Earned Media and AI Answers: A Guide for MENA PR Agencies

The client's CMO starts the monthly meeting with her phone in her hand. Over the weekend she asked an AI assistant: "What are the best dermatology clinics in Riyadh?" It named three clinic groups. Hers wasn't one of them. Under the second name sat a link to an Arabic article in a Saudi business publication, from April.

"We've had more coverage than them this year," she says. "So why them?"

Illustrative example: the clinic group, the question and the answer are invented.

The honest answer is uncomfortable for a PR agency, and useful. Answer engines repeat what credible third parties have already written. Coverage that answers a buyer's question gets reused; coverage that only announces something rarely does. This guide is for PR agencies in Saudi Arabia, the UAE and Egypt: why this happens, what to do differently in Arabic and English outlets, and a monthly check that shows the client where they stand.

1. Why the rival was named

Answer engines repeat what credible outlets already wrote.

The evidence points one way.

• Earned media dominates citations. Muck Rack's What Is AI Reading? study, May 2026 edition, looked at more than 25 million links cited by ChatGPT, Claude and Gemini across 17 industries. Earned media made up 84% of citations, journalism alone 27%, paid and advertorial content 0.3% (release, May 2026). More than half of the journalism cited was under a year old.
• There's no trick to it. Google says there are no additional requirements or special optimizations to appear in AI Overviews or AI Mode. A page has to be indexed and eligible to show in Search with a snippet, and the features use "query fan-out", running several related searches to build one answer (Google Search Central, updated December 2025).
• It's here, in Arabic. AI Overviews reached MENA, and Arabic globally, in May 2025 (Google, 2025).
• People click less when an answer appears. Pew Research Center tracked the browsing of 900 US adults in March 2025. With an AI summary on the page, they clicked a traditional result on 8% of visits, against 15% without one; links inside the summary got clicked on 1% of visits (Pew, July 2025). That's US data, but the direction is what matters.

Put together: for many category questions, the answer is now the shop window, and it's assembled from what other people published about you. That's earned media. It's a PR agency's home ground.

2. What changes for a PR agency in MENA

The coverage that counts answers a question a buyer would ask.

Most agencies already have the service. What's missing is the aim.

• Questions before announcements. "Clinic group opens third branch" is news for a day. "How to choose a dermatologist in Riyadh: what to ask before you book", with the client's medical director quoted, answers something people will type for years.
• Arabic and English are two different answers. Ask the same question in each language and you'll often get different names, cited from different outlets. An English win in a Dubai business title does little for an Arabic question asked in Jeddah.
• Local detail lives in local outlets. Global titles rarely write about a Riyadh clinic group. The specifics an answer needs sit in Saudi, Emirati and Egyptian publications and trade titles, so that's where the coverage has to land.
• Twenty copies are one story. A press release republished word for word on twenty Arabic news sites looks like twenty hits in the clippings report. It's one story, and usually an announcement. One original article that explains something tends to be worth more.
• Spelling splits the client in two. When the client's name is written three ways in Arabic and two in Latin letters, answer engines may treat it as several organizations. Pick one Arabic spelling and one English spelling, and use them everywhere.

3. The framework: five moves

Map the questions, earn the answers, fix the facts, check monthly, report it.

Google's own guidance supports move 3: keep structured data consistent with the visible text on the page, and keep Business Profile information current. Move 1 deserves a few trend questions too ("what's changing in aesthetic clinics in Saudi Arabia?"). Muck Rack found industry-trend queries drive journalism citations at more than twice the rate of how-to questions.

4. The monthly AI-answer check

Same prompts every month, in Arabic and English, with every cited source logged.

Set it up once, then run it in the first week of every month.

• Twelve prompts: six English, six Arabic. Write some of the Arabic in Modern Standard Arabic and some the way people actually type it. Illustrative prompts:
  • "What are the best dermatology clinics in Riyadh?"
  • "Which clinic in Dubai is best for laser treatment?"
  • «ما أفضل عيادات الجلدية في الرياض؟» (Modern Standard Arabic)
  • «وش أفضل عيادة جلدية في جدة؟» (Saudi)
  • «إيه أحسن عيادة جلدية في القاهرة؟» (Egyptian)
  • «شو أحسن عيادة ليزر في دبي؟» (Emirati)
• The engines your client's buyers use: usually ChatGPT, Gemini, Google AI Overviews or AI Mode, Copilot and Perplexity.
• Three runs per prompt, each in a fresh session, signed out where possible, noting the country you ran it from. Answers change from run to run. A single screenshot proves very little.
• One log, every month:

Then read the log for patterns. Outlets that keep getting cited become the pitch list. Wrong facts go to the fact sheet. The weaker language is where next month's coverage goes.

5. Worked example: one quarter for a clinic group

Illustrative: from 2 of 12 prompts to 7 of 12 in one quarter.

Illustrative example. Lumen Skin Clinics is a fictional clinic group in Riyadh, Jeddah and Dubai, and its agency is fictional too. Every number below is invented to show the method.

• Month 1, the baseline. Lumen is named in 2 of 12 prompts, both English. Two rivals appear in 9. The cited sources: two English regional business sites, a "top clinics" list on a Saudi lifestyle site, and an Arabic news story about a rival's new branch. Two answers get Lumen's founding year wrong, and the Arabic answers spell its name two different ways.
• What the agency does. Fixes the Arabic and English spelling everywhere, starting with the website and Google Business Profile. Places the medical director's explainer on choosing a dermatologist with a Saudi Arabic health and lifestyle outlet. Offers a short data story on summer skin-care questions to an English-language title in the UAE. Sends the Saudi site behind the "top clinics" list a fact sheet and a briefing.
• Month 3. Lumen is named in 7 of 12 prompts: 4 English, 3 Arabic. The Arabic explainer is cited in two Arabic answers. The founding year is right everywhere. Arabic answers asked from Dubai still skip Lumen, so next quarter's pitch plan targets Arabic outlets read in the UAE.
• In the client report. One slide: prompts where Lumen appears, month by month, by language, next to coverage and share of voice. The CMO watches an article she approved being quoted back by an answer engine. That's a retainer conversation.

6. Explaining it to clients without overpromising

Promise the work and the measurement, never a place in the answer.

• Report ranges, not a rank. Answers change between runs and between models, and Muck Rack's own report notes that citation behaviour may shift as models are updated. Show the trend over months.
• Be clear about control. No one can guarantee a mention. The agency controls the questions covered, the outlets reached, the facts kept straight and the check run on schedule.
• Keep it inside PR. Presence in answers is one more outcome of earned media, reported next to the coverage that caused it.

If the client wants the wider background, our explainer on what GEO is and our piece on whether a brand is visible in Arabic AI answers cover it. This guide is the agency's work plan. For where the answer slide sits in the monthly report, see the monthly client report, rebuilt.

What actually decides it

You can't shape what answers say in Arabic if you can't read what was written in Arabic.

The coverage that gets cited is spread across Saudi, Emirati and Egyptian publications, in Arabic and English. The conversation around it runs on social platforms in dialect, often with the client's name spelled three ways. Agencies that read all of it know which outlets matter, which facts are drifting and where a rival keeps getting quoted. dima reads online and print press, TV and radio, and social platforms in the dialect each mention was written in, and has captured 50M+ mentions. That's the specific problem it was built to solve.

When your client asks an assistant about their category in Arabic, do you know which article it will quote?

See which Arabic and English coverage is shaping your clients' categories, in one place.