MarketLensMost SEO case studies show you a hockey-stick chart and hide the work. This one does the opposite. Below is the complete anatomy of a campaign for a Romanian dental clinic’s content site: what the baseline looked like, exactly what changed in which order, which interventions moved numbers and which barely mattered — and the honest timeline, because nothing here happened in two weeks.
It also does not pretend the site was healthy to begin with. Organic traffic here had been falling for two years before anyone touched it — 31,369 sessions across 2024, 17,783 across 2025, bottoming out at 725 sessions in December 2025. What follows is a recovery, not a growth story, and the difference matters when you are deciding whether the playbook applies to you.
The headline results, year over year: Search Console clicks up +130.8% (8,122 → 18,747 across May–August), impressions up +116.4%, and direct contact actions — calls and form submissions — up +125.6% (503 → 1,135, January–August). Total paid search spend on the domain across the entire period: $0. Every figure here, and the exact API query behind it, is documented in the full dentalclinica.ro case study.
The Baseline: A Site That Published and Prayed
The clinic entered with roughly 50 blog articles about treatments — implants, whitening, orthodontics, extractions — written over several years. The content was professionally accurate but structurally invisible:
- Titles were generic (“Everything About Dental Implants”) and matched no actual query phrasing.
- No FAQ structured data anywhere, despite articles full of question-shaped content.
- Headings were essay-style (“A Modern Approach”) rather than search-shaped (“How much do dental implants cost?”).
- Articles rarely linked to each other; each page was an island.
- Search Console showed dozens of pages stuck at positions 4-20 — ranking well enough to prove relevance, badly enough to earn almost nothing.
That last point defined the strategy. Pages in positions 4-20 already have Google’s approval on relevance; they lack presentation and depth. The whole campaign was built around harvesting that stranded equity before writing anything new — the logic we later formalized in our striking-distance keyword optimization guide.
Reading the site’s own Search Console query report made the gap obvious. The impressions were landing on price phrasings, pain and anxiety phrasings, before-and-after phrasings, and treatment comparisons. Almost none of the clinic’s articles led with any of the four. That single insight drove most of the content decisions that followed.
Layer 1 (Weeks 1-4): Titles, Descriptions, and FAQ Schema
The first pass touched only metadata and structured data — no body content:
- Titles rewritten to match the highest-impression GSC query per page, kept to 55-60 characters. “Everything About Dental Implants” became the equivalent of “Dental Implant Costs & Stages: What to Expect in 2026.”
- Meta descriptions rebuilt at 150-160 characters with the primary keyword and a concrete hook (a price range or timeline beats “learn more” every time).
- FAQ schema added via front matter to every article that contained genuine Q&A material — about 30 pages gained FAQ rich-result eligibility in one sprint.
Result: the fastest wins of the campaign. Several pages sitting at positions 4-7 moved into the top 3 within three to six weeks, purely on better query alignment and richer SERP presentation. Layer 1 is low effort and real, but note its ceiling: it accelerates pages that already deserve to rank. It did almost nothing for pages stuck at 12-20.
Layer 2 (Months 2-3): Headings Rebuilt from Query Data
Every article’s H2/H3 skeleton was rebuilt against its own GSC query report. Where a page had impressions for “how long do dental implants last” but no section addressing it, a section was created with that phrasing as the H2. Essay headings became question headings; buried answers moved directly under their headings in 40-60 word blocks that snippets and AI Overviews could lift cleanly.
This is unglamorous work — roughly 45 minutes per article — and it was the campaign’s best effort-to-impact ratio after Layer 1. The method is documented as a standalone process in restructuring H2 and H3 headings for search intent.
Layer 3 (Months 3-9): The Depth Work That Carried the Campaign
The highest-effort layer, and the one responsible for most of the recovery. For each priority article:
- Real prices in tables — ranges honest about variation, e.g. implant costs broken down by stage (consultation, placement, abutment, crown) instead of “costs vary.”
- Clinical timelines — week-by-week healing expectations, number of visits, when normal life resumes. This is what anxiety-driven searchers actually want.
- Misconception sections — direct answers to fears (“extraction pain peaks in the first 24-48 hours and is managed with…”) written from clinical experience.
- E-E-A-T reinforcement — visible medical authorship and credential signals, which in healthcare is not optional. Both Google and the answer engines weight clinical trust signals heavily on YMYL topics. The specifics of those signals are their own topic, covered in E-E-A-T clinical trust signals in the age of generative AI search.
Depth work took months to compound, and that is the honest shape of the curve: Layer 3 gains arrived steadily from month 4 onward and kept climbing after the project’s active phase ended.
Layer 4 (Ongoing): Hub-and-Spoke Internal Linking
Every article received 2-5 contextual links: treatment overview pages (hubs) linked down to detail articles (cost, pain, recovery, alternatives), and every spoke linked back up plus sideways to its siblings — always with descriptive anchors (“how long implant healing takes”), never “click here.” Isolated pages stopped existing.
The Results, Properly Attributed
Search Console, comparing 1 May – 30 Aug against the same four months a year earlier:
| Metric | 2025 | 2026 | Change |
|---|---|---|---|
| Clicks | 8,122 | 18,747 | +130.8% |
| Impressions | 483,969 | 1,047,189 | +116.4% |
| Mobile average position | 23.8 | 8.6 | −15.2 places |
| Desktop average position | 43.5 | 25.6 | −17.9 places |
Analytics, comparing 1 January – 30 August year over year:
| Metric | 2025 | 2026 | Change |
|---|---|---|---|
| Organic sessions | 13,817 | 29,384 | +112.7% |
| Direct contact actions | 503 | 1,135 | +125.6% |
| Paid search spend | $0 | $0 | — |
Two things deserve emphasis. Conversions grew slightly faster than sessions — because cost- and comparison-intent pages, the ones optimized hardest, attract patients who are ready to book, not just read. And the mobile position shift, from 23.8 to 8.6, is where the click growth actually came from: page two to the bottom of page one is the move that turns impressions into visits.
The third thing deserves emphasis more than either: this was a rescue. The monthly organic curve falls all the way through 2024 and 2025 to a floor of 725 sessions in December 2025, then climbs to 6,483 by July 2026. A recovering site has cheap wins available that a healthy site does not, and you should discount the percentages accordingly. Conversion measurement was a prerequisite either way — calls and form submissions had to be firing as reliable GA4 events before any of this could be counted, which is harder than it sounds once a consent banner is in the way, as the GA4 timing and consent problem explains.
What This Means If You Run a Practice
The replicable core is not “write more content.” It is: mine Search Console for stranded positions 4-20, fix presentation first, then spend your depth budget on cost, pain, proof, and comparison intent — with visible clinical authority throughout. No ads, no purchased links, no new platform. If you want the same diagnostic run on your own site, the MarketLens Premium audit produces the full striking-distance inventory and page-by-page action plan this campaign started from.