Quick answer
Meta Ads audit checklist for UAE accounts
A Meta Ads audit covers Pixel + Conversions API setup, attribution settings, campaign architecture, audience structure, creative testing process and reporting. Most UAE accounts fail at least one of the first two.
1. Pixel + Conversions API
Pixel installed correctly, server-side container (preferably stape.io or GCP), CAPI events deduplicated against pixel, event match quality 7+ for key events, offline events wired from CRM.
2. Campaign architecture
Advantage+ Shopping where feeds are healthy, prospecting / lookalikes / retargeting layers clear, audiences not over-fragmented, budget set at campaign level (CBO) where appropriate.
3. Creative testing
Structured hook/visual/CTA matrix, weekly drops, killing criteria documented, native vertical assets, bilingual where the audience requires it.
4. Audiences
Broad targeting tested, lookalikes from high-quality seeds, interest stacks tested vs broad, exclusions clean.
5. Reporting
CAPI-aware reporting, blended ROAS including platform + GA4 + offline, weekly creative scoreboard.
Events Manager checks you can finish in 20 minutes
Most Meta audits find their biggest problems inside Events Manager, and the checks take less time than any other part of the process. Open the pixel or dataset and work through the items below with the site live in another tab. What you are confirming is that every conversion the business cares about reaches Meta once, from both browser and server, with enough customer information attached for Meta to match it to a user. A dataset that passes these checks gives the optimisation model a clean signal; one that fails them wastes budget no matter how good the creative is. Write down the event match quality score for the lead or purchase event before any fixes are made, so the improvement can be shown later. In UAE accounts, the most frequent failures are a lead event firing on button click instead of successful submission, phone numbers passed without the country code, and server events arriving without the event ID needed for deduplication.
- Domain verified and the pixel owned by the business's own Business Manager, not an agency or freelancer.
- Both browser and server events present for the key event, with a deduplication rate above 90 percent.
- Event match quality of 7 or higher on lead and purchase events.
- Customer information parameters include email, phone with country code, first name and external ID.
- Test events show the correct event firing once per real submission on the live site.
- Offline event set or CRM integration receiving qualified and closed stages at least weekly.
- Custom conversions and standard events are not both counting the same action as a primary result.
Reading the account structure: learning phase, fragmentation and budget
The structure section of an audit answers one question: is each ad set receiving enough conversions to let Meta optimise, or is budget spread so thin that everything stays in learning? Open the ad set view, add the delivery column, and count how many ad sets show learning limited. Meta needs roughly 50 optimisation events per ad set per week to exit learning; at a CPL of AED 100, that means AED 5,000 a week per ad set, which is why a AED 20,000 a month account can barely sustain a single ad set, let alone six. Fragmentation usually comes from well-meaning segmentation: one ad set per emirate, per gender, per interest, or per age band. Consolidate until each ad set clears the threshold, and let placement and audience expansion do the segmentation. Check that campaign budget optimisation or Advantage+ budget is on where ad sets share a goal, that retargeting audiences are excluded from prospecting ad sets, and that the campaign objective matches the business outcome. A leads campaign optimising for link clicks because conversions were too few is a common and expensive compromise.
Creative audit: the metrics that predict cost before it moves
Creative is the largest cost lever on Meta, so the audit needs more than a glance at the ads. Build a creative scoreboard for the last 60 days with one row per ad and the columns below, then sort by spend. The pattern to look for is concentration: in most UAE accounts, the top three ads carry 70 percent or more of spend, which means the account is one fatigue cycle away from a CPL spike. Check how many new ads launched each week; fewer than three in a lead generation account usually signals a creative pipeline problem. Compare hook rate and hold rate across formats to find which openings earn attention and which lose it, and compare cost per lead by format to see whether short vertical video, static image or carousel is winning. Review the language mix. Arabic-first audiences in Sharjah, Abu Dhabi and Saudi Arabia respond poorly to English-only ads, and bilingual accounts that have not tested Arabic creative are leaving the cheapest audiences untouched.
- Hook rate: 3-second video plays divided by impressions; under 25 percent usually means the opening is weak.
- Hold rate: ThruPlays divided by 3-second plays, which shows whether the ad keeps attention.
- Link CTR, not all-clicks CTR, which includes likes and profile taps.
- Frequency on prospecting ad sets; above 3 in a 7-day window signals fatigue.
- Cost per lead and cost per qualified lead by ad, pulled from the CRM where possible.
- Days live per ad and the week-on-week CTR trend, to catch decline before CPL rises.
Attribution settings and how to cross-check reported results
Meta's reported results depend on the attribution setting, and an audit must state which one was used. The default of 7-day click and 1-day view counts a lead if someone saw an ad, did not click, and submitted a form the next day. For UAE lead generation that often inflates results, because retargeting ads are shown to people who were already on their way to converting. Use the compare attribution settings feature to show results under 7-day click, 1-day click and 1-day view side by side for the last 30 days. If 1-day click captures less than 60 percent of the 7-day click total, a large share of results are slow or loosely attributed and should be verified against the CRM. Then reconcile: count leads in the CRM by source for the same period and compare with Meta's figure. A gap of 10 to 20 percent is normal; a gap above 40 percent in either direction points to a tracking fault or an attribution setting that is too generous. Finally, check whether bidding is optimising on the same event the reports use, and whether the attribution setting changed mid-period, which breaks trend lines.
UAE-specific checks most templates miss
Generic audit templates were written for US and UK accounts and skip the settings that matter most in the Gulf. Location targeting is the first. Many UAE accounts target the whole country when the business only serves Dubai, or target Dubai with people recently in this location and collect tourists; choose people living in this location for services. Language targeting is the second: an account with no Arabic ad sets in Abu Dhabi, Sharjah or Saudi Arabia is paying English-market CPMs for an audience it half reaches. Lead quality controls are the third. Instant forms should use the higher intent form type and at least two qualifying questions, and WhatsApp campaigns should land in a business account with a CRM integration, not a personal number. Fourth, check payment and account health: a Business Manager with a single admin, a card about to expire, or an unresolved policy flag is a risk to delivery that an audit should name. Finally, review scheduling against the follow-up team's hours; leads generated at midnight and contacted at noon convert poorly, and the account should either pace toward business hours or have an auto-reply that sets expectations.