If your website's only ordering feature is a button that sends customers to Talabat, your site is a brochure for someone else's checkout. The fix is not to leave Talabat. It is to split the channels deliberately: keep Talabat for discovery, where its audience and delivery fleet earn their keep, and build direct ordering on your own site for the repeat orders you currently rent back, one commission at a time. This guide covers the decision and the design — what the owned journey must contain, which orders belong on which channel, and how to run the per-order numbers with your rate.
The short answer: own the repeat order, rent the discovery
Talabat is a discovery and delivery marketplace. A customer who has never heard of your restaurant finds you there, and that first order is worth paying for. The problem starts when a loyal customer who already knows your name and menu orders through the app for the twentieth time — and you pay marketplace economics on an order you did nothing to win. Your own channel exists for that customer. Own the repeat order, rent the discovery, and design your website for the handover between the two.
The problem: a website that can only send customers away
Many UAE restaurant websites follow the same pattern: attractive photos, a PDF menu, opening hours — and an order button that hands the customer, the payment and the relationship to an aggregator. Every weakness compounds. You cannot offer a loyal customer a reason to return, because you do not know who they are. You cannot fix a bad ordering experience, because it is not yours. You cannot see your own demand clearly, because the data sits in someone else's dashboard. Our guide to restaurant website design covers the wider design picture; this article deals with the ordering decision underneath it.
What an order actually costs on each channel (worked per-order table — bring your own commission rate)
Talabat does not publish a single public commission rate — your rate is the one in your own partner agreement, so use that figure and no one else's. Take a basket of AED 100 to keep the maths simple: every percentage point of commission on that basket is AED 1. Fill in your own numbers:
| Per AED 100 order | On Talabat | On your own site |
|---|---|---|
| Commission | Your contracted rate × AED 100 | None — no marketplace takes a cut |
| Payment processing | Per your partner terms | Your gateway's fee, per your agreement |
| Delivery | Per your Talabat fulfilment arrangement | Your rider or courier cost, or zero for pickup |
| Promotions | Campaign contributions, if you join them | Your own offers, set and stopped by you |
| Customer data | Held by the platform | Yours, with the customer's consent |
The comparison is not "commission versus free". Direct ordering has real costs — gateway fees, delivery, the ordering system, and the marketing that replaces the app's traffic. The honest question is narrower: on orders from customers who would have found you anyway, does your contracted commission cost more than running those orders yourself? Run the table at your order volume before deciding.
What Talabat genuinely gives you (discovery, delivery fleet, demand you cannot buy cheaper)
A fair decision starts with what the aggregator does well. Talabat puts your restaurant in front of customers actively looking for food in your area, including people who would never search for you by name. Its delivery fleet solves a logistics problem most single restaurants cannot solve alone at peak hours. Its app also sets the convenience benchmark — saved addresses, stored payment details, live tracking. Buying equivalent discovery yourself, through advertising, would cost money too, with no guarantee of the same conversion. None of that argues for sending every order there; it argues for keeping the channel and using it for what it is good at.
What only your own site gives you (margin, customer data, menu and brand control)
Your own channel returns three things no marketplace will. First, margin: a repeat order carries no marketplace commission, so the same basket contributes more. Second, customer data: with consent, you learn who orders, how often and what they reorder, which makes loyalty offers and win-back messages possible. Third, control: your menu, your prices, your photography and your story, presented without a competitor's sponsored listing one swipe away. You also control the experience when something goes wrong — a direct apology and a replacement dish build more loyalty than a platform support queue. Features like online payment, ordering and customer accounts are standard ecommerce development work; the difference is applying them to a menu rather than a product catalogue.
Designing the owned ordering journey: menu, cart, checkout, kitchen
Direct ordering fails when it is treated as a button instead of a journey. Design all four stages:
Menu. Build a real ordering menu, not a PDF: items with modifiers (size, spice, extras), clear availability, honest preparation times and photographs of the dishes. Mark items unavailable the moment the kitchen runs out — nothing kills a direct channel faster than accepting payment for a dish you cannot serve. If you serve Arabic-speaking customers, the menu needs an Arabic version with right-to-left layout, not a machine-translated afterthought.
