A new in-home physical therapy practice. No website yet. No rankings, no reviews, no traffic history. This page documents the seven-phase plan we’re implementing — the architecture, the page counts, the quality gates, and the reasons we are not building the 200-page keyword matrix most agencies would sell.
Read the Seven Phases See the RoadmapMost agency case studies are written after the fact, once the numbers are flattering. This one is published from the planning stage — before launch, before rankings, before a single lead. You can see the decisions as they are being made.
| Phase | What it covers | Status |
|---|---|---|
| Design & Build | Three page templates, approved before the site is built | In progress |
| Phase 1 | 36-page launch architecture | In progress |
| Phase 2 | 20 local service pages | Gated on Phase 1 data |
| Phase 3 | 10 nested condition pages | Gated on Phase 2 data |
| Phase 4 | 12 selective local pages | Gated on Phase 3 data |
| Phase 5 | Google Business Profile, reviews, citations | Runs in parallel |
| Phase 6 | Conversion optimization and maintenance | Starts at launch |
| Phase 7 | Resource content | Optional · annual review |
Because the plan is the part you can actually evaluate. Anyone can show a traffic graph and claim credit for it. Far fewer can explain, in advance, why a site should launch with 36 pages instead of 200 — and hold to it when the client asks for more.
The client is a mobile in-home physical therapy practice serving Nassau County, New York. The engagement starts from a genuinely blank page.
No existing site. No legacy URLs. No redirect map. No historical Search Console data to lean on. Every ranking, every review and every tracked lead has to be created from zero — which means every architectural decision has to be right the first time, because there is no traffic data to correct it with.
| Factor | Redesign / migration | New build (this project) |
|---|---|---|
| Existing rankings | Protect them | None to protect |
| Redirect map | Mandatory, high-risk | Not applicable |
| Keyword data | Historical Search Console available | Zero — research only |
| Page count decision | Constrained by what exists | Completely open — and that is the danger |
| Biggest risk | Losing traffic at cutover | Overbuilding before any evidence exists |
That last row is the whole reason this framework is phased. When you can build anything, the temptation is to build everything. A 20-town by 10-service matrix is 200 pages, it can be generated in an afternoon, and it looks like enormous value on a proposal. It is also the fastest way to get a new healthcare site treated as a doorway network.
This practice treats patients in their homes. It has no clinic in any of the towns it serves. That single fact dictates the URL structure, the page templates, the schema, the Google Business Profile setup and the copy on more than twenty pages.
A practice that shows up when a patient in the service area searches for help — and turns that search into a phone call.
Everything in this framework serves that one outcome. The seven phases are ordered so each one earns the right to the next, and no page gets built before there is evidence it should exist.
Services -> Service Areas -> Local Services -> Subservices -> GBP/Reviews -> CRO -> ResourcesMoney pages first. Pages that answer “what do you do,” “where do you do it,” and “how do I book” come before anything educational. Blog and resource content is not skipped — it is deliberately delayed until the foundation is earning.
Most local SEO failures come from collapsing two buckets into one — usually by turning service-area pages into service pages, or by building “location” pages for a business with no locations.
What the business does. Core service pages, built in Phase 1, measured on county-level service rankings.
Where the business serves. Town and county pages, Phase 1, measured on [service] [town] rankings and map pack presence.
What problems patients need solved. Condition pages, Phase 3, measured on long-tail condition queries.
Local map pack authority. Off-site work in Phase 5, aligned from Phase 1, measured on map rank and review velocity.
Turning traffic into leads. Sitewide from Phase 1, ongoing in Phase 6, measured on calls, forms and conversion rate.
Each bucket has a different job, a different page type and a different success measure. Merge them and you cannot tell which part is failing.
This practice delivers care in patients’ homes. It has no clinic in Hicksville, no office in Garden City, no location in Plainview. The site must never suggest otherwise — and that single distinction drives decisions on more than twenty pages.
