Answer Engine & Generative Engine Optimisation
Your customers have started asking ChatGPT, Perplexity and Google's AI Overview for recommendations instead of scrolling a results page. We measured whether businesses across the Gulf and Africa are even capable of being the answer. Most are not — and the reason is mechanical, specific, and fixable.
Short answer: we audited 149 live company websites across Qatar, Saudi Arabia, Oman, Kuwait, Ghana, Côte d'Ivoire and Tanzania. 97% had no FAQ schema, 60% had no structured data of any kind, and only 18 published an llms.txt. The average AI-readiness score was 36 out of 100. When an answer engine has no machine-readable facts to lift, it names a competitor or a directory instead of you.
Every figure below comes from fetching the live site, its robots.txt and its llms.txt, then parsing the JSON-LD. Nothing here is estimated.
| Market | Sites audited | Avg score | No FAQ schema | No structured data | |
|---|---|---|---|---|---|
| Qatar | 33 | 38 | 91% | 61% | |
| Saudi Arabia | 34 | 39 | 100% | 53% | |
| Oman | 17 | 33 | 94% | 71% | |
| Kuwait | 29 | 35 | 100% | 59% | |
| Ghana · Côte d'Ivoire · Tanzania | 36 | 34 | 100% | 64% |
In Saudi Arabia, Kuwait and all three African markets, not one audited site had FAQ schema. That is the single strongest signal an answer engine uses to decide what to quote.
Organization and LocalBusiness JSON-LD, so an engine knows who you are, where you operate and what you sell.
FAQPage and Speakable schema — short, self-contained answers a model can lift without rewriting you.
Whether GPTBot, PerplexityBot, ClaudeBot and Google-Extended are allowed to read you at all.
A plain-language brief written for models rather than people, stating your facts so they are cited accurately.
The question answered in the first two lines, before the marketing copy — how snippets get won.
Meta descriptions, H1s, canonicals and sitemaps, because AEO sits on top of SEO rather than replacing it.
Dubai is saturated: dozens of agencies compete on the same terms and prices are bid down. The rest of the Gulf is not. In these four markets the audited baseline is close to zero, which means being first to structure your site properly is often enough to become the cited answer in your category — before the competition understands the category exists.
SEO gets you a blue link on a results page. AEO gets your business named inside the answer itself — the AI Overview, the voice reply, the featured snippet. The mechanics differ: answer engines need machine-readable structure and short, self-contained answers they can lift. A site can rank well on Google and still never be quoted.
GEO is being cited by generative models — ChatGPT, Perplexity, Claude, Gemini — when someone asks a question in your category. It needs three things: the AI crawlers allowed to read you, content parseable as facts rather than marketing prose, and a clear machine-readable identity. Most Gulf sites fail all three.
Usually because the model cannot parse who you are. With no Organization or LocalBusiness schema, no FAQ schema and no llms.txt, an answer engine has no reliable facts to quote — so it names a competitor that does, or a directory listing instead of you.
The technical layer is typically shipped in two to four weeks. Being picked up and cited follows re-crawling, which generally shows movement inside one to three months depending on how often the engines crawl your domain.
Yes. Send your domain and you get the same audit shown above, run against your site, with the specific gaps listed and what each one costs you. No charge and no obligation.
Free, no obligation
Send your website address. You get back the same audit — structured data, FAQ schema, crawler access, llms.txt — with your score out of 100, the exact gaps, and what each one is costing you in AI answers. If nothing is wrong, we will tell you that too.
Request my free audit →Method. 173 company websites were selected across the seven markets from verified business contact data; 149 responded over HTTP and form the sample. For each we fetched the homepage, /robots.txt and /llms.txt, parsed all JSON-LD blocks and typed them, and scored: structured data present (20), Organization/LocalBusiness identity (20), FAQPage schema (25), llms.txt (10), AI crawlers not blocked (15), meta description (5), H1 (5). Audited 12 August 2026. Individual companies are not named — the figures are reported only in aggregate.