Persian SEO: search realities for clinics targeting Iran
Persian SEO for clinics attracting patients from Iran: the role of organic search when ads are closed, Persian query patterns, hreflang and accessibility.
By Roozbeh Nazari · CEO
Patient demand from Iran is, in most of the clinics we work with, the second most frequently raised topic after the Arabic-speaking markets; yet the Persian pages are usually the last language version to open and the least maintained. This article explains why Persian SEO for the Iranian market is a different job. Let us say it upfront: we give no search volumes, patient counts or growth rates here. In the environment where these lines were written we had no access to a tool that returns search volumes, and we do not write a number we cannot verify. What we describe are decisions you can repeat with your own data and the behaviour patterns we see in the field as an agency.
Ads closed, organic open
The first fact that sets the Iranian market apart is that paid search is not on the table. Google's own help page states that, because of US sanctions, Google Ads is not available to advertisers in Iran and Iran cannot be selected as a campaign target. So the "bridge with ads until organic arrives" plan you are used to in Arabic markets does not work here. Demand comes on its own; if there is no page to meet it, it goes to another clinic through recommendations, intermediaries and messaging apps. That is why, for Persian pages, organic search is not a "nice to have" for clinics but the only scalable channel.
The second consequence is a change in measurement expectations. With no ads, the comparison points you are used to, cost per click and cost per conversion, do not exist; the performance of Persian pages is read only through organic impressions, clicks and the contacts that later turn into conversations. That measurement setup has to be in place before the pages open; we come back to it below.
The structure of the Persian query
Persian is written in the Arabic script but it is not Arabic; the technical traps of letter variants, code points and the zero-width non-joiner were covered in our keyword research in four languages article and we do not repeat them here. This article is about the query itself.
Our first observation is that the user in Iran often builds the query not as the treatment name alone but as treatment plus city or treatment plus country; the words "Istanbul" and "Turkey" appear inside the query more often than the clinic's name. That means the city name, in its Persian spelling, belongs in page titles and in the first paragraph. Second, the weight of cost queries: because of payment and currency constraints, queries built with هزینه (cost) and قیمت (price) form a distinct group in our own data. Meeting them requires a Persian version of the price page and a plain explanation of the payment method; we described a regulation-compliant price page structure in the price pages article.
Third, Persian written in Latin letters, so-called "Finglish". Users who do not type Persian on a mobile keyboard, who live abroad or who simply type in Latin letters out of habit search the treatment name in Latin script. Rather than opening a separate page for these queries, it is enough to include the Latin spelling of the treatment name once, in a natural sentence, on the existing Persian page. Fourth, the Persian versus Dari distinction: a user from Afghanistan uses the same alphabet but some treatment terms and everyday words differ; writing the page in Iranian Persian without opening a separate Dari page, while not stating a region in the language code so that neither audience is excluded, is a practical middle path.
The technical layer: hreflang, targeting, access
The language code decision follows from that last point. Google's documentation on localized pages says the hreflang value can be given as a language code alone, without a region; using "fa" instead of "fa-IR" for Persian pages connects the Persian-speaking audience outside Iran to the same page. The same document recommends the x-default value for the page shown when no language matches; letting the English version take that role is the right call on most clinic sites.
On country targeting, Google's multi-regional site documentation explains that a country-code domain is a strong signal, but that structures such as subfolders and subdomains also work with hreflang and on-page signals. A separate domain for Iran is not practical; an /fa/ folder under the main domain, hreflang tags and Persian contact details on the page provide enough signal. Right-to-left writing, direction attributes and typeface decisions were collected in the technical SEO for RTL sites article.
The real overlooked issue is accessibility. Because of sanctions, some hosting providers, content delivery networks and third-party services block requests originating from Iran; the result is a Persian page that is visible on Google but does not open from Iran. Search Console does not show this, because Googlebot sees the page without any problem. Before launch, verify that the page opens from Iran with a real user or a testing tool located in that country, and check that forms and messaging buttons work. For the same reason page weight matters: on a page opened over a distant, constrained connection, large images and excess scripts produce a visit that never turns into a contact.
Measurement: which traffic is really from Iran
Search Console's performance report allows filtering by country; the first reading of Persian pages is done there, with the country set to Iran and the page filtered to /fa/. But there is a distortion we see in the field: because of connectivity restrictions, some users in Iran appear to come from other countries. So the country filter should not be read on its own but together with the language dimension in GA4; sessions whose browser language is Persian are the audience of the Persian pages whatever the country shows. We cannot say how much shifts this way without looking at your site; in your own data, placing the two dimensions side by side makes the gap visible immediately.
The second measurement layer is contact. We observe that the user from Iran moves to a messaging app rather than filling in a form; so every messaging button on the Persian pages needs a click event and the source of the conversation needs to land in the CRM. Setting this up is part of the analytics and data work; in a market where no channel comparison is possible, the only comparison point is which Persian page turns into a conversation.
A note on content and regulation
The new text of the Regulation on Promotion and Information Activities in Health Services, published on 12 November 2025, requires that promotion aimed at international health tourism be run through a separate website or social media account directed abroad and that the health tourism authorisation certificate be published on the site. Building this structure from the start when planning Persian pages is easier than migrating later. In the content itself the rule does not change: no promises about treatment outcomes, an explanation of mechanism and process instead; and on every treatment page a clear statement that results vary from person to person.
Conclusion
What makes Persian SEO different in the Iranian market is not the language but the channel structure: ads closed, organic search the only scalable channel, access and measurement distorted in ways you do not encounter in other markets. A Persian version that meets the query on the city and cost axis, is tagged with the "fa" code, is verified to open from Iran and measures messaging contacts is the frame of what can be done in this market. The numbers are in your Search Console and GA4 data; in this article we explained how to read them rather than inventing them.