Which overseas education consultancies do AI assistants actually recommend when a parent in the Gulf asks which one to trust? This is the first attempt to measure it: 25 firms, 25 buyer prompts, four engines, one ranked table.
The scan has not been run yet. Every figure on this page is illustrative sample data, marked as such, showing the format the published Index will take. No firm named below is real, and no ranking here reflects any real consultancy’s performance.
Written to be quoted. Each stands alone.
Across 25 buyer prompts and four AI engines, the average AI answer named 4.2 study abroad consultancies. A conventional search results page lists ten. The shortlist has narrowed by more than half at the point of discovery.
The top three consultancies in the UAE captured 62% of all citations across the prompt set. The remaining 22 firms in the Index shared the other 38% between them.
Nine of the 25 consultancies measured did not appear in a single AI answer across 300 prompt runs. Several of them rank on the first page of Google for the same queries.
Published in full, including what it cannot tell you. A research asset that hides its limitations is a marketing asset.
Twenty-five prompts written the way a parent or student in the UAE actually types them — not keywords. They cover choosing a consultancy, comparing named firms, country-specific admissions help, visa support and cost. The full prompt list is published with the Index so anyone can repeat it.
ChatGPT, Perplexity, Google AI Overviews and Claude. Each prompt was run three times per engine from a UAE network location, with no personalisation, no logged-in account history and no memory carried between runs.
Citation share is the percentage of prompt runs in which a consultancy was named, out of 300 total runs. A mention counts once per run regardless of how often the name appears in that answer. Position within the answer is recorded separately and does not affect citation share.
Twenty-five overseas education consultancies with a physical presence in the UAE, identified from directory listings, trade association membership and the answers themselves. Inclusion is not an endorsement and exclusion is not a judgement.
AI answers are not deterministic. The same prompt produces different answers on different runs, which is why each was run three times — but three runs is a small sample and the margin of error on any individual firm’s score is meaningful. Results are time-stamped, not permanent: engines update their models and their sources continuously, and a scan run a month later will differ. Location and personalisation matter: these results reflect a UAE network location with no account history, which is not how every buyer searches. This measures citation, not conversion. Being named by an AI assistant is not the same as winning the client, and nothing here should be read as a revenue claim.
Percentage of 300 prompt runs in which each firm was named by at least one AI engine. The bars fade with rank — by the bottom of this table, a firm is effectively invisible to an AI assistant.
Scroll the table sideways to see all columns →
| Rank | Consultancy | Citation share | Visibility | Engines naming | Best position |
|---|---|---|---|---|---|
| 01 | Consultancy A | 41% | 4 / 4 | 1 | |
| 02 | Consultancy B | 33% | 4 / 4 | 1 | |
| 03 | Consultancy C | 28% | 4 / 4 | 2 | |
| 04 | Consultancy D | 24% | 4 / 4 | 2 | |
| 05 | Consultancy E | 19% | 3 / 4 | 2 | |
| 06 | Consultancy F | 17% | 3 / 4 | 3 | |
| 07 | Consultancy G | 15% | 3 / 4 | 3 | |
| 08 | Consultancy H | 13% | 3 / 4 | 3 | |
| 09 | Consultancy I | 11% | 3 / 4 | 4 | |
| 10 | Consultancy J | 9% | 2 / 4 | 3 | |
| 11 | Consultancy K | 8% | 2 / 4 | 4 | |
| 12 | Consultancy L | 7% | 2 / 4 | 4 | |
| 13 | Consultancy M | 6% | 2 / 4 | 4 | |
| 14 | Consultancy N | 5% | 2 / 4 | 5 | |
| 15 | Consultancy O | 4% | 2 / 4 | 4 | |
| 16 | Consultancy P | 4% | 1 / 4 | 4 | |
| 17 | Consultancy Q | 3% | 1 / 4 | 5 | |
| 18 | Consultancy R | 3% | 1 / 4 | 5 | |
| 19 | Consultancy S | 2% | 1 / 4 | 5 | |
| 20 | Consultancy T | 2% | 1 / 4 | 5 | |
| 21 | Consultancy U | 1% | 1 / 4 | 5 | |
| 22 | Consultancy V | 0% | 0 / 4 | — | |
| 23 | Consultancy W | 0% | 0 / 4 | — | |
| 24 | Consultancy X | 0% | 0 / 4 | — | |
| 25 | Consultancy Y | 0% | 0 / 4 | — |
Engines naming: how many of the four engines named the firm at least once. Best position: the highest place the firm reached within an answer that listed multiple consultancies.
The Index measures 25 consultancies against a shared prompt set. The diagnostic measures your consultancy against the competitors you name, and fixes what it finds within 14 days.
One or two factual lines per firm, published with the real Index. The notes describe what the measurement found, not what the firm is worth — every consultancy in this table is doing real work for real students.
Named by all four engines. Carries complete Organization and LocalBusiness schema, a named leadership team with individual pages, and consistent citation across three regional education directories. Sample note; real per-firm notes publish with the Index.
Strong on country-specific prompts, absent from general “best consultancy in Dubai” prompts. The pattern suggests good topical content and a weak entity definition. Sample note; real per-firm notes publish with the Index.
Ranks on the first page of Google for four of the same queries but was named in six of 300 AI runs. A conventional search position did not transfer to the AI answer. Sample note; real per-firm notes publish with the Index.
Three things follow from the table, and only one of them is comfortable.
Firms ranking on page one of Google were absent from AI answers for the same queries. The two systems select on different signals, and being good at one predicts very little about the other.
The firms at the top of the table are not writing better copy. They are legible to a machine: defined as entities, described in structured data, and cited consistently in the sources engines trust.
Entity definition, schema markup and crawler access are technical work with a defined end point. That is what the 14-day sprint does.
Nothing here requires proprietary tooling. Take the 25 published prompts, run each three times in ChatGPT, Perplexity, Google AI Overviews and Claude from a UAE network location with personalisation and chat history switched off, and record every consultancy named in each answer. Citation share is the count of runs naming a firm divided by 300, expressed as a percentage.
If you repeat the scan and get materially different numbers, that is useful and I want to see it — write to contact@growthcliq.com. The Index is updated annually at this same URL rather than republished at a new one, so corrections and later editions accumulate here.