Clickworkzz is a Dubai-based digital marketing strategist specialising in AI-first growth, GEO, and SEO for GCC businesses. led digital strategy for healthcare, real estate, and professional services clients across the UAE.

AI Search Optimization Dubai: What We Learned Testing 12 Businesses for ChatGPT & Google AI

We ran a structured AI Search Optimization (AISO) experiment across 12 Dubai businesses in 6 industries over 90 days. Businesses that implemented structured schema, conversational content, and authority signals became the ones ChatGPT, Gemini, and Google AI Overviews actually recommend. One client, Dr. Hafsa Ali Idrissi at Dubai Healthcare City, ranked for “best prosthodontist in Dubai” on ChatGPT within 28 days of implementation. Here’s the full breakdown.

Why We Ran This Experiment

If you’ve typed “best dermatologist in Dubai” or “affordable interior designer Dubai Marina” into ChatGPT or Google recently, you’ve noticed something: the same businesses keep appearing. Over and over.

It’s not random. It’s not just who has the most Instagram followers. And it’s definitely not traditional SEO working in some mysterious new way.

We wanted to know exactly what signals AI search engines use to decide who gets recommended and who disappears entirely.

So we ran a controlled experiment. 12 Dubai businesses. 6 industries. 90 days of systematic, documented changes, tracked against one question:

What actually gets a Dubai business recommended by ChatGPT, Gemini, and Google AI Overviews?

Here’s everything we found, including a detailed breakdown of how we got a Dubai Healthcare City dental specialist to rank for one of the most competitive healthcare keywords in the UAE in under 30 days.

The 12 Businesses (And Why We Chose Them)

We selected businesses across industries where Dubai consumers actively use AI tools to make decisions:

None had done any form of AI-specific optimization before we began. All had functional websites. Most had reasonable traditional SEO in place.

We split them into a control group (6 businesses, minimal changes) and a test group (6 businesses, full AISO implementation). Dr. Hafsa’s practice was our most closely monitored test case.

The AISO Stack: What We Actually Changed

We call our methodology the AISO Stack, five layers that work together. Implementing any single layer in isolation produced limited results. The compounding effect of all five is what drove the numbers.

Layer 1: Structured Data (Schema Markup)

The single highest-impact change across every business we tested, bar none.

AI models don’t browse your website the way a human does. They pull structured, machine-readable signals. If your data isn’t structured, you are effectively invisible to them regardless of how good your content is.

For each test business, we implemented:

• LocalBusiness / MedicalBusiness / LegalService / DentistSchema — telling AI exactly what the business is, what it does, who it’s licensed by, and where it operates
• FAQPage schema — pre-answering the specific questions Dubai consumers type into AI tools
• Review schema — surfacing verified Google ratings as machine-readable data
• Service schema with AED pricing context — critical for any service business targeting price-aware AI queries
• OpeningHours and ContactPoint — essential for “is X open now” and “how do I contact X” query types
• Physician / Dentist schema with specialisation markup — for healthcare clients

The result for Dr. Hafsa Ali Idrissi: Her practice had zero AI mentions at baseline when tested across ChatGPT, Gemini, and Google AI Overviews for prosthodontics-related queries. After implementing Medical Business schema with her DHA license number, Dentist + ProsthodonticSpecialty markup, and FAQPage schema covering implant, veneer, and smile design queries — she appeared in 7 out of 10 tested ChatGPT queries for “best prosthodontist in Dubai” within 28 days.

Layer 2: Conversational Content Architecture

Traditional SEO optimises for keywords. AI search optimises for answers.

AI tools retrieve information by matching a user’s question to the most direct, credible, locally-relevant answer they can find. Your content needs to be structured so AI can extract and cite it with confidence.

We rewrote key pages using a Question → Direct Answer → Evidence framework:

1. Identify the 15–20 questions your ideal Dubai customer is asking AI tools right now
2. Answer each one directly in the first sentence — no burying, no preamble
3. Follow immediately with proof: credentials, DHA/RERA licensing, case outcomes, before/after metrics

The Dubai context signal is non-negotiable. Explicit references to “DHA-licensed,” “Dubai Healthcare City,” “approved by RERA,” “serving JBR and JLT,” and pricing in AED dramatically increased AI citation rates — because AI models are trained to weight geographic and regulatory specificity in local queries.

For Dr. Hafsa’s practice, we rewrote her treatment pages to open with direct answers: “Dr. Hafsa Ali Idrissi is a DHA-licensed prosthodontist at Dubai Healthcare City specialising in dental implants, porcelain veneers, and full-mouth rehabilitation for patients across Dubai and the UAE.” Previously her page opened with clinic history.

