Skip to content
Neaptidestudio
case studies

SEO case study, car dealership: organic traffic ×9 in five months

August 19, 2026 · 6 min read

An anonymized case from the studio archive: a regional light-commercial-vehicle dealer. From 11 to 67 queries in the top 10, organic traffic ×9, paid-search spend cut by 70%. Figures from primary reports, the method, and honest limitations.

One of the cases from our archive: a regional official dealer of light commercial vehicles (LCV) — sales, service, parts. In five months organic traffic grew ninefold, commercial queries in the top 10 went from 11 to 67, and paid-search buying was cut by 70%. The brand, city and domains are withheld; every figure is taken from the primary reports unchanged.

The starting point: an invisible site amid live demand

Demand in the niche was large and measurable: the collected semantics totalled about 47,500 monthly searches, and the queries were hot — “buy [model] in [city]”, “[model] price”, “official [brand] dealer in [city]”. The starting benchmark across 181 queries showed only 11 in Yandex's top 10 and 18 in Google's, with 137 queries out of tracking range entirely. The dealer compensated for invisibility with paid search — roughly 8,000 visits a month from the paid channel.

The method: semantics → structure → the external layer

Three layers of work. First: a full rebuild of the semantics — about 190 commercial queries with real search volume, covering the model range, service and parts, with a dedicated landing page for every group. Second: on-page optimization — structure, copy and metadata built around that semantics. Third: the external layer — within a month, links to the project were placed on 300+ vetted sites. That was the industry norm at the time; today we build authority differently — see the limitations.

The dynamics: from invisibility to 65% of demand in the top 10

  • Queries in Yandex's top 10: 11 (February) → 49 (March) → 59 (April) → 67 (May) out of ~190 tracked.
  • By May: 28 queries in the top 3, 18 ranked #1; 59 in Google's top 10.
  • Organic traffic by month: 254 → 468 → 1,032 → 1,818 → 2,349 visits (five consecutive months).
  • By May, queries in the top 10 accounted for 65% of the total semantic search volume (30,879 of 47,553 monthly searches).
  • Main-site traffic: March 4,318 → April 5,694 visits (+32%).

The economics: organic replaced paid traffic

At the start: organic — 254 visits a month, paid search — 8,006. Five months later: organic — 2,349 (×9.2), paid cut to 2,418 (−70%). Every organic visit on a hot commercial query replaced a paid click: SEO paid for itself not through abstract “visibility” but through a direct cut in ad budget while total traffic grew.

“80% of our sales now come from the internet,” the dealership's owner said during our work together. The attribution is withheld along with the brand — but that phrase describes the change more precisely than any table: the online channel went from auxiliary to the dealer's main source of sales.

What this case proves — and what it doesn't

It proves that the discipline of “semantics with real search volume → a landing page per query group → measurable month-by-month dynamics” works in a competitive commercial market, and that the economic effect of SEO can be measured through paid-traffic replacement. It does not prove the same result is reproducible today with the same method. Mass link placement no longer works and is risky, and part of the demand has moved from classic results into AI-assistant answers. Only the link layer of the method is obsolete; semantics and structure matter twice as much now — they are exactly what language models read and cite.

faq

The short version

Why is the case published without the brand and region?

We don't publish client data without explicit consent. The figures come from primary reports — rankings, traffic, semantics — and are quoted unchanged; the brand, city and domains are replaced with generic terms.

This case is from the “classic SEO” era — is it still relevant?

The mechanics changed; the discipline didn't. The “mass links” layer is dead, but the chain of real-volume semantics → structure → measurability transferred into our GEO/AEO work unchanged: today that same structure is read not only by search engines but by ChatGPT and Perplexity.

What would you do differently today?

Entity corroboration and citable sources instead of link buying; share of voice in AI answers plus rankings instead of rankings alone; llms.txt, schema.org and answer-first pages from day one. The “before” benchmark would stay exactly the same — without a baseline there is nothing to prove.