What's the difference between a one-off audit and a monthly retainer?

A one-off audit is a single wave: you ask the AI engines a fixed set of buyer questions once and get a snapshot of where you and your competitors stand. A monthly retainer repeats that same question set on a cadence, so you can see whether the fixes you made actually moved the answers, and whether a competitor's move pushed you out. AI answers shift as models update, so a single snapshot ages fast without a retainer behind it.

Why do AI visibility platforms cost so much more per month than an audit?

Enterprise AI-visibility software platforms typically list from around USD 500 to USD 3,000 or more per month, scaling with prompt volume, engines tracked, and seats. That's software you run yourself - dashboards and prompt tracking, not a scoped diagnostic with a human reading the answers and a fix plan attached. It solves a different problem than a one-off or retained audit, and the two aren't always substitutes for each other.

How do I know if a cheap AI visibility audit is worth paying for?

Ask whether every number in the report can be re-derived: can you see the exact questions asked, the exact engine, the exact answer text, and recount the score yourself? If the provider can't show you that, you're not buying measurement, you're buying a score nobody, including them, can reproduce. A frozen question set and per-engine, re-checkable answers are the minimum bar for a report to be worth the invoice.

The honest number

How much does an AI visibility audit cost in Dubai?

Q: How much does an AI visibility audit cost in Dubai?

One-off AI visibility audits in the UAE market typically run AED 8,000-40,000, depending on how many markets and engines are covered and whether competitor benchmarking and a fix plan are included. Monthly measurement retainers, which re-run the question set on a schedule, run roughly AED 4,000-20,000 a month. The low end buys a single-market snapshot; the high end buys multi-market coverage with named competitor positions and re-measurement built in.

The price range is wide because the work isn't standard yet. Here's what you should actually be paying for, and what the number should buy you.

28 Labs · September 2026

One-off AI visibility audits in the UAE run roughly AED 8,000-40,000, depending on scope. The low end is a single-market snapshot; the high end is multi-market, multi-engine, with named competitor benchmarking and a fix plan attached. Monthly measurement retainers, which re-run the same question set on a schedule, run roughly AED 4,000-20,000 a month. Enterprise AI-visibility software platforms are a different category entirely - they typically list from around USD 500 to USD 3,000+ per month for self-serve dashboards, not a scoped diagnostic.

Why the range is so wide

There's no rate card for this yet, because there's no standard unit of work. An audit that asks 20 questions in one language on one engine is not the same job as one that asks 200 questions across three markets on four engines with competitor positions logged for every answer. Both get called "an AI visibility audit." Only one of them tells you anything you can act on.

Treat the price as a proxy for scope, not a proxy for quality. A provider can charge AED 35,000 and still hand you a screenshot deck. The number alone doesn't tell you what you're getting - the next section does.

What actually drives the price

Number of markets and languages

A UAE-only, English-only audit is a different job from one covering Arabic queries or a second GCC market. Each market roughly doubles the question set.

Question-set size

More real buyer questions means more coverage of how people actually ask, and more cost to run and read every answer.

Number of engines

ChatGPT, Claude, Gemini, and Perplexity retrieve differently and disagree often. Covering one engine is cheap and tells you almost nothing about the others.

Competitor benchmarking depth

Logging which named competitors show up, in what order, and why, takes more analyst time than just checking whether your brand appears at all.

Fix plan and re-measurement

A report that stops at "here's your score" costs less than one that ends with a prioritized fix list and a second wave to confirm the fixes worked.

One-off vs wave-over-wave

A single snapshot is cheaper per unit than a retainer, but it ages the moment a model updates. Retainers price in the re-runs.

What a real audit has to include

Below AED 8,000, most of what's sold as an "AI visibility audit" is a general-knowledge check: someone typed a handful of prompts into ChatGPT once and screenshotted the results. That's not measurement, it's an anecdote with a logo on it.

A report worth paying for gives you:

A frozen question set of real buyer questions - not generic prompts, the actual phrasing people use when they're close to a purchase decision.
Per-engine answers you can re-check yourself - the exact question, the exact engine, the exact response, timestamped, so you can rerun it and see if it's changed.
Named competitors' positions - not just "you appeared" or "you didn't," but who the engine recommended instead, and how often.
Source and citation data - which pages the engine pulled from to build the answer, so you know what's actually feeding the recommendation.
Honest methodology limits stated up front - what the measurement can and can't tell you, including the difference between what a model was trained on and what it retrieves live. Anyone who claims to measure the former directly is overselling.

If the price seems too goodAsk whether you can re-derive every number in the report - the exact question, the exact engine, the exact answer, recounted by you. If you can't, you didn't buy measurement. You bought a mood board with a score on it.

The buying checklist

Ask any provider these five questions before you pay for an audit or sign a retainer:

Is every number re-derivable? Can you hand me the exact question, engine, and answer behind each count?
Are the denominators frozen? Does "visible in 40% of runs" mean the same 40% next wave, or does the question set quietly change?
Which engines are covered, and how many? One engine is a data point. Four engines is a picture.
What happens after the audit? Do I get a prioritized fix list, or just a score and a goodbye?
What are the contract terms? Is this a one-off, or does it auto-renew into a retainer? What's the cancellation notice?

If a provider hesitates on the first two, stop there. Re-derivability and frozen denominators are the difference between a measurement and a marketing claim.

Why we don't publish a rate card

28 Labs quotes a fixed price after a short scoping call, not off a published rate card. The reason is the same one driving the wide market range above: market count, language count, engine coverage, and competitor benchmarking depth change the actual work by a factor of five or more between the smallest and largest scope. A rate card that fits both would either overcharge the small scope or undercharge the large one. A short call gets you an honest number for your actual scope, not an average.

What we do about it

The question underneath "how much does this cost" is really "what am I getting." We build a frozen question set of real buyer questions for your market, run it across ChatGPT, Claude, Gemini, and Perplexity, and hand back a report where every count traces to a specific question and answer you can re-check yourself - named competitor positions included. For the mechanics of what "visible," "recommended," and the other terms in a report like that actually mean, see our AI visibility metrics glossary. If you're comparing providers before you commit, our AEO agency buyer's guide for Dubai covers the same re-derivability test in more depth.

28 Labs measures how AI engines answer real buyer questions - and what moves the counts. Every number we publish re-derives to specific questions and engines. try28labs.com