Original THI Research

Inside Turkey's Hair Transplant Industry

Clinic Disclosure Study: What 25 Turkish Hair Transplant Clinics State Publicly

A dated count of what a defined set of Turkish hair transplant clinics state on their own English-language pages about price, daily patient workload, surgical team, follow-up and who performs each surgical stage.

Of the 24 clinics whose English pages could be read, 21 state who performs at least one surgical stage, 14 state a price, 13 describe their surgical team, 11 state a follow-up schedule and 10 state how many patients they treat per day. One site could not be read and is reported separately.

Clinic TransparencyResearch by Turkey Hair Index ResearchPublished by Turkey Hair IndexPublished 2026-10-17Methodology Clinic Disclosure Study protocol v1.0

At a glance

Research Snapshot

Sample: 25 clinics^[1]
Underlying data: The 25 clinics with a completed research dossier and a published THI Score on the freeze date, 19 in Istanbul and 6 in Ankara, as fixed in the study protocol
Data collection: 2026-10-16 – 2026-10-17
Geography: Istanbul, Turkey (19 clinics), Ankara, Turkey (6 clinics)
Last updated: 2026-10-17

Research Question

Among the 25 clinics Turkey Hair Index researched and scored under methodology v1.0 in September and October 2026, how many state each of five patient-decision items on their own standing English-language web pages, and how does that differ between the two cities in the set?

What the data shows

Key Findings

Finding
Of the 24 clinics whose English-language pages could be read, 21 (88%) state who performs at least one of four surgical stages, 14 (58%) state a hair transplant price, 13 (54%) describe who works with the lead doctor, 11 (46%) state a follow-up schedule and 10 (42%) state how many patients they treat per day.^[1][2]
Underlying data: Counts of the 24 readable clinics coded disclosed and not disclosed for each of the five items, and of the 1 clinics coded not confirmed, computed from the 125 coded cells
Interpretation
These are counts of statements on clinic pages. They do not show whether a statement is accurate, or whether a clinic that makes no statement works differently.
Finding
2 of the 24 readable clinics state all five items, 10 state four, 2 state three, 6 state two, 1 states one and 3 state none.^[1][2]
Underlying data: Number of the five items stated by each of the 24 clinics with no not-confirmed cell, counted per clinic from the coded cells
Finding
17 of the 125 coded cells are borderline readings. Between a strict reading (borderline disclosed cells become not disclosed) and a loose reading (borderline not-disclosed cells become disclosed) the count for an item spans at most 8 clinics (surgical team: 10 to 18 of 24). Surgical roles remain the most often stated item under both readings (21 or 22 of 24); the order of the other four items changes.^[1][2]
Underlying data: Counts of borderline cells per item and the disclosed count if every borderline cell were read the strict way or the loose way, computed from the coded cells
Finding
Among the 21 clinics that state who performs at least one surgical stage, 16 name the doctor alone for hairline design and 16 for creating the recipient sites. For graft implantation, 7 name the team alone, 4 the doctor alone and 1 both, and 9 do not say.^[1][2]
Underlying data: Who is stated to perform each of four surgical stages, counted for the 21 clinics coded disclosed on surgical roles
Interpretation
A clinic that names who performs some stages and is silent on others has not told a reader who performs the stages it leaves out.
Finding
Most clinics that state no figure still address the item in general terms: all 13 clinics without a follow-up schedule do, as do all 11 without a team description, 9 of the 10 without a price and 7 of the 14 without a daily patient count.^[1][2]
Underlying data: Counts of not-disclosed cells for which the coder recorded that the pages address the item only in non-qualifying general terms

Why We Studied This

Patients comparing hair transplant clinics in Turkey rely heavily on what clinics publish about themselves. A price, the number of patients a team treats in a day, who works beside the doctor, what follow-up looks like and who performs each surgical stage are all things a patient can reasonably ask. Whether a clinic writes them down on its own pages is something that can be counted.

Turkey Hair Index research already records these items for each clinic it evaluates, and the guides explain why each one matters. This study asks a narrower question that a count can answer: how often do the clinics we have researched state them publicly, in plain text, in English?

The result is a description of public statements by a defined set of clinics on two days. It is not a ranking, a measure of quality or an estimate for the whole industry.