The difference in AI citation pick-up was immediate and measurable.

Layer 3: Authority Signal Architecture

AI models are trust engines. They recommend businesses they have encountered across multiple credible, independent sources. A business that exists only on its own website, however well-optimised, carries far less weight than one whose expertise is corroborated externally.

The authority signals that produced measurable lift in our experiment:

Third-party citations:

• A single mention in Gulf News, Khaleej Times, or Zawya produced measurable lift in AI recommendation rates within weeks of the article being indexed.
• Official directory listings — DHA practitioner portal, Dubai Chamber, RERA, carried disproportionate weight, especially for healthcare and legal clients. These are high-trust domains that AI models treat as ground truth for professional credibility in the UAE.

Google Business Profile optimisation: This surprised even us. Businesses with fully completed GBPs, including accurate specialty categories, Q&A sections populated with real patient/client questions, regular posts, and high-quality photo updates, significantly outperformed incomplete profiles in AI-generated local recommendations.

One of our fitness clients in JLT doubled their AI recommendation rate after GBP optimisation alone, before any on-site changes were made.

NAP consistency (Name, Address, Phone): We found NAP mismatches on 4 of our 12 test businesses — small discrepancies between the website, GBP, DHA listing, and social media. These actively degraded AI entity confidence. Standardising them produced improvement within 30 days in all four cases.

Layer 4: Review Velocity and Sentiment Signals

AI tools consume review data heavily when constructing local recommendations. But it’s not just about having reviews, it’s about the shape of your review profile.

What our experiment showed:

• Recency matters more than volume. A business with 22 reviews in the past 90 days outperformed one with 180 reviews over 3 years in AI recommendation frequency.
• Review language is a signal. Reviews containing specific service terms, “Dr. Hafsa fitted my dental implants in Dubai Healthcare City” or “best prosthodontist I’ve visited in Dubai” were more impactful than five-star generic praise. AI models extract named entities and service references from review text.
• Owner responses are factored in. Google’s AI systems appear to treat owner response activity as a business engagement and legitimacy signal. Non-responding businesses performed consistently lower in our data.

We implemented a structured post-visit review request system for all six test businesses. Results: an average of 34 new reviews per business in 90 days, versus 4 in the control group. For Dr. Hafsa’s practice, 19 new reviews were gathered in the 28-day sprint, several explicitly mentioning “prosthodontist,” “Dubai Healthcare City,” and specific treatments.

Layer 5: Entity Consistency

The most technical layer, and the most frequently overlooked.

AI models construct a knowledge graph of your business as an entity: a cluster of consistent, corroborated signals that together confirm you are who you say you are, located where you say you are, offering what you say you offer.

Entity fragmentation , when your name, address, speciality, or credentials appear differently across platforms, causes AI models to lose confidence in recommending you. They default to recommending businesses whose identity is unambiguous.

We audited and standardised entity signals across:

• Website (About, homepage footer, all metadata)
• Google Business Profile
• DHA / RERA / Dubai Chamber directory listings
• Instagram, LinkedIn, and other social bios
• All schema markup
• Any third-party mentions or citations

Dr. Hafsa’s practice had her name formatted three different ways across platforms — “Dr. Hafsa,” “Dr. Hafsa Idrissi,” and “Dr. Hafsa Ali Idrissi” — and her speciality listed as both “Prosthodontics” and “Cosmetic Dentistry” in different places. After standardisation to a single consistent entity identity across all touchpoints, AI recommendation frequency for prosthodontics queries accelerated noticeably.

Case Study: Dr. Hafsa Ali Idrissi – “Best Prosthodontist in Dubai” on ChatGPT in 28 Days

This was our most closely tracked test case, and the one that best illustrates the compounding effect of the full AISO Stack.

The client: Dr. Hafsa Ali Idrissi, DHA-licensed prosthodontist at drhafsaalidrissi.com, practising at Dubai Healthcare City. Specialisations: dental implants, porcelain veneers, full-mouth rehabilitation, smile design.

The baseline (Day 0): Zero appearances across 10 tested ChatGPT queries related to prosthodontics in Dubai. Zero mentions in Google AI Overviews for any tested keyword. GBP 61% complete. No schema markup. Name inconsistencies across 4 platforms. 8 Google reviews (most recent: 4 months prior).

