9 times the traffic for a UK car finance site, won on the clusters competitors left alone

Starting point — a 1-year-old site with 90% of its traffic gone

I picked this project up in April 2021. The broker works the UK market only, and the whole proposition is that buying a car on credit stops being the borrower’s problem: they submit one application and a personal manager carries the rest.

• Selects the most favourable car finance programs available to that borrower.
• Sends the borrower's details to the banks for assessment and a decision.
• Helps choose the car itself, new or used, from a database running to hundreds of offers.
• Prepares the paperwork for what is a 3-way transaction between borrower, lender and dealer.

The site arrived damaged. A previous contractor had run exactly one play, cross-linking from their own network of sites, and a Google update caught it. 90% of the traffic went, and leads fell by a factor of 10.

Diagnosis — why the head terms were the wrong target

Car finance is a brutal niche. Page one belongs to banks and to platforms sitting on their own vehicle inventory, and on budget alone I was never going to outspend them. What search still had going for it is cost per lead, which beats advertising in this market by a distance. The client set me 2 goals: more traffic, more organic leads.

The client wanted the high-volume, high-competition terms, which is what every client wants. I went through the results, the competitors and the converting queries, and came back with the opposite answer: chasing them would spend the whole budget and produce nothing. 3 reasons.

• The site was 1 year old; the competitors had 5 or more.
• Positions in the top-10 were contested hard. Even the sites sitting at 9 and 10 had spent 2 years on links and content to get there.
• Individual pages could scrape into the top-10 while the rest stayed far off, which produces too little traffic to matter.

Rather than argue the point, I put together what the competitors were spending and what displacing them would cost, and put the arithmetic in front of the client. The head terms were a bad bet at this stage. But there was another way in.

Strategy — take the clusters the market left alone

So where do you go when the front door is bolted? Round the back. Car finance semantics run far wider than the obvious terms, and buying intent sits right across that long tail. The competitors were working the head and ignoring the rest. In Ahrefs, Semrush and Rush Analytics I:

• Stripped out off-topic and incorrect queries.
• Clustered what remained in Rush Analytics, cleaning repeatedly rather than once.
• Identified which competitors to analyze.
• Studied their indicators and calculated how long a top placement would take.

I then filtered the clusters against 3 conditions, and all 3 had to hold at once:

That gave me a list of low-competition pages to build for, and the client signed it off. I ordered the work by how fast each part pays back: commercial-intent pages first, informational blog content after, SEO copy and internal linking alongside both, then the linking and anchor strategy.

Work — 90% of the commercial pages built from scratch

The site's structure was narrow, so most of what the strategy needed simply didn't exist yet. I built roughly 90% of the commercial pages from scratch, plus an FAQ section and the blog.

Every page went through the same pipeline. I designed the layout with all its blocks and elements, wrote the development specification for the programmers, and prepared the text brief in Rush Analytics, accounting for full, partial and diluted keyword placement.

What a cluster page contained

Take the page for young drivers. I put 4 things on it: messaging around the lowest-cost credit available at that age, a calculator for estimating the monthly payment, an explanation of why credit is harder to get when you're young and what the service does about it, and answers to the questions that come up.

Scaling from 6 articles a month to 13

I took production from 5 or 6 articles a month to 12 or 13, and brought in English-language copywriters I had already tested on other projects.

At first the writers produced everything by hand, which stops scaling the moment you need 13 articles a month on top of the landing pages. So I split the process: copywriters took the material that mattered most, and I scaled the template pages against the page structures already ranking in the top-10, standardized brief templates, mass AI-generated text, and a quality check with revisions on the way out. That split is what let template landing pages keep growing month over month.

Work — outreach links, and the ones the blog earned

I picked the weakest of the top competitors as the ones to challenge and took their link strategies apart: what types of links they held, how those spread across sites, how they had grown by year and by month, and what percentage of anchors carried keywords.

The strategy I built on that ran mainly on outreach: I picked the donor sites and negotiated every placement myself. UK links are expensive and a lot of what gets offered is spam dressed up as a media outlet, so selection mattered far more than count.

One advantage compounded on its own, with no budget line behind it. Blog content that answered the audience's questions properly got republished by other sites, with active links back. Earning links rather than buying them is rare in the UK market, and it came straight out of the quality of the top-performing pages.

Early signals — which clusters moved first

My hypotheses held. Pages built on low-competition, traffic-generating queries reached the top-10 within months, then the top-5, and some of them the top-2 and top-1. Leads started arriving.

3 commercial groups moved fastest:

• Car finance for young people.
• Lending for privileged categories.
• Vehicle leasing.

Blog queries began converting into leads within months of launch, which was earlier than informational traffic usually pays back.

Holding — through 10 algorithm updates in 2022

The previous contractor's positions had vanished completely at the first algorithm update, which is what prompted the change of agency. This time the positions came back and then held.

Google shipped 10 updates during 2022. Daily traffic sometimes fell 15–20%, and recovered within 1 to 2 months each time.

The discipline is the part that transfers, and mostly it consists of sitting on your hands. After an update I changed nothing in reaction to it. Where a site meets Google's requirements the traffic comes back on its own, and it did: the weekly trend turned positive again every single time. Left unoptimized, that run of updates would have finished off whatever traffic the site had left.

Result — 9 times the traffic in 22 months

In absolute terms that is roughly 500 queries in the top-3 and over 1,500 in positions 4 to 10. By October 2022 more than half the promoted set was in the top-10, and virtually all of it was in the top-100.

April 2020 to April 2021 is left out: it is the year before the engagement, and both series sit flat on the axis through it.

Tracking runs back to April 2020, a year before the engagement began, and that year is the useful control: queries in the top-3 moved between 0 and 48 and stayed there. The table starts in May 2021, once the work was underway.

Where the traffic sat, by section

The blog is the number worth pausing on: launched from nothing in February 2022, it accounts for 37% of every click the site earned across the period, and it ranks better on average than the site as a whole — position 12.0 against 14.6.

2 measurements, so each line shows a slope rather than a shape. The FAQ and the location pages were built during the window and have no opening figure; they are in the table.

The rest of the picture:

• Commercial queries — traffic up 6x.
• The blog — launched February 2022 with no visitors at all; by February 2023 it was drawing 550 visits a day, over 13,000 a month.
• The FAQ section, added later, reached 30 visits a day, roughly 1,000 a month.
• Location pages — London, Manchester and others. Geo-dependent car finance queries are low frequency individually and produce leads every month collectively.
• Template pages went on producing steady traffic and leads after they were built.

I mapped out what comes next as well. Pull low- and mid-competition queries out of the larger semantic pool. Write informational articles that carry commercial intent: best used cars under a given budget, best used cars for tall people, best cars for dog owners. Build a vehicle catalog to cover the inventory.

What transfers to another site

Taking first-page positions in a hard finance niche on a weak domain came down to 5 things I would repeat anywhere:

2 years after this period closed, the same broker took the head term itself. That case is here.