Sample

The population is every clinic for which Turkey Hair Index had completed an internal research dossier and published a THI Score on October 16, 2026: 25 clinics, 19 in Istanbul and 6 in Ankara. The clinics are listed below as the population, without results.

The set is selective and evidence-led. Inclusion in the Turkey Hair Index is not an endorsement, absence is not a negative assessment, and these clinics are not a random sample of Turkish clinics. The results describe these clinics only.

Included

A clinic with a completed Turkey Hair Index research dossier and a published THI Score on October 16, 2026.
The clinic's official website, and its standing English-language pages.

Excluded

Clinic suggestions and leads that Turkey Hair Index has not researched and published.
Blog and news articles, patient stories, galleries, testimonials and review pages.
Pages in languages other than English, and contact, booking, privacy, cookie and terms pages.
Social media, review platforms, forums, messaging and quotes obtained by inquiry.
Text that exists only in images, PDFs, videos, chat widgets, calculators or forms.

Anonymization: The clinics are listed as the population, but no result is attributed to a named clinic. The study publishes aggregate counts only.

Methodology

For each clinic, the study records whether its standing English-language pages state each of five items. Each of the 125 clinic-by-item cells is coded disclosed, not disclosed or not confirmed. Disclosure means that a statement exists on a page. It is what the clinic says, not a checked fact, and the study does not judge whether a statement is true or good.

The population, definitions, pages in scope, retrieval patterns and coding rules were written down and frozen on October 16, 2026, before any cell was coded. The disclosure rate for an item is the number of clinics coded disclosed divided by the number coded disclosed or not disclosed. Clinics coded not confirmed are reported as a separate count and sit on neither side of the rate.

Authorship and independence. This report is credited to Turkey Hair Index Research, the organization's research desk, and no individual is named, so that clinics and doctors address the organization through the public contact page and corrections page rather than individuals. Turkey Hair Index has no commercial relationships with any clinic. No clinic paid for, requested, reviewed or approved this study, and no clinic was contacted for it.

Suggested citation: Turkey Hair Index Research, “Clinic Disclosure Study,” Inside Turkey's Hair Transplant Industry, October 17, 2026, https://turkeyhairindex.com/research/clinic-disclosure-study.

Selection criteria

Clinics with a completed Turkey Hair Index research dossier and a published THI Score on the freeze date.
Each clinic's official website, as listed in its Turkey Hair Index profile.
Standing English-language pages only: the home page, the pages it links to, relevant pages those link to, and relevant pages in the site's XML sitemaps.

Collection process

A script collected the pages on October 16, 2026, with a small number of additions on October 17. It identifies itself as TurkeyHairIndexResearch/1.0 with the address of the methodology page, reads robots.txt first and respects it, waits 0.8 seconds between requests to the same site, and reads up to 160 pages per site. Where a crawl reached that limit, the omitted addresses were reviewed and standing pages that the limit had left out were added.

Four sites refused the script. Their English menu pages were opened one at a time in an ordinary browser window on October 17, after robots.txt was read, and their text was not archived. For two of them the browser window showed a refusal earlier that day, and the site opened normally later. One of those two then limited the request rate after a few quick requests, so its pages were read more slowly. One site could not be read by either method: its server did not accept connections from our network. No workaround was tried, and this site is reported as not confirmed.

In total 1,437 pages were read: 1,359 by the script and 78 in the browser.

Validation process

Coding used patterns fixed before collection to retrieve the passages that could qualify. Every retrieved passage was read in context. Each coded cell records the page address, the check date and a short description of the statement. An automated check confirms that every cited address is among the collected pages with a matching date, and that a cell is coded not disclosed only for a site that was read completely.

After the first pass, the states of all 125 cells were frozen. A second reading of 25 cells chosen at random with a fixed seed matched the first pass in 25 cases; 2 of those cells belonged to sites that could not be read at the time. The second reading was done by the same AI-assisted workflow, was not blind to the first pass and used the same list of pages, so it shows repeatability, not independent reliability.

