, chauffeur-driven executive transport, Lincolnshire to anywhere in the UK, and it has been doing it since 1 September 2006. It holds British Chauffeurs Guild permit 2722CT. Those two facts, nearly twenty years of trading and a Guild permit, are the spine of the whole site: this is a business whose product is reliability, sold to people who cannot afford a no-show at 3am. The about page states both plainly, because "trading since 2006" beats "years of trusted experience" every time a sceptical customer reads it.

The old WordPress site was not doing that business justice. Every one of its 17 pages had an empty meta description. Its location pages were near-identical boilerplate; three of them were word-for-word the same apart from the town name. The About page, on a business whose entire proposition is trust, ran to roughly 140 words. And its own locations hub linked to a Bourne page that had never actually been built, so the link had been quietly returning a 404. None of this is unusual for a site of its age. All of it is fixable, and fixing it properly meant a rebuild.

We build for this sector deliberately: chauffeur services have their own page in our portfolio, and the pattern that repeats across transport clients is that the website's job is certainty. The passenger has already decided to pay more than a taxi fare. What they are buying is the confidence that the car will be there.

One ground rule set before any design work: URL preservation. Every URL on the outgoing site resolves on the new one. The eight old location pages 301-redirect to their new homes under the locations hub, and the retired blog returns 410 Gone rather than being redirect-sprayed at pages it has nothing to do with, the same discipline we applied on the Brewood Removals rebuild. Nothing the old site had earned was thrown away, and nothing it should not have had was carried forward.

The Design: Certainty, Not Glamour, on a Split-Flap Board

Almost every chauffeur website sells glamour: gold on black, champagne flutes, a man in a peaked cap. We think glamour is the wrong product. Nobody in Woodhall Spa books a chauffeur because they want to sit in a nice car. They book because their flight leaves Heathrow at 07:15, and the alternative is driving the A1 at three in the morning and paying two weeks of airport parking. The design thesis, agreed with the client before a pixel was drawn, was to sell calm and precision instead.

The palette comes from the material world of an executive car before dawn: blue-cast charcoal rather than black, the warm amber of an interior ambient light strip, cold pre-dawn windscreen blue. As on the Cordley Solutions build, every colour pairing was measured before use rather than eyeballed: primary text on the night background sits at 11.09:1, the amber accent at 7.88:1, both past WCAG's stricter AAA threshold. The call-to-action buttons are dark text on amber, not white on amber, because white on amber fails AA. It also happens to read more expensive.

Type is three self-hosted variable fonts totalling 114.8 KB: Switzer for display, Source Serif 4 for body text, and Martian Mono for anything that is data, times, durations, flight numbers, airport codes. That third face powers the site's signature device: a recurring instrument-panel language borrowed from the departures board and the dashboard, the two objects this audience actually stares at.

The instrument language repeats with a different job on every page type. The home page has the split-flap departures board cycling the nine towns. Each town page carries a destination board that flips through the airports and seaports the business actually serves. The airport and seaport transfers page, the run the business does most, adds a live-flight-monitoring instrument, because tracking the inbound flight and adjusting the pickup is a genuine part of the service. The services hub presents its five services as a numbered board index. And the quote form carries a live "Your journey" summary panel that fills in as the customer types.

Full disclosure, in line with our editorial standards: the new site is hand-coded PHP, not WordPress. Twenty-three pages, no CMS requirement, aggressive performance targets; hand-coding was the honest tool, exactly as it was on Brewood and Cordley. The design system, content model and schema patterns transfer directly to the WordPress design work that makes up most of what we do.

Nine Town Pages of Genuinely Local Airport-Transfer Content

The commercial heart of the site is nine town pages, one for each place the business genuinely serves: airport transfers from Boston, executive travel from Grantham, airport transfers from Lincoln, airport runs from Sleaford, transfers from Spalding, the Skegness coast run, early starts from Horncastle, Woodhall Spa and a Bourne page with a story of its own, all under one locations hub.

Nine location pages is exactly the shape that goes wrong. The old site proved it: its location pages were interchangeable, and interchangeable location pages at scale are doorway pages, the pattern that buried Brewood Removals under 160 of them. So the rebuild enforces one hard rule, borrowed from our work on non-commodity content: if a page could have its town name swapped and still read correctly, it is not finished.

"Does the content provide original information, reporting, research, or analysis?"

