A healthcare specialist, measured against generalists. 30% more impressions by July.
In May 2026 I took on the SEO strategy for a US software company that does one thing: reputation management for healthcare providers. Google already showed the site often, near the foot of page one, but it had few pages for the ways practices actually shop. I benchmarked it against eight rivals and laid out a phased page roadmap. By July, monthly impressions stood 30% above their May level.
Results
Google impressions
Visible already, thin where buyers look
The programme was onboarded on 8 May 2026, and the site was not starting cold. Google showed it roughly 150,000 times that month, at an average position of about 9, so it hovered at the bottom edge of page one. The product serves medical practices, group practices, hospitals and health systems in the US and nobody else, which puts it up against review tools sold to every industry and against large patient-experience vendors.
Where it fell short was coverage. My May review counted four feature pages, three pages by type of organisation and four specialty pages, while rivals carried ten, twenty or more specialty pages each. The software connects with more than a hundred health record and practice-management systems, yet not one of those integrations had a page of its own.
Mapping the routes buyers take into the site
The onboarding answers set the frame: healthcare only, the US market, and visibility in both search results and AI answers within scope. The client's own team kept day-to-day editing and some page creation in house, so my part was the strategy and the site architecture behind it.
- 01
Benchmark against eight rivals
I compared the site with eight competitors, from multi-industry review platforms to enterprise patient-experience suites. Side by side, their page sets showed where a buyer could find a relevant page elsewhere and nothing here: specialties, integrations and pages for particular kinds of healthcare organisation. That comparison became the backbone of the plan.
- 02
Phase the new pages
The gap analysis set out somewhere between eighty and a hundred-plus new pages, spread across a six-month roadmap in phases rather than one big launch. Phasing lets the in-house team absorb the work next to its own editing, and lets the most commercially useful page types go first instead of waiting for the complete set.
- 03
Give each page family its keywords
I mapped keywords to every page family - features, specialties, organisation types, buyer personas and use cases - and reviewed the mapping with the client. A single target per page keeps new pages from competing with each other, and the persona and use-case sets reach searches a features-only site never appears for.
- 04
Split paid from organic
A URL plan separated the landing pages built for ads from the pages meant to rank, with six new URLs reserved for SEO. Drawing that line early keeps thin ad pages out of the index and each set doing its own job.
- 05
Read the link profile for what it is
A May export showed roughly 31,000 links from about 3,800 referring domains, around two thirds of them image links, mostly from the review badge customers embed on their own websites. Those links are genuine but say little about subject matter, so I pointed future link building at editorial and press coverage.
The opening stretch of a six-month plan
Across the four months, monthly impressions ended 14% above their May level after peaking 30% above it in July. Average position stayed between about 9 and 10 every month, so the site kept its place on page one while being shown more often.
The roadmap runs six months from May, so this window covers its first stretch. The page families rivals already own - specialties, integrations, organisation types, personas and use cases - are now mapped with keywords and URLs instead of missing from the plan. That is the baseline I would hold the next phase against.
Other case studies
Want a result like this for your site?
A 30-minute call, no pitch deck. Tell me where you are stuck and I will tell you honestly whether I can help - and if I cannot, who can.