Because a missed statement is the error that matters most in a disclosure study, every cell coded not disclosed was then searched again with stricter and broader patterns, and every cell coded disclosed was checked against its cited page. The four sites read in a browser have no archived text. For two of them the cited pages were read again by hand and the daily-workload cells were searched again; the other two, read last, were read once. These audits changed 3 of 125 coded states, all from not disclosed to disclosed. The states as first coded stay frozen in the research record, so each correction is traceable.

Cells where reasonable readers could differ are marked as borderline, and the findings show how the counts would change if each were read the other way. Every figure on this page is computed by a script from the coded cells, and an automated test fails if a published number differs from the coded data.

Variables measured

Price stated: The standing pages state a hair transplant price in a named currency: a fixed price, a starting price, a range or a per-graft price. Average prices for other countries, prices for other services only, discounts without a total, a price given only on request, and prices that appear only in images, PDFs, forms, calculators or chat do not count. Unit: clinics.
Daily patient workload stated: The standing pages state how many patients the clinic, the lead doctor or the surgical team treats per day or at one time, as a typical figure, a maximum or a policy. Annual or cumulative totals, years of experience, wording such as “limited number of patients,” graft counts and staff-to-patient ratios do not count. Unit: clinics.
Surgical team described: The standing pages identify who works with the lead doctor on the procedure: named staff with roles, roles with numbers, or additional named surgeons who take part in hair transplant care. Wording such as “experienced team,” role categories without numbers or names, the lead doctor's biography alone, administrative or coordinator staff, and staff named only in testimonials do not count. Unit: clinics.
Follow-up schedule stated: The standing pages state a schedule of the clinic's own post-operative contact with at least two time points, or a recurring interval, in person, by phone, by video or by photo update. Generic assurances, a bare duration such as “12 months of follow-up,” patient recovery timelines and a single contact point do not count. Unit: clinics.
Surgical roles stated: The standing pages state, for at least one of four surgical stages (hairline design, creation of the recipient sites, graft extraction, graft implantation), who performs it: the doctor, other team members or both. Statements that the procedure is performed or led by the doctor without naming a stage, statements about technique, general rules about how clinics should be staffed, and roles in consultation or follow-up only do not count. Unit: clinics.
Form of the stated price: For clinics that state a price: fixed price, starting price, range or per-graft price. Where a page gives more than one form, the main form is recorded. Amounts are not compared and currencies are not converted. Unit: clinics.
Who performs each stage: For clinics that state at least one stage: for each of the four stages, whether the doctor, the team, both or nobody is stated. A statement that the doctor performs the whole procedure counts for all four stages. Unit: clinics.

Source model

Clinic-controlled information
Direct Turkey Hair Index research

Missing information

A site that could not be read is coded not confirmed on all five items. It is left out of the rates and reported as a separate count, never imputed and never counted as silent. Not disclosed is used only when a site was read completely and no qualifying statement was found on the pages in scope. It describes the pages, not the clinic's practice.

Deduplication

A statement counts once per clinic however often it is repeated, and a page that repeats text found elsewhere on the same site is not independent evidence. Where a site publishes separate English versions for different regions, both were read and a statement counted once.

How Turkey Hair Index Research Works

Findings

What the 24 readable clinics state

Of the 25 clinics in the study, 24 had English-language pages that could be read. The table shows, for each item, how many of those 24 state it on their own pages. “Surgical roles” means that a clinic states who performs at least one of four surgical stages. The 1 clinics whose pages could not be read are in neither of the first two columns and are counted separately.

Items stated by readable clinics
Item	State it	Do not state it	Share that state it	Not confirmed
Surgical roles	21	3	88%	1
Price	14	10	58%	1
Surgical team	13	11	54%	1
Follow-up schedule	11	13	46%	1
Daily patient workload	10	14	42%	1