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

In practice that means every town's copy is grounded in its actual geography. Lincoln's page talks about the A46 pointing straight at the A1 and the ring of bypasses around the city. The coastal pages deal with the long cross-county reality of starting a Heathrow run from the seaside. Each town gets its own routes, its own pickup considerations, and its own destination board flipping through ten airports and seaports with per-town drive times. One validated template renders all nine, but the template carries none of the words; the words are written per town, which is the entire difference between a local landing page and a doorway page.

One of the nine deserves its own sentence. The old site's hub and homepage both linked a Bourne page that was never built, so for years that link 404'd. The new site closes the loop: Bourne got a real, individually written ninth page, and the old dead path 301s to it so any external link that ever pointed there finally lands somewhere useful.

The Honesty Layer: Journey Times Published as Estimates

Here is the section that makes this build worth studying rather than just admiring. The journey clock and the destination boards run on drive times, and at launch the client had not yet verified every figure from his own experience. The tempting move is obvious: publish the desk research as fact, because "Lincoln to Heathrow, 3h 30m" looks more authoritative than an estimate. We did the opposite, in three layers.

• On the page, every time is labelled an example: "Approximate driving time, with a margin built in for traffic. Your quote confirms the exact pickup time for your flight." The board headers say "example drive times" in so many words.
• In the figures themselves, each estimate carries a deliberate safety margin rather than a best-case number, because an optimistic drive time is the most expensive kind of wrong.
• In the data layer, every journey time is flagged as unverified until the owner confirms it. As he corrects each figure from real runs, the flag flips, and the disclaimer will eventually retire itself.

Why bother, on numbers a competitor would happily invent? Because the audience for a chauffeur site includes people who do these exact runs every month. A customer who knows the A1 will spot a fantasy drive time instantly, and once one number on an instrument panel is wrong, every number on it is. The same honesty governs pricing: the business quotes each journey individually, so the site publishes no fares at all, and the quote form is the only route to a price. It gives up price-intent queries on purpose and competes on trust and route specificity instead.

The same register runs through the copy. The site states "trading since 2006" and the Guild permit number instead of reaching for superlatives, and it never once calls the business a taxi, while still answering the "is this an airport taxi?" question a real searcher asks. That taxi-versus-chauffeur distinction gets its own section further down, because it drives the whole keyword strategy. First, the performance story.

PageSpeed 99, Every Vital Green, and Why Not 100

Performance was treated as a launch gate, as it is on everything we ship, including our own site. The launch-day mobile Lighthouse run came in at 92, respectable and not good enough. Three fixes took it to 99, and the sequence is a small case study in engineering trade-offs.

First, the responsive hero. The homepage hero was a single 286 KB image served at 2560px to every device, which meant a phone was downloading a desktop-sized file as its largest paint. We generated three proportional variants and served them by viewport, so a phone now fetches 37 KB instead of 286 KB. Largest Contentful Paint dropped from 3.2 seconds to 1.4.

Second, the win exposed a new fault: with the hero arriving that fast, the three custom fonts were now swapping in visibly, causing a layout shift of 0.097. Preloading the font files fixed it, layout shift back to a flat zero. But the preloads compete with the hero image for early bandwidth, which nudged LCP from 1.4 seconds to 2.0. That is the trade: perfect layout stability cost 0.6 seconds of paint time.

"To provide a good user experience, sites should strive to have Largest Contentful Paint of 2.5 seconds or less."

web.dev, Largest Contentful Paint (LCP), Google

The final numbers: First Contentful Paint 0.9s, Largest Contentful Paint 2.0s, Cumulative Layout Shift 0, Total Blocking Time 0ms, Speed Index 1.2s. Score: 99. Every vital comfortably inside Google's thresholds, and one point on the table if we wanted to trade the zero layout shift back for a fraction of LCP. We decided not to. A stable page at 99 serves the user better than a twitchy one at 100, and Google measures vitals, not vanity. The decision is recorded in the project log with the date on it, and you can re-run the test yourself. The site runs on 365i's standard Linux hosting with the bundled CDN; the same discipline drives our WordPress speed optimisation work, where the budget is the same even when the stack is heavier.

