"Knee replacement cost in Bengaluru", "knee replacement cost in Hyderabad", "knee replacement cost in Nagpur". A hospital chain with branches in twelve cities and thirty procedures can produce three hundred and sixty of these pages in an afternoon with a template and a spreadsheet. Whether it should is a different question.
This article is for SEO leads and web teams at hospital groups and multi-branch clinics who already know the basics of programmatic pages and need a decision rule. The question is narrow: which structured inputs make a cost page for one city genuinely different from another, and which inputs are only decorative. Anything in the second group should be merged, not multiplied.
We covered the general case in location pages at scale. Here we go deeper on cost pages specifically, because price is the one piece of data that is both highly searched and easy to get wrong.
What Google is likely to treat as thin
Google's spam policies describe two relevant patterns. Scaled content abuse is "when many pages are generated for the primary purpose of manipulating search rankings", whatever the generation method. Doorway abuse is "when sites or pages are created to rank for specific, similar search queries". A set of city cost pages that differ only by the city name and a changed number meets both descriptions uncomfortably well.
The policy is about purpose and value, not templates. A template is fine. Hospitals have used templates for years. What matters is whether each resulting page tells a visitor something they could not get from the page for the neighbouring city.
The inputs that can justify a separate page
We think about cost-page inputs in three groups. Think of each as a column in your content table.
Group A: price drivers that really vary by location
| Field | Why it can differ by city |
|---|---|
room_category and tariff | Room types and rates differ by branch |
implant_tier | The implants stocked or the choices offered at that site |
stay_nights | Typical length of stay can differ with clinical protocols and facilities |
payer_mix | Which insurers, TPAs or schemes are empanelled at that branch |
surgeon_team | Which departments and specialists operate there |
facility_equipment | Whether a particular technology is available |
Group B: local context that helps the reader decide
Travel and stay logistics for patients from outside the city, local airport and rail access, languages spoken, branch-specific package inclusions, and the process for getting a written estimate at that branch.
Group C: decorative inputs
City population, the city name repeated in headings, generic "why choose us in [city]" paragraphs, and stock text about the state. These add words, not information.
The rule is simple: a page needs at least two Group A or B inputs that differ in substance from the nearest sibling page, and those inputs must be backed by data you hold for that branch. Otherwise it merges.
The swap test
Before building, take a draft page for City A and replace City A with City B in the text, the table and the schema. Read it. If everything still makes sense and nothing in the numbers or inclusions is wrong, the pages are not different. Merge them into one page for the procedure with a branch comparison table, or keep a single page with a selector.
A second test is the reader's next action. If the visitor in City A and City B would do the same thing (call the same central number, request the same estimate) then one page serves both.
A worked example, clearly hypothetical
This example is invented to show the logic. The numbers are not real prices and do not describe any hospital.
Imagine a chain with branches in three cities offering unilateral knee replacement. The data table holds:
| Branch | Room categories | Implant tiers offered | Typical stay (nights) | Insurers empanelled | Hypothetical package range |
|---|---|---|---|---|---|
| City A | Shared, single, deluxe | Standard, premium | 4 | 18 | ₹X1 to ₹X2 |
| City B | Shared, single, deluxe | Standard, premium | 4 | 17 | ₹X1 to ₹X2 |
| City C | Single, deluxe | Premium only | 3 | 9 | ₹Y1 to ₹Y2 |
Branches A and B have the same room categories, implant tiers, stay and range. Their insurer lists differ by one name. That is not a reason for two pages. Merge them: one page for "knee replacement cost in [region or chain]" with a branch table and a note on insurer empanelment.
Branch C genuinely differs: no shared room, one implant tier, a shorter stay and a different range. This justifies its own page, because a visitor deciding between options would find different answers.
The result is two pages instead of three, and each is defensible. Multiply by thirty procedures and the saving in thin pages is large.
The data model behind the page
To make this repeatable, store the inputs as fields instead of writing prose per page. A workable table for a cost page includes:
procedure_id,branch_id,last_verified_dateroom_category(list),implant_tier(list),stay_nights_typicalprice_low,price_high,price_basis(package, estimate range, or case-by-case)inclusions(list) andexclusions(list)payer_listandscheme_list, each with a verified dateestimate_route(how a patient gets a written estimate at this branch)clinical_reviewerandclinical_review_datefor the explanatory text
A build script can then apply the rule automatically: if two branches have identical Group A fields, emit one page and set the other branch's URL to redirect or canonicalise to it. Google's guidance on consolidating duplicate URLs explains canonical signals; a redirect is the stronger choice when the second URL has no reason to exist.




