MEDICAL
Patients do not search for a clinic's name. They describe their own symptoms in a sentence and ask where to go. You can be the answer to that question without using any of the expressions Korea's Medical Service Act prohibits.
In one paragraph
Healthcare AEO is the work of designing a medical institution's documents so they get cited as evidence when patients describe a symptom or a situation to an AI answer engine. Most questions in this industry start with no brand name in them — 'which department should I see for these symptoms', 'when is this procedure indicated'. To be the answer you need an explanatory document about that question, not a clinic introduction.
At the same time this is the most heavily advertising-regulated field there is. Korea's Medical Service Act prohibits expressions that guarantee treatment outcomes, expressions comparing one institution with another, superlatives with no objective basis, and advertising in the form of treatment testimonials. Yet that list of prohibitions overlaps substantially with the sentences answer engines rate poorly. There is a stretch where the document that respects the regulation is also the one favoured for citation.
Last verified
WHEN YOU NEED THIS
These are the situations we hear repeatedly in consultations. If any of them apply, start by measuring.
Brand-named and unnamed questions call for entirely different documents. New patients mostly arrive through the unnamed ones, and there is usually no document on the site that answers them.
When a medical institution's site consists only of department listings and directions, third-party media take the answer slot for questions that need explaining. Here the range of content types itself has to grow.
When the notation on map services, portal business listings, public lookup systems and the website all differ, an answer engine has no basis for deciding which is correct. Aligning the notation comes first.
What is prohibited is guarantees, comparisons, superlatives and testimonials — not explanation itself. A document setting out indications, the procedure, recovery time and precautions with evidence attached is inside the regulation and favourable for citation.
WHAT WE DO
Written as units of work rather than abstract proposals. The scope of an engagement is set from this list.
Before starting, we list the expressions the Medical Service Act prohibits and confirm which media carry advertising subject to pre-review. Nothing gets written before that list is settled.
We organise the questions the way a patient describes their own condition. The standard is the words patients actually use, not medical terminology.
We write so that one question maps to one paragraph. Answer engines do not use a document whole — they cut it — so the meaning has to complete inside the paragraph that gets cut out.
Medical institutions have unusually many places where their information scatters. When the notation diverges character by character, an answer engine cannot bundle it as the same clinic.
Under the Medical Service Act, prices for treatments not covered by national health insurance are subject to notice and reporting. Where the website's notice diverges from what is actually charged, it is both a regulatory problem and a trust problem for an answer engine.
We measure repeatedly under the same conditions and record the change, and where AI answers something contrary to fact we build the supporting document and track it as a correction target.
PROCESS
What you receive at each stage is stated alongside it. Durations are the working time Navirang controls; they are not a promise about when results appear.
We confirm the departments and the advertising media, determine whether pre-review applies, and build the list of prohibited expressions.
Draft expression guide
Questions get organised as patients describe symptoms, and the set to be measured gets fixed.
Question set document
We put the questions to the answer engines, organise the currently cited domains by character, and record any incorrect statements.
Observation record · misinformation list
Name, address and opening hours across website, maps and portals get compared, aligned to one form, and declared through structured data.
Notation comparison table
We decide which explanatory documents to make per type and hand them over sorted by effect against cost.
Content design document · improvement priorities
DELIVERABLES
We do not do work that ends in conversation. The documents below remain, and become the baseline for the next measurement.
The expressions the Medical Service Act prohibits paired with the evidence-based expressions usable instead. Every stage of writing gets reviewed against this document.
The full set of patient questions used in measurement, with their type classification. It is the basis for re-measurement, so the institution keeps it.
A table recording mention, citation and accuracy per question × answer engine combination, along with the domains cited alongside.
The differences in name, address, phone and opening hours across website, map services and portal business listings, item by item.
The sentences where AI answered contrary to fact, listed with the sources those sentences appear to have drawn on.
The description of these provisions is guidance on what to confirm before starting, not legal advice. Whether a particular piece of wording is lawful, and whether it passes review, rest with the competent authority and the self-regulatory review body, and final responsibility lies with the medical institution as the advertiser. Navirang prepares the wording and the evidence in a form that can be reviewed.
HOW IT CONNECTS
Our work moves as one piece. SEO builds the foundation for being found by search engines, AEO raises the odds of that information being cited in an answer, structured data helps machines understand the facts, and content supplies the evidence there is to cite.
| Area | Relationship to this work |
|---|---|
| AEO audit | The single measurement of current citation state using the symptom question set |
| Citable content design | Rewriting indication and procedure explanations into citation-sized units |
| Structured data | Declaring departments and opening hours as machine-readable facts |
| Citation monitoring | Finding misinformation, building the evidence to correct it, and tracking through to re-citation |
| AEO by industry | The parent page for comparing question types across other industries |
| Legal, tax and labour | The same condition — advertising regulation blocking performance claims — so the way trust gets evidenced is similar |
FAQ
A It does not. What the Act prohibits is guaranteed treatment outcomes, comparison with other medical institutions, treatment testimonials and superlatives with no objective basis — not explaining medical information. The documents that actually get cited in AEO are mostly ones setting out indications, procedure and precautions with evidence, and that form does not overlap with the prohibitions. If anything, documents stripped of promotional adjectives are cited more readily by answer engines, so there is a stretch where the regulation and the optimisation point the same way.
A We do not. Advertising in the form of treatment testimonials is explicitly prohibited by the Medical Service Act, and Navirang does not produce content that crosses that line. Where evidence of trust is needed we use verifiable facts instead of reviews — departments, equipment held, the clinicians' board certifications and academic activity, the treatment process, the update date. From an answer engine's point of view a confirmable fact is a better citation basis than an anonymous review anyway.
A They ask differently. Rather than asking for a clinic by name, they more often describe their symptoms in a sentence and ask which department to see, or when a given test is done. That is why a clinic introduction page does not answer these questions. What answers them is a document explaining that symptom or procedure, with the institution cited inside it as the source. Which domains are actually being cited for which questions is something we measure directly for your departments in the free audit.
A The applicant for review is the medical institution. Navirang prepares the wording to be submitted and the evidence supporting it in a reviewable form, and filters out sentences that fall foul of the prohibited-expression list in advance. Whether particular wording passes is the self-regulatory body's judgement, and final legal responsibility lies with the medical institution as the advertiser. We set this boundary out in writing at the contract stage before starting.
A It can. These errors mostly arise because the website, the map services and the portal business listings state different things. With no basis for judging which is current, an answer engine uses whichever is referenced more often. Navirang first compares the notation on each channel into a table, designates one as canonical and aligns the rest, then re-declares that fact through structured data. After that we measure the same question repeatedly and track whether the answer changes.
A We will not. Advertising that compares one medical institution with another is prohibited by the Medical Service Act. There is another way to answer comparative questions, though — documents comparing medical options rather than institutions, such as the difference in indications between procedure A and procedure B, or between two testing methods. Most patient comparison questions are option comparisons rather than clinic comparisons, so this form alone answers a large share of them.
A Both. Healthcare is an unusually location-bound industry, so conditional questions in the form of 'district name plus department' are common. The industry axis decides which document answers which question type; the regional axis looks at which domains currently hold the ground in that area. Navirang has published field measurement of the composition of the top 10 Naver web documents across all 17 Korean provinces, and the regional competitive picture starts from that data. Those regional pages exist in Korean only, since the queries behind them do not hold as search intent in English.
Give us the departments and the website address and we will run the free audit on patient questions. Expressions that fall foul of the regulation get flagged at the audit stage.
We reply within one business day.