21 of 24 clinics (88%) state who performs at least one of the four surgical stages; 3 do not.^[1][2]
Underlying data: Counts of the 24 readable clinics coded disclosed and not disclosed for each of the five items, and of the 1 clinics coded not confirmed, computed from the 125 coded cells
14 of 24 clinics (58%) state a hair transplant price; 10 do not.^[1][2]
Underlying data: Counts of the 24 readable clinics coded disclosed and not disclosed for each of the five items, and of the 1 clinics coded not confirmed, computed from the 125 coded cells
13 of 24 clinics (54%) describe who works with the lead doctor; 11 do not.^[1][2]
Underlying data: Counts of the 24 readable clinics coded disclosed and not disclosed for each of the five items, and of the 1 clinics coded not confirmed, computed from the 125 coded cells
11 of 24 clinics (46%) state a follow-up schedule; 13 do not.^[1][2]
Underlying data: Counts of the 24 readable clinics coded disclosed and not disclosed for each of the five items, and of the 1 clinics coded not confirmed, computed from the 125 coded cells
10 of 24 clinics (42%) state how many patients they treat per day; 14 do not.^[1][2]
Underlying data: Counts of the 24 readable clinics coded disclosed and not disclosed for each of the five items, and of the 1 clinics coded not confirmed, computed from the 125 coded cells
Interpretation
Daily patient workload is the item stated least often as coded. A daily figure is informative only when it is matched to the number of teams, overlapping cases and stage responsibilities behind it.

How much the coding choices matter

17 cells were borderline: a reasonable reader could have coded them the other way. The table shows each count as coded, and the count if every borderline cell had been read the strict way (a borderline disclosed cell becomes not disclosed) or the loose way (a borderline not-disclosed cell becomes disclosed).

Counts under three readings
Item	As coded	Strict reading	Loose reading	Borderline cells
Surgical roles	21	21	22	1
Price	14	13	14	1
Surgical team	13	10	18	8
Follow-up schedule	11	7	13	6
Daily patient workload	10	10	11	1

17 of the 125 coded cells are borderline: 8 coded disclosed and 9 coded not disclosed.^[1][2]
Underlying data: Counts of borderline cells per item and the disclosed count if every borderline cell were read the strict way or the loose way, computed from the coded cells
Between the strict and the loose reading the count for an item spans at most 8 clinics: 13 to 14 for price, 10 to 11 for daily patient workload, 10 to 18 for the surgical team, 7 to 13 for the follow-up schedule and 21 to 22 for surgical roles, of 24 readable clinics.^[1][2]
Underlying data: Counts of borderline cells per item and the disclosed count if every borderline cell were read the strict way or the loose way, computed from the coded cells

How many of the five items each clinic states

The 24 clinics with no unreadable cell are counted by how many of the five items they state. A clinic can appear in only one row.

Items stated per clinic
Items stated	Clinics
5	2
4	10
3	2
2	6
1	1
0	3

2 of 24 readable clinics state all five items.^[1][2]
Underlying data: Number of the five items stated by each of the 24 clinics with no not-confirmed cell, counted per clinic from the coded cells
10 of 24 readable clinics state four of the five items, the most common result; 3 state none.^[1][2]
Underlying data: Number of the five items stated by each of the 24 clinics with no not-confirmed cell, counted per clinic from the coded cells

How clinics state a price

A clinic can state a price in several ways. The study records the main form stated. It does not compare amounts, convert currencies, or treat a lower or a higher price as better or worse.

How clinics state a price
Group	Form	Clinics
State a price	Fixed package price	6
State a price	Per-graft price	4
State a price	Range	3
State a price	Starting price	1
Do not state a price	Say the price is given on request	7
Do not state a price	Do not mention a price	3

Of the 14 clinics that state a price, 6 state fixed package prices, 4 per-graft prices, 3 a range and 1 a starting price.^[1][2]
Underlying data: Form of the stated price for clinics coded disclosed, and whether a price on request is stated for clinics coded not disclosed, counted from the coded cells
Of the 10 clinics that do not state a price, 7 say that the price is given on request or after a consultation and 3 do not mention a price for hair transplantation.^[1][2]
Underlying data: Form of the stated price for clinics coded disclosed, and whether a price on request is stated for clinics coded not disclosed, counted from the coded cells

Which surgical stages clinics assign

Among the 21 clinics that state who performs at least one stage, the table shows who each says performs each of four stages. A statement that the doctor performs the whole procedure counts for all four.