Cart. Keep the basket visible and editable, show delivery fees and minimums before checkout, and let customers save a favourite order for one-tap reordering — repeat orders are the whole point of the channel.
Checkout. Ask for the minimum: name, phone, address or pickup choice, payment. Offer the payment methods your customers expect, including card and cash on delivery where you support it. Every extra field costs orders. Send an immediate confirmation with an order number and a realistic time.
Kitchen. The order must land where staff actually work — integrated with your POS where you have one, or on a dedicated kitchen display or printer where you do not. Define who accepts the order, what happens if nobody accepts within a few minutes, and how refunds and cancellations are handled. Test the full path — order, payment, kitchen receipt, refund — before you promote it.
The hybrid split: which orders to push direct, and how (packaging, loyalty, Google, WhatsApp)
With both channels live, steer each order to the channel that suits it. First orders from new customers: Talabat, where discovery happens. Repeat orders, large family and group orders, and pickup orders from nearby regulars: direct, where the margin and the relationship are yours. Then make the direct channel easy to find and worth using. Put the ordering link on your Google Business Profile and social profiles, and point dine-in customers to it with table cards and packaging for orders you fulfil yourself — where your aggregator terms allow promotional material, a loyalty invitation in the bag moves repeat orders over time. Give direct customers something the app cannot: a loyalty programme, a direct-only dish, or priority pickup. Offer WhatsApp as a fallback for customers who prefer to message, with a clear handover to staff — but treat it as a fallback, because unstructured chat orders are where mistakes breed.
Decision criteria: order volume, repeat rate, delivery radius, POS in place
Four questions settle most cases. Order volume: at very low volumes, the fixed cost of running your own channel may outweigh the commission saved — the table above tells you where your break-even sits. Repeat rate: restaurants live on regulars; the higher your repeat share, the more you are paying commission on demand you already own. Delivery radius: if most orders are pickup or come from a tight radius your own rider can cover, direct delivery is practical; a wide, sparse radius is where the aggregator fleet earns its cut. POS in place: if your POS can receive web orders cleanly, the build is mostly design and configuration; if orders would be re-typed by hand, fix the routing first or the kitchen will reject the channel.
Common mistakes
Quitting aggregators outright is the biggest: you lose discovery before your direct channel has an audience. Second is the PDF menu with an order button bolted on — it adds friction without adding a journey. Third is building direct ordering and telling nobody: an unpromoted ordering page changes nothing, because customers default to the app they already have. Fourth is copying your Talabat prices without giving customers a reason to order direct. Finally, neglecting the kitchen end — if staff experience direct orders as extra chaos, they will quietly discourage them.
FAQ
Should I leave Talabat completely if I build my own ordering?
No. Keep Talabat for discovery and delivery reach, and move repeat, pickup and regular customers to your own channel. Leaving outright usually costs more in lost new customers than it saves in commission.
What commission does Talabat charge?
There is no single published rate to quote: commission and fees are set in each restaurant's partner agreement and vary with the services you use. Check your own contract, and be wary of any article that states one figure as universal.
What do I need in place before direct ordering can work?
A structured menu with modifiers and live availability, online payment and cash-on-delivery options, a reliable route for orders into your POS or kitchen, defined delivery zones and times, and a plan to tell customers the channel exists.
Can customers just order through WhatsApp instead?
WhatsApp works as a fallback and suits some regulars, but chat orders are unstructured: items, addresses and payments get missed, and there is no cart, confirmation or order history. Use it alongside a proper ordering journey, not instead of one.
How do I move repeat customers from Talabat to my own site?
Make direct ordering visibly better: loyalty rewards, direct-only items, easy reordering and your link on Google, social profiles and dine-in materials. Move customers gradually with reasons to switch, rather than removing the channel they know.
Scope a direct-ordering website for your restaurant (CTA)
If your repeat customers are your profit, your website should be able to take their orders. WebStackRank designs and builds restaurant websites with owned ordering journeys — menu, cart, checkout and kitchen routing — scoped in a fixed-price, itemised proposal. See what we build for restaurants and food businesses, compare our web design team in Dubai and fixed-price web design approach, then tell us your order volume and current channels and we will scope the direct channel your repeat orders deserve.