What we implemented (Days 1–14):

• Full MedicalBusiness + Dentist + ProsthodonticSpecialty schema with DHA license number
• FAQPage schema covering 8 high-volume questions (implants, veneers, costs in AED, Dubai Healthcare City location, appointment booking)
• Treatment page rewrites using Question → Answer → Evidence structure
• GBP audit: updated to 100% completion, corrected primary category to “Prosthodontist,” added 14 clinical photos, seeded 6 Q&A entries, enabled messaging
• Entity standardisation across website, GBP, Instagram, DHA portal
• Homepage meta rewrite: title and description targeting “prosthodontist Dubai” and “dental implants Dubai Healthcare City”
• Internal linking between treatment pages and the blog
• Post-appointment review request workflow activated

Results (Day 28):

New reviews in 28 days: 19 (vs. 0 in prior 4 months).

The key unlock: The combination of DHA-licensed entity schema + treatment-specific FAQPage markup + GBP category accuracy created what we call an authority triangle — three independent corroborated signals that AI models use to confidently identify and recommend a specialist. No single element achieved this alone.

The Results: Control Group vs. Test Group (90 Days)

Across all 12 businesses, after 90 days:

The control group didn’t just grow more slowly, several businesses in it effectively lost AI visibility as competitors in the same categories were optimised and the share of AI citations consolidated around the stronger entities.

What Didn’t Work (Saving You Time and Budget)

We’re as interested in the negatives as the positives. A few tactics that showed no meaningful impact:

❌ Posting more on Instagram and social media — Social activity had no direct effect on AI recommendation rates in our data. Social media supports brand discovery and follower growth separately, but it does not signal authority to AI search systems.

❌ Publishing more blog posts without structure — Volume without conversational architecture moved nothing. AI does not reward frequency. It rewards structured, locally-relevant, directly-answering content.

❌ Meta keyword tags — Irrelevant to both traditional and AI search in 2026.

❌ Paid advertising (Google/Meta) — Ad spend had zero effect on organic AI recommendation rates. AI search is not pay-to-play. A business spending AED 50,000/month on ads with no schema or authority signals was outperformed in AI search by a smaller competitor with a fully structured GBP and three DHA directory citations.

❌ Generic “about us” rewrites — Cosmetic copy changes without structural changes (schema, Q&A architecture, entity standardisation) produced no measurable improvement.

How to Apply This to Your Business: A Step-by-Step Checklist

You don’t need an agency to start. Here is the prioritised sequence based on what drove results in our experiment:

Week 1 — Foundation

• Audit your business name across all platforms — website, GBP, social, directories. Standardise to one exact form.
• Complete your Google Business Profile to 100% — photos, categories, Q&A, description, services list with AED pricing.
• Identify your DHA, RERA, Dubai Chamber, or relevant regulatory listing. Ensure it matches your website exactly.

Week 2 — Schema

• Implement LocalBusiness (or MedicalBusiness / LegalService) JSON-LD with your full address, phone, hours, and regulatory license number.
• Add FAQPage schema to your homepage and top service pages — minimum 5 questions your customers ask AI tools.
• Add Review schema pointing to your verified Google rating.

Week 3 — Content

• Rewrite your top 3 service pages using the Question → Direct Answer → Evidence structure.
• Include explicit Dubai/GCC context: neighbourhood, licensing body, pricing in AED.
• Add a “Who is [Your Name/Business]” paragraph to your homepage that reads like an answer to a ChatGPT query.

Week 4 — Authority & Reviews

• Set up a post-service review request (WhatsApp message or email) asking customers to leave a Google review mentioning the specific service and location.
• Identify one credible publication or directory where your business could be mentioned or listed. Pursue it.
• Add internal links between your service pages and blog content.

The Industries Where AISO Is Most Urgent in Dubai Right Now

Based on our experiment data and ongoing client work, these are the verticals where AI search is already driving significant discovery — and where not being optimised is costing real leads daily:

Act now:

• Healthcare specialists (DHA-licensed clinics, dental, dermatology, cosmetic surgery)
• Legal and financial advisory (DIFC and mainland)
• Real estate — brokerage, property management, and off-plan

Growing fast:

• Restaurants and hospitality (AI is the new “where to eat tonight” engine)
• Fit-out, interior design, and home services
• Education and tutoring

Early-mover window still open:

• Logistics and freight
• Automotive services
• Events and wedding planning

What This Means For Your Business

The businesses ChatGPT and Google AI recommend in 2026 are not necessarily the biggest, oldest, or highest-spending. They are the businesses that speak the language AI systems understand: structured, local, authoritative, consistent, and recently validated.

The window for early-mover advantage in Dubai is still open, but it is closing. As more consumers use AI as their first search step -bypassing traditional results entirely, the businesses that have built compounding AI authority will increasingly dominate recommendations, and the ones that haven’t will become invisible to a growing share of their potential customers.

Dr. Hafsa’s results in 28 days demonstrate what is achievable when the right signals are in place. The same approach applies across every industry, healthcare, legal, real estate, hospitality, and beyond.