Who performs each stage
Stage	Doctor	Team	Doctor and team	Not stated
Hairline design	16	0	1	4
Creating the recipient sites	16	0	0	5
Graft extraction	8	2	3	8
Graft implantation	4	7	1	9

Of the 21 clinics that state at least one stage, 8 state who performs all four, 3 state three, 7 state two and 3 state one.^[1][2]
Underlying data: Who is stated to perform each of four surgical stages, counted for the 21 clinics coded disclosed on surgical roles
16 of the 21 name the doctor alone for hairline design and 16 for creating the recipient sites; 4 do not say who designs the hairline and 5 do not say who creates the recipient sites.^[1][2]
Underlying data: Who is stated to perform each of four surgical stages, counted for the 21 clinics coded disclosed on surgical roles
For graft extraction, 8 name the doctor alone, 2 the team alone and 3 both, and 8 do not say. For graft implantation, 4 name the doctor alone, 7 the team alone and 1 both, and 9 do not say.^[1][2]
Underlying data: Who is stated to perform each of four surgical stages, counted for the 21 clinics coded disclosed on surgical roles

Istanbul and Ankara

The set has 19 clinics in Istanbul and 6 in Ankara. 18 and 6 of them could be read. Because both groups are small, the table shows counts only, with no percentages or significance tests, and the report draws no conclusion about differences between the cities.

Items stated, by city
Item	Istanbul	Ankara
Surgical roles	17 of 18	4 of 6
Price	12 of 18	2 of 6
Surgical team	11 of 18	2 of 6
Follow-up schedule	9 of 18	2 of 6
Daily patient workload	7 of 18	3 of 6

Clinics stating each item, as counts of readable clinics in Istanbul (18) and Ankara (6): surgical roles 17 and 4, price 12 and 2, surgical team 11 and 2, follow-up schedule 9 and 2, daily patient workload 7 and 3.^[1][2]
Underlying data: Counts of clinics coded disclosed and not disclosed in Istanbul and in Ankara for each item, computed from the coded cells

What clinics say when they state no figure

Not stating an item does not mean that a clinic says nothing about it. For every cell coded not disclosed, the coder also recorded whether the pages address the item in terms that do not qualify: a price given on request or a market range, “regular check-ins” without an interval, or “a limited number of patients” without a number.

Items addressed in general terms
Item	Do not state it	Address it in general terms
Surgical roles	3	3
Price	10	9
Surgical team	11	11
Follow-up schedule	13	13
Daily patient workload	14	7

All 13 clinics without a follow-up schedule and all 11 without a team description address the item in general terms. So do 9 of the 10 without a price (7 of them say that the price is given on request), 7 of the 14 without a daily patient count and all 3 without a surgical role.^[1][2]
Underlying data: Counts of not-disclosed cells for which the coder recorded that the pages address the item only in non-qualifying general terms

Coverage and checks

24 of the 25 sites were read: 20 by the collection script and 4 in an ordinary browser window after the script was refused. 1 could not be read and is reported as not confirmed, not as silent.^[1][2]
Underlying data: Sites and pages read, the comparison of the frozen first-pass states with the final states, and the agreement of the second reading, computed from the coding files
1,437 pages were read in total: 1,359 by the script and 78 in the browser.^[1][2]
Underlying data: Sites and pages read, the comparison of the frozen first-pass states with the final states, and the agreement of the second reading, computed from the coding files
A second reading of 25 randomly chosen cells matched the first pass in 25 of 25; 2 of those cells belonged to sites that could not be read at the time and are not confirmed by construction.^[1][2]
Underlying data: Sites and pages read, the comparison of the frozen first-pass states with the final states, and the agreement of the second reading, computed from the coding files
Audits after the second reading changed 3 of 125 coded states, all from not disclosed to disclosed. 1 of the 3 was in the second-reading sample, where both readings had missed the same page.^[1][2]
Underlying data: Sites and pages read, the comparison of the frozen first-pass states with the final states, and the agreement of the second reading, computed from the coding files

Surgical roles: 88%
Price: 58%
Surgical team: 54%
Follow-up schedule: 46%
Daily patient workload: 42%

Underlying data: Counts of the 24 readable clinics coded disclosed and not disclosed for each of the five items, and of the 1 clinics coded not confirmed, computed from the 125 coded cells, Counts of borderline cells per item and the disclosed count if every borderline cell were read the strict way or the loose way, computed from the coded cells

Interpretation

What the counts support. They describe what 24 clinics write on their own English pages. They support statements such as “most of the clinics read here name the doctor for hairline design” and “fewer than half state how many patients they treat in a day.” They do not support statements about the industry, about a named clinic or about quality.