LocalBusiness entities, no fabricated addresses.That gap is the strategy.
| Phase | What it adds | Running total |
|---|---|---|
| 1 — Launch Foundation | 36 | 36 |
| 2 — Local Service Expansion | 20 | 56 |
| 3 — Nested Subservice Pages | 10 | 66 |
| 4 — Selective Local Expansion | 12 | 78 |
| 5 — GBP / Reviews / Authority | 0–2 | 78–80 |
| 6 — CRO / Ongoing Maintenance | flexible | flexible |
| 7 — Resources (optional) | 3–6 articles | — |
Phase 2 does not start until the Phase 1 thirty-day report exists. Phase 3 waits on the Phase 2 sixty-day report. Phase 4 waits on Phase 3’s ninety-day data. The operating principle after launch is the same every time: build the strongest opportunities first, monitor results, expand on evidence.
Each phase opens to its full work order — the goal, the individual steps, the templates, the quality gates, and the decision points that determine whether the next phase happens at all.
Design three templates. Approve them. Build them. Then build the other 33 pages.
3templatesRuns alongside Phase 1. Every page on this site is one of three layouts — approve those three and the rest is production, not design.
1. Homepage /
2. One core service page /services/physical-therapy/
3. One location/service-area page /service-areas/[town]-ny/Could you produce the next town’s page by duplicating and changing content only — no layout work? If not, it is not a template. It is a page that happens to look like one.
A change to the location template touches 21 pages. A header change touches all 36. Every change request is priced on pages affected, not on the effort of the single edit.
Tell Google five things and no more. Not too thin, not overbuilt.
36pagesThe launch foundation communicates the business type, the primary service, the local service footprint, the provider’s trust signals, and the conversion path. That is the sweet spot. A five-page site cannot establish a 20-town service footprint. An 80-page matrix means dozens of near-identical pages competing with each other before there is a single data point to justify any of them.
HOME /
|
+-- SERVICES HUB /services/
| +-- /services/physical-therapy/
| +-- /services/chronic-pain-management/
| +-- /services/orthopedic-physical-therapy/
| +-- /services/sports-physical-therapy/
|
+-- SERVICE AREAS HUB /service-areas/
| +-- /service-areas/nassau-county-ny/
| +-- 20 town service-area pages
|
+-- ABOUT US /about-us/
+-- CONTACT US /contact-us/
+-- INSURANCE & PAYMENT /insurance-payment/
+-- FAQS /faqs/
|
+-- UTILITY / COMPLIANCE
+-- /privacy-policy/
+-- /terms-of-use/
+-- /accessibility-statement/
+-- /thank-you/ (noindex - form redirect target)Non-page work that does not add pages but does block launch.
noindex or Disallow: / still in place.
Shipping with a staging noindex or Disallow: / still in
place. It is a checkbox on the runbook for a reason — the site can look
perfect and be invisible for weeks before anyone notices.
Five priority towns by four core services. Twenty pages, not eighty.
+20to 56 total
Phase 2 targets genuinely more specific intent —
orthopedic physical therapy Hicksville NY,
chronic pain management Garden City NY — while avoiding the trap.
Building all 80 town-by-service combinations at once means most end up too similar to
be useful, and the site starts to look like a doorway network.
URL pattern: /service-areas/[town]-ny/[service]/
5 priority towns x 4 core services = 20 pagesEvery Phase 2 page must be meaningfully different from its parent service page, its parent town page, and its sibling pages. If it cannot clear all three, the page gets merged or dropped — not published thin.
Before any page is built, an availability matrix confirms the service is genuinely offered in that town. Any blank cell means that page does not get built. No aspirational coverage.
Global authority pages for the problems patients actually search.
+10to 66 totalTen condition pages, nested under their parent service to reflect a real clinical hierarchy.
