Arabic + English Hook Bank for Dubai E-Commerce Video Ads

A Dubai skincare brand spent AED 40,000 on a bilingual product film last year. Beautiful footage, one Arabic voiceover, one English voiceover, the same edit. The English version reached a hook rate of 22% and the Arabic version 8%, and the agency's conclusion was that "Arabic doesn't work for this audience."

It was not the language. It was that the first line had been written in English for an English reflex, then translated into Arabic that was formally correct and socially wrong. The Arabic viewer met a sentence that no one from Dubai would say out loud, in the two seconds where the decision is made.

The mistake is translation, not language

A hook is not a sentence you move between languages. It is a local reflex. Arabic ad copy that reads fluent, warm and slightly indirect in a Gulf feed sounds vague when translated word for word into English — and English hook writing, which is front-loaded, blunt and often rude by Gulf standards, sounds like a discount flyer when translated into Arabic.

The result is two mediocre ads that share a production budget. One is too soft for a price-driven audience, the other is too aggressive for a trust-driven one, and neither names the objection the viewer is actually holding.

So stop translating. Write two hook banks from the same product truth.

Why the Arabic hook has to be written first

Most Dubai brands write in English, then localise. Reverse it and every part of the ad improves, because Arabic forces you to name the specific objection instead of the abstract benefit. English hook writing reaches for aspiration — "glow up your routine". Arabic hook writing, done well, reaches for the doubt — "هل هذا أصلي؟", is this the real one.

In a market where the first question is authenticity, delivery or price, the objection beats the benefit every time. Write the Arabic first and you find the objection for free. Write the English first and you inherit a benefit-led structure that has to be dismantled before it can be localised.

The hook bank

Twelve openers that work in a Dubai feed, with the objection each one answers. Treat them as templates, not scripts: swap the product and the specific.

Two things to notice. Every line is short enough to read in a muted feed, because most of these views will be sound-off on a phone held vertically at arm's length. And none of them describe the product. The hook's only job is to buy the next three seconds; the body does the selling.

Two seconds is the entire budget

Dubai feeds are dense, ad-saturated and thumb-driven. The realistic window before a scroll is under two seconds, which means the hook must survive three separate conditions at once:

• Muted. Your line has to work as text. If the hook only lands as audio, most of the audience never receives it.
• No context. The viewer has not read the caption, does not know the brand, and has not decided to care.
• One idea. Not a benefit and a price and a limited-time offer. One.

The read-aloud test catches most failures before you spend: say the line to a Gulf native speaker with no picture and no caption, and ask what is being sold and who it is for. If either answer takes more than a second, the hook is doing too much.

Dubai is four audiences, not one

The Arabic you choose tells the viewer which Dubai they are in. That is a targeting decision disguised as a copy decision.

• Gulf (Emirati and Khaleeji) dialect — reads local and carries weight for anything sold as premium, authentic or Dubai-specific. Highest trust per impression.
• Egyptian dialect — reaches the largest Arab expat buying segment in the UAE, and it is the warmest register in the region. Best for household, beauty and food.
• Levantine — reads stylish and modern; strong for fashion, jewellery and interiors.
• English-first expats — the fastest to close on price and delivery promises, and the cheapest to reach because English CPMs in the UAE are lower.

Run the same product with a Gulf hook and an English hook before you spend a dirham on anything cleverer. If the Arabic version wins on hook rate, you have found cheap reach for the whole catalogue. If the English version wins, your offer is price-led and your copy should stay blunt.

One product, twenty openers, no studio

Hooks are the cheapest asset in the pipeline, and AI generation makes them nearly free to version. The pattern that works:

1. Generate one clean product run — the product on a surface, in a hand, in use. Never re-generate the product in each variant, so it stays identical across every ad and the comparison stays honest.

2. Produce eight to twelve openers as separate two-to-three second clips against the same body and CTA.

3. Keep the body identical across every variant. If the body changes too, the test tells you nothing about the hook.

4. Ship as one campaign, one ad set per language, so dialect is the only variable moving.

The same discipline is set out in six hooks, one product: the weekly Meta test grid. The UAE difference is that the language dimension doubles the grid, so you cut the creative dimensions in half — two formats, not five — to keep the test readable.

Dialect, agreement and pronunciation: what translation always breaks

Since Arabic is the fastest way to lose a Dubai viewer, these three failures are worth naming precisely. None of them are vocabulary problems, which is why no translation tool catches them.

Dialect against market. Modern Standard Arabic is the register of news and formal retail. Gulf dialect is the register of conversation. An MSA voiceover on a casual product ad reads like a government announcement; a Gulf dialect voiceover on a bank's corporate film reads unserious. Both are errors, in opposite directions. The workhorse for paid social is mixed: a dialect hook to stop the scroll, MSA or light dialect through the body to keep it credible.

Gender agreement. Arabic verbs and adjectives agree with speaker and addressee. A female-sounding voice with masculine verb forms, or a script written for a mixed audience in a single gender, is immediately jarring — and it happens constantly when English scripts, where "you" carries no gender, are machine-translated.

Brand pronunciation. Arabic transliteration of a Latin brand name has more than one defensible rendering, and a model will pick one. If it picks the one your client does not use, you have produced an ad that misnames the product.

The fix is a native review pass: one native speaker with copywriting judgement listens without the script, marks problems by timecode, rewrites rather than just re-records, and signs off in writing. Thirty to sixty minutes per one-minute ad. It is the cheapest insurance in the entire production and the first thing cut when a pipeline is optimised for volume.

Judge on hook rate, then hold rate

Likes are noise in the Gulf, where the feed is generous with them. Two numbers decide whether an ad lives:

• Hook rate — three-second video views divided by impressions. Under 15% means the opener is the problem, not the offer. Rewrite the first line before touching the body.
• Hold rate — those three-second viewers who reach fifteen seconds. Under 40% means the ad promised something the body did not deliver.

Run each hook for AED 100-150 of spend or three days, whichever comes first, and kill on the number rather than the comment section. Arabic comments are frequently more useful than the metrics for finding product objections — read them, but do not let them set the budget.

Five failures that kill Dubai e-commerce ads

1. One bilingual cut. Two languages in one edit means neither has a native hook, and the ad performs at the average of its two weakest moments.

2. A benefit where an objection belongs. "Premium quality" answers nothing. "Same quality, half the price, here's the proof" answers the doubt that was already there.

3. Generated product footage. If the model has invented the packaging, the ad is misrepresenting the product and the returns will say so.

4. Formal Arabic on a casual product. The fastest way to sound like a government notice in a beauty feed.

5. Testing twelve things at once. Language, hook, format, offer and audience all moving together produces a result nobody can act on.

The cheapest win most Dubai brands skip

Take your five best-performing English ads from the last quarter and rewrite only their first line in Gulf Arabic, on the same body. No new shoot, no new product generation, no new edit — five new openers, one week of spend. That single change is the cheapest test available to a Dubai e-commerce brand, and it is the one most have never run, because localisation usually starts at the script instead of the hook.

If you are still choosing a production partner rather than writing hooks in house, ask how they produce Arabic hooks and who reviews the dialect — the second half of the Dubai buyer's checklist. If you are already producing daily reels, this is the layer that makes daily volume worth the money: volume without hook variety is the same ad repeated, which is the arithmetic in Dubai retail's shift to daily AI reels.

This week

Write five Gulf Arabic openers and five English openers for one product. One campaign, two ad sets, AED 1,200 total. At the end of seven days you will know which language owns your audience, which is a decision most Dubai brands are still making with an opinion.