What the counts do not show. A statement is what a clinic says, not a checked fact, and the study checks only that a statement exists on a page. Not stating an item is not evidence of a problem, and stating one is not evidence of quality. The study ranks no clinic and does not change any THI Score. THI Scores are assigned under the separate, published methodology after a dossier review, and commercial relationships never influence them.

For a reader comparing clinics. An item that is missing from a clinic's pages is a reasonable question to put to the clinic in writing before booking: how many patients the team treats in a day, who performs each surgical stage, who else works with the doctor, which follow-up contacts are scheduled, and what the price covers. The guides explain what to look for in the answers.

Where this fits. Clinic profiles show what each clinic has documented, with the evidence and its limits. This report adds a count across the researched set. A choosing guide ties the questions together.

Limitations

The 25 clinics are a selective, evidence-led set in two cities. They are not a random sample of Turkish clinics, and inclusion in the Turkey Hair Index is not an endorsement.
Effect on interpretation: The counts describe these clinics only. They cannot be generalized to the industry, to other clinics or to other cities.
The study measures what clinics state on their own pages. A statement is a clinic's own claim, not a checked fact, and a missing statement says nothing about how a clinic works.
Effect on interpretation: A clinic that states an item may not follow it, and a clinic that does not state it may answer it in writing or in a consultation.
Only standing English-language pages count. Blog posts, case pages, other languages, messaging and quotes obtained by inquiry are outside the study. In at least three cells the only statement found was on an excluded page, so the cell is coded not disclosed under the study's definition.
Effect on interpretation: The counts describe the standing pages a patient reaches from a clinic's English menu, not everything a clinic has published.
Coding was carried out by an AI-assisted research workflow that followed written rules. The second reading used the same workflow, was not blind to the first pass and used the same list of pages, so it shows that the coding is repeatable, not that it is correct or independent. An audit after it found 3 missed disclosures in 125 cells, one of them in the second-reading sample.
Effect on interpretation: Other missed statements may remain, which would make the counts too low. A human spot-check of a sample of cells would add independence and can be added.
17 cells were borderline readings of strict definitions, for example a follow-up described by one time point plus a period, or a team described by role without numbers or names.
Effect on interpretation: The table of strict and loose readings shows how far each count could move. For the surgical team the range is 10 to 18 of 24.
One site could not be read, and four more were read page by page in a browser window with their text not archived. Page text is not republished.
Effect on interpretation: The rates cover 24 clinics, not 25. The unread clinic is not described as stating or not stating anything.
The definitions were fixed before collection and are strict: a follow-up schedule needs at least two time points or a stated interval, a team needs names or numbers, and a daily workload needs a patient count rather than a graft count.
Effect on interpretation: Looser definitions would produce higher counts. The strict and loose readings show part of that effect.
The pages were read on October 16 and 17, 2026. Clinic websites change.
Effect on interpretation: The counts describe those two days and may not describe a clinic's pages later.
Only English-language pages were read. A clinic may state more in other languages or in direct communication.
Effect on interpretation: The counts describe public English pages only.
The city groups are small: 18 readable clinics in Istanbul and 6 in Ankara.
Effect on interpretation: City counts are shown without percentages or tests and support no comparison between the cities.

Research & Sources

Turkey Hair Index clinic disclosure coding records: 125 clinic-by-item cells for 25 clinics, collected October 16–17, 2026. A research record; the aggregate counts are published in the tables on this page — Turkey Hair Index · Published 2026-10-17 · Source family: THI Direct Research · Evidence role: Underlying Data
Official English-language websites of the 25 clinics in the study population, as listed in their Turkey Hair Index profiles · Accessed 2026-10-17 · Source family: Clinic Controlled · Evidence role: Supporting
Clinic Disclosure Study protocol, frozen October 16, 2026: population, definitions, pages in scope, retrieval patterns, coding rules and denominators. Held in the Turkey Hair Index research record and summarized on this page — Turkey Hair Index · Published 2026-10-16 · Source family: THI Direct Research · Evidence role: Supporting