The Keyword Strategy: Taxi in the URLs, Chauffeur in the Brand

The search reality for this business is blunt: people do not search for "chauffeur-driven executive travel". They search for "airport transfers Lincoln", "airport taxi Boston", "executive airport taxi". The brand reality is equally blunt: this is not a taxi firm, and describing it as one would be both wrong and self-defeating. The site resolves the tension deliberately. The URLs say airport-taxis- because that is what customers type and those slugs carry the site's authority. The copy positions pre-booked executive chauffeur hire, and an FAQ on every town page answers "Is this a Lincoln airport taxi?" in the searcher's exact words while explaining what the service actually is. Both halves are on purpose, and neither may be "fixed" into the other.

The queries this site needs to win, in order: "airport transfers [town]" across the nine towns, "executive airport taxi" and its variants, and "chauffeur service Lincolnshire". Generic head terms like "taxi Lincoln" are conceded to actual taxi firms and the aggregators; a quote-only executive service pretending to be a cheap cab would attract exactly the wrong clicks and convert none of them.

Then we checked our own work, and this is the part worth stealing. The day after launch, a programmatic audit of all nine rendered town pages found the target phrasing thinner than it should have been: "airport transfer(s)" appeared just 2 to 4 times per roughly 1,400-word page, about 0.2% density, "executive" twice, and on every town page the H1 was the only heading carrying the target phrase. Every other H2 was rhetorical: well written, and invisible to search intent.

The fix went in at template level on 22 July, so one edit reached all nine towns. The board heading became "[Town] airport transfers, every terminal.", which supplies the exact word order of the search query that the H1's "Airport transfers from [Town]" phrasing did not. The local-knowledge section became "The local part of your airport transfer." And one "airport transfer" and one "executive car" were woven into each town's unique prose, with distinct phrasing per town so the nine pages stayed individually written. After the pass: "airport transfer(s)" 6 to 9 times per page, around 0.5% density, "executive" 4, "airport taxi" 3, the latter growing only through the FAQ negation and navigational anchors, never a claim. Single clean occurrences, no stuffing, and every count re-checkable on the rendered pages; the Lincoln town page is the one we audited line by line, and the template-level fix reached all nine towns at once.

Whether it works is a question only Search Console can answer, and it has had one day. But the method stands regardless: write for the reader first, then audit the rendered pages against the queries you actually need, and close the gaps surgically. Keyword coverage you assume is keyword coverage you usually do not have.

AI Discovery Files From Day One

The full set of AI Discovery Files went live at the site root on launch day, alongside a standard security.txt: llms.txt, a human-readable llms.html, ai.txt, ai.json, identity.json, brand.txt, faq-ai.txt, developer-ai.txt and robots-ai.txt, implementing the AI Discovery Files specification. The llms.html file is deliberately indexable, a designed page in the site's own brand rather than a bare text dump.

The reason to do this before any rankings exist is the same one we set out on the Cordley Solutions launch: AI assistants do not wait for rankings. When someone asks ChatGPT or Gemini about airport transfers in Lincolnshire, the assistant answers from whatever machine-readable material it can find, and a site's own structured, honest, first-party files are the strongest signal a small business can put in front of it. The files also encode the restraint: they carry the "never a taxi, never a minicab" positioning and the quote-only pricing policy, so an AI describing the business cannot honestly get those facts wrong. We watched this groundwork pay off within three weeks on the Lockerfella build, which hit number one on ChatGPT and Gemini in 21 days.

And in keeping with a study built on re-checkable claims, this one takes a minute to verify: run the domain through the free AI Visibility Checker, or open the site's public AI visibility profile, where its verified discovery-file status is listed.

What We Can't Claim Yet: Rankings and Traffic

Our case studies carry this section by policy, and this one carries it further than any before it. The site went live on 21 July 2026. This study was published on 22 July 2026. The site has been live for one day, which means we claim no rankings, no traffic, no leads, no AI citations, and no movement on any of the keyword targets described above. Not even the brand query. One day is not a data set; it is a launch date.

We publish the study now anyway, because the decisions are at their most honest before the results exist to flatter them. The estimates-not-facts call, the 99-not-100 call, the keyword pass with its before-and-after counts: all of it is on the record, dated, ahead of any outcome. When real Search Console data lands, for the town queries, the brand query and the "airport transfers" cluster, this page will be updated with the figures and the update will be logged, whichever way they point. If the strategy underperforms, that goes on this page too. It is the same standard we set on Brewood and Cordley, and it is the house style now.

What is verifiable today: the live site with all 23 pages, the labelled journey estimates, the PageSpeed figures, the discovery files at the root, and the build dates in the project's version history. That is the evidence this study rests on, and every reader is welcome to check it.

Frequently Asked Questions