Follow
Follow The Observatory
Clinic websites fill the organic selections, but AI citations introduce a recurring platform host and recorded model accounts foreground other names.
Quick answer
Which aesthetic-clinic websites recur in Tokyo's search results?
Six clinic-site domains account for 100% of the 65 rows (65/65) in The Observatory's 13 published Tokyo organic top-five readings from June 22–September 27, 2026. Jiyugaoka, Matsurika and Azabu's domains each appear in every selection. On the citation surface, a clinic domain occupies every published first row: kojihifu.com is first in 8/12 citation top fives and biancaclinic-tokyo.com in 4/12. The concentration therefore extends beyond organic membership without turning repeated rows into separate clinics.
At a glance — key facts
- A direct clinic-site core: six domains supply all 65 organic rows; three of those domains appear in 13/13 publications.
- Clinic domains occupy every published citation first row: kojihifu.com is first in 8/12 citation top fives; biancaclinic-tokyo.com in 4/12.
- A recurring host still sits beside clinic sources: Instagram appears in all 12 citation top fives and in 11/11 matched citation windows.
- Selection depth changes the host comparison: Instagram appears in 11/11 matched citation top fives but only 4/11 opening top fives at equal depth.
- The question still matters: a broad clinic-search pattern does not determine the first result for an expert-dermatologist query.
Repeated websites, not an expanding list of clinics
The organic selections use six exact domains: jiyugaokaclinic.com, matsurika-clinic.com, azabu-skinclinic.com, kojihifu.com, biancaclinic-tokyo.com and plazaclinic.net. Their own pages identify the corresponding clinics: Jiyugaoka, Matsurika, Azabu, Kojimachi, BIANCA and Plaza. The June 22–28 and September 21–27 readings bracket the 14-publication series.
Jiyugaoka, Matsurika and Azabu each occur in all 13 selections. The order still changes: Matsurika is first in August 17–23, while Kojimachi's domain is first in September 7–13. Continuity of source type does not mean an unchanging first name.
This makes Tokyo a specifically clinic-site-heavy organic record. It does not describe every query or everything below the published fifth position. The ordering follows frequency of holding the top organic position; the displayed presence percentages describe another attribute. The ranking guide explains how to keep those measures separate.
An expert-dermatologist query reaches beyond the six domains
The exact published query “expert aesthetic dermatologist Tokyo” supplies an instructive exception to treating the aggregate selection as a universal clinic list. The July 6–12 and August 3–9, 2026 Spotlights put e.takadanobaba-derm.com at median best organic position 1. That domain is outside the six-domain set supplying the 13 aggregate top fives. The August 31–September 6 reading instead puts Matsurika Clinic Ginza first on that same query.
These are single-query Google desktop organic results. The evidence establishes which source appears first for the question; it does not establish a practitioner's expertise, care quality or the reason for the position. Its value is to qualify the broader recurrence: a domain absent from the aggregate top five can still head a specific local search.
Instagram enters the cited-source record
Both kojihifu.com and biancaclinic-tokyo.com appear in 12/12 published Tokyo AI Overview citation top-five readings, spanning June 22–September 20, 2026. Instagram also appears in all 12. The first and last publications show the host beside clinic sources.
A host-level entry does not identify the underlying account, its author or the clinic discussed. Instagram's recurrence therefore cannot be assigned wholesale to patient opinion, independent commentary or any one provider.
There are 11 windows with both opening and citation publications. Because opening lists contain ten rows and citation lists five, the first useful comparison checks equal selection depth before interpreting the difference.
| Instagram in Tokyo's Google desktop Japan record, 11 shared windows from June 22–September 20, 2026 | Publications containing the domain |
|---|---|
| AI Overview citation, published top five | 11/11 |
| AI Overview opening, first five published positions | 4/11 |
| AI Overview opening, full published top ten | 10/11 |
The difference remains after matching depth: Instagram is more consistently among the first five cited-source entries than the first five opening entries. Matching dates and list length does not make opening and citation the same measure.
Two examples show what the comparison contains. In July 27–August 2 opening, Instagram is sixth, while the citation counterpart puts it fourth. In August 3–9 opening, it is outside the published ten; the citation counterpart includes it fifth. One example is affected by list depth; the other remains an omission even at ten opening entries.
“Tokyo's recurring clinic websites describe the source pattern in the published selections; the question and the answer surface determine which part of that pattern a reader sees.”
The Observatory
Model-selected names add a different account of Tokyo
The September 21, 2026 Model Watch compares GPT-4o native and Claude Sonnet 4.6 native across six paired prompts. Its published named-actor comparison emphasizes Shinagawa Cosmetic Surgery Clinic in GPT-4o's account and Takagi Clinic in Claude's. These are names foregrounded in the recorded answers, not additional entries in the organic ranking or a recommended provider list.
The same comparison records different roles beyond clinic names: local innovators and regulatory bodies in GPT-4o's account; device distributors and concierge coordinators in Claude's. This broadens the narrative beyond the website-heavy measured selections without establishing who controls Tokyo's clinical market.
A separate Gemini 2.5 Pro native reading on September 28 acknowledges limits concerning Japanese-language information and Tokyo's internal clinic hierarchy. That is a recorded model qualification, not a measured exclusion rate. There is no same-date partner in this set from which to infer model superiority or improvement.
Questions readers ask
Does the 65-row count mean 65 clinics were observed?
No. It counts five published rows in each of 13 organic selections. Those repeated rows use six exact domains, with identities and branches requiring separate interpretation.
Do the six recurring domains cover every Tokyo clinic search?
No. The published expert-dermatologist Spotlights put e.takadanobaba-derm.com first in two dated windows even though that domain is outside the aggregate six-domain set. A selected market table and a single-query result answer different questions.
Can Instagram's citation presence be credited to a particular clinic?
The domain-level row does not resolve the account, author or provider discussed. More specific source evidence is required before assigning the appearance.
Do these counts establish a rising or falling share over time?
They establish selection membership. Historical aggregation rules differ: the June 29–July 5 citation reading uses mean presence across runs where a source appeared, while September 14–20 uses period-wide cumulative citation volume. Recurrence does not supply a normalized growth series.
Evidence basis and limits
The Tokyo research cutoff was September 30, 2026. Its 93 published readings comprise 51 measured publications and 42 narrative publications. Organic and opening evidence reach September 27; citation and Spotlight reach September 20; the narrative set includes September 28. Some Lens and Model Watch publications share underlying answers.
The organic count covers all 13 published top fives. The citation count covers all 12 published top fives. The Instagram comparison uses only 11 windows with both answer-family publications: July 20–26 has neither, September 7–13 has citation without opening, and September 21–27 has opening without citation.
Measured records identify Google desktop capture in Japan for the Tokyo market; the room is configured in English. They do not establish a Tokyo-city capture coordinate or comprehensive Japanese-language coverage. These observations do not measure clinical outcomes, safety, suitability or provider quality.
The Observatory does not claim to measure real-world revenue, market share, or business leadership unless explicitly stated. It publishes observed AI and search positioning within a defined scope and time window.
Intelligence by AAI Kernel. Published by The Observatory.
Continue the reading
Because repeated source rows are not independent clinics, the Tokyo identity context separates brand, branch and practitioner before these appearances are attributed. The Aesthetic Clinics context places Tokyo's direct-source pattern beside intermediary-led examples elsewhere. A claim that representation is changing requires more than repeated dates, so Method's comparability test sets the conditions for moving from recurrence to a trend.
Related reading
Tokyo observations · Citation Surface & Source Share · Official vs Narrative Layer