URL pattern: /services/[parent-service]/[subservice]/
Physical Therapy
+-- Post-Surgical Rehabilitation
+-- Balance & Fall Prevention
+-- Senior Physical Therapy
Chronic Pain Management
+-- Back Pain Physical Therapy
+-- Neck Pain Physical Therapy
+-- Sciatica Treatment
Orthopedic Physical Therapy
+-- Knee Pain Rehab
+-- Shoulder Pain Rehab
+-- Hip Pain Rehab
Sports Physical Therapy
+-- Sports Injury RehabGoogle does not require nested URLs and does not reward them with a ranking boost. Nesting is chosen here because these services have a real clinical hierarchy and the URL should reflect it — for hierarchy, user understanding, crawlable internal linking, breadcrumbs, content organization and future scalability.
Post-surgical rehabilitation fits under both Physical Therapy and Orthopedic PT. It gets one URL under Physical Therapy, cross-linked from Orthopedic. Two URLs for one topic is self-competition.
Localize proven winners only. Three signals must converge first.
+12to 78 totalPhase 3 creates global authority pages. Phase 4 creates local versions of the ones that earned it.
The global subservice page is getting impressions.
The town page is getting traffic.
Patients from that town are calling about that problem.
|
v then, and only then
/service-areas/garden-city-ny/back-pain-physical-therapy/
Global subservice URLs nest, because the clinical hierarchy is real:
/services/chronic-pain-management/back-pain-physical-therapy/
Local subservice URLs stay shallow, because a third level adds
depth without meaning:
/service-areas/garden-city-ny/back-pain-physical-therapy/
| Option | Build | Total |
|---|---|---|
| A — Wider town coverage | 5 towns x 4 core services = 20 | 86 |
| B — Deeper local intent (recommended) | 3 towns x 4 top subservices = 12 | 78 |
| C — Hybrid | 12 | 78 |
| D — Build nothing, optimize instead | 0 | 66 |
Option D is a real answer. If the data says existing pages are under-converting rather than under-ranking, the correct Phase 4 is no new pages at all.
Off-site work. Runs alongside the build, not after it.
0–2pagesPhase 5 drives map pack visibility and local trust, and it can start any time after launch. If the map pack is weak while organic is fine, Google Business Profile and reviews are the bottleneck, and running Phase 5 beats building more pages.
A review response can never confirm that someone was a patient or reference their condition — even when the reviewer disclosed it themselves. The response templates in this phase exist specifically so that nobody has to improvise it in the moment.
No end date. The rule: optimize before expanding.
∞ongoingPhase 6 never finishes. It is the monthly system that stops everything else from decaying, and it shares one session with Phase 5 — the GBP checklist and the site checklist are the same meeting, not two.
Nothing breaks visibly when the monthly session stops. Rankings and leads slide about four months later and nobody can say when it started. A named owner is the only real control.
Blogs are not skipped. They are deliberately delayed.
3–6articlesAn article about back pain competes for attention and internal link equity with the page that actually books appointments for back pain. Published first, it dilutes. Published after the service pages rank, it feeds them.
A Resources section, not a generic blog — the framing matters. “Resources” signals patient education tied to the practice’s services. A blog invites unfocused posting.
/resources/
What to Expect During In-Home Physical Therapy
In-Home PT After Surgery
Physical Therapy for Seniors at Home
Balance and Fall Prevention at Home
How Physical Therapy Helps Back PainEach article maps to an existing service or subservice page and links to it. None of them exist to rank on their own — they exist to answer the questions that stop a patient from calling, and to pass authority to the page that converts.
At the annual review, and only if the data says visibility is fine but trust or conversion is the gap. If leads are still constrained by rankings or map pack presence, Phase 5 and Phase 6 work comes first.
If a decision conflicts with one of these, the decision is wrong.
Every page must answer a question a real patient actually asks, in a way that no other page on the site already answers. If it cannot, it does not get built — regardless of search volume.
This is the framework we use. If you are launching from zero — or inheriting a site that was built as a keyword matrix — we can walk you through what the first ninety days should look like.
Talk to Us Re-read the Phases