Moving Talabat and Deliveroo customers to WhatsApp
A UAE restaurant can pay an aggregator commission on every order for three years and still not know a single customer's name. That is not a pricing problem. It is an ownership problem, and it has a narrow but real solution.
Short answer. You cannot take the aggregator's customer data, and you should not try. What you can do is give the customer a reason to start a conversation with you — a QR on the packaging that opens WhatsApp. They initiate, which opens a free 24-hour window and gives you consent you can actually evidence.
What the commission actually is
Published UAE ranges: Talabat commonly around 15–30% and Deliveroo around 25–35%, varying by partnership type, volume, location and category. And the headline rate is not the whole cost — service fees, featured placement, peak-hour surcharges and in-app advertising push the effective take rate meaningfully higher.
But the number is not the real issue. Plenty of restaurants would happily pay 30% for genuinely incremental demand. The issue is that the same customer, ordering from you for the twentieth time, still costs 30%, because the relationship belongs to the platform. You are paying an acquisition fee on a retained customer.
What you cannot do
Being direct about this, because a lot of advice in this space is quietly reckless:
- Do not scrape numbers from order sheets or delivery labels. Even where a number is visible, using it for marketing is a consent problem, and it is usually a breach of your platform agreement.
- Do not upload aggregator order data as a marketing list. Those customers never opted in to hear from you.
- Do not expect it to go unnoticed. Blocks and reports degrade the quality rating on your WhatsApp number, and that number is the one you use for everyone else too. See bulk WhatsApp and UAE law.
The commercial argument against it is stronger than the legal one: a cold message to someone who ordered biryani through an app converts badly and costs you the number you send everything else from.
What works: give them a reason to scan
The mechanism is simple and the execution is where it fails. A QR code in the bag that says "follow us" gets scanned by nobody. A QR that offers something the customer wants right now, while the food is in front of them, gets scanned. Ranked by what actually works:
- Something about this order. "Scan to tell us if anything was missing" — scanned by the people with a problem, which is exactly who you want to reach before they leave a one-star review.
- A loyalty stamp. Scan to collect. The value is immediate and it recurs, which is the only way a loyalty scheme survives contact with reality. See WhatsApp loyalty programmes.
- Direct ordering at a better price. Honest and effective: the same meal costs you 30% less through your own channel, so share some of that. This is the offer aggregators dislike most, so read your agreement.
- The full menu. Aggregator menus are truncated and out of date. Yours is not.
Every one of these results in the customer opening the chat. That matters technically as well as legally: an inbound message opens a free 24-hour window, so the entire first conversation costs you nothing, and you have a timestamped record of them contacting you.
Print it so it survives the journey
- On the bag or the box, not on a loose flyer. Flyers go in the bin with the bag.
- Sticker beats printed packaging while you are testing. You can change the offer weekly.
- The QR must be big and matte. Grease and condensation kill glossy small codes.
- Print the wa.me link as text too. Some people will type it.
- Unique code per campaign or branch, so you can tell what worked.
Then be worth the switch
Getting the first message is the easy part. Customers go back to the aggregator because it is easy: saved cards, saved address, live tracking, a refund button. Your WhatsApp channel has to be at least as easy on the parts that matter:
- Reorder in one tap from their last order.
- A saved address, not "please send your location" every time — see address capture.
- A payment link that takes two taps.
- Real delivery updates, not silence.
- A human when something is wrong. This is the one aggregators are worst at, and your biggest advantage.
Keep the aggregators
This is not a plan to leave Talabat. Aggregators are genuinely good at discovery, and delisting to protect margin usually costs more volume than it saves. The goal is narrower and more achievable: stop paying an acquisition rate on customers you have already acquired. If a quarter of your repeat orders move to a channel you own, the economics of the whole business change while the discovery engine keeps running.
Adjoltz builds these flows for UAE brands on Meta's official Cloud API — correct template categories, the buttons, and the inbox behind them. Messages at Meta's rate with zero markup, from $149/month.
Frequently asked questions
Can I message my Talabat or Deliveroo customers directly?
Not using data taken from the platform. Aggregators own the customer relationship and the contact data, and using numbers from order sheets or delivery labels for marketing is a consent problem and usually a breach of your platform agreement. What you can do is give customers a reason to message you first, typically with a QR on the packaging, because then they initiate and you have consent you can evidence.
What commission do Talabat and Deliveroo charge in the UAE?
Published ranges are commonly around 15 to 30 percent for Talabat and 25 to 35 percent for Deliveroo, varying by partnership type, order volume, location and category. Service fees, featured placement, peak-hour surcharges and in-app advertising push the effective rate higher than the headline number.
What should the QR code on the packaging actually offer?
Something the customer wants while the food is in front of them. Asking whether anything was missing works because it reaches unhappy customers before they leave a review, a loyalty stamp works because the value recurs, and a better price for ordering direct works because your own channel genuinely costs you less. Follow us on social gets scanned by nobody.
Should restaurants leave the delivery aggregators?
Usually not. Aggregators are good at discovery and delisting typically costs more volume than the margin it saves. The realistic goal is to stop paying an acquisition rate on customers you already acquired, by moving a share of repeat orders to a channel you own while the aggregator keeps producing new customers.