Aahfil.

for Kuwait restaurants & cafés

Own the order. Keep the margin.

We build the direct ordering app that takes your repeat customers off the aggregator: full menu with modifiers, KNET and cash on delivery, area and time-slot logic, kitchen notifications, and the WhatsApp reorder message that gets the second order. Live in 30 days.

Casual dining Cafés & specialty coffee Cloud kitchens Bakeries & desserts Healthy meal plans Multi-branch chains Casual dining Cafés & specialty coffee Cloud kitchens Bakeries & desserts Healthy meal plans Multi-branch chains

the actual problem

Why the current way is costing you

Aggregators are excellent at discovery and expensive at everything after it. They take a commission on the customer who already loves you, they own the relationship, and they will not tell you who ordered twice. Meanwhile the direct orders arrive as Instagram screenshots at 9pm and someone reads them out to the kitchen.

What ships, and when

  • Booking & ordering web app

    30–45 days

    Full menu with modifiers and combos, KNET and cash on delivery, governorate and time-slot delivery rules, kitchen order notifications, and a branch view.

  • WhatsApp-driven system

    14–30 days

    Order confirmation, 'ready / on the way' updates, and the timed 'same as last time?' reorder message with a one-tap KNET link.

  • Storefront build

    21–45 days

    For meal plans, retail bags and merchandise that belong in a proper cart with subscriptions rather than an order form.

Day ranges are for a first-version scope — we fix them in the proposal.

the schedule, published

From day one to launch

  1. 1

    Days 1–5 — menu and rules

    Your menu as structured data with modifiers and prices — this is the single biggest delay when it is not ready — plus delivery areas, minimums and slots. Gateway paperwork starts now.

  2. 2

    Days 6–18 — build ordering

    Menu, cart, checkout with KNET and cash on delivery, delivery rules, and the kitchen notification screen. You place a test order in week two.

  3. 3

    Days 19–28 — the reorder loop

    WhatsApp confirmations and the reorder flow timed per customer, plus server-side tracking so ads optimise on paid orders rather than menu views.

  4. 4

    Days 29–40 — launch and migrate

    A first-direct-order offer to everyone who has ordered from you before, an insert in aggregator bags, a QR on the table, and retargeting anyone who viewed the menu.

  5. 5

    Day 41 onward — grow direct share

    Monthly: direct order share, reorder rate, average basket, commission saved. The aggregator stays for discovery; the repeat business moves home.

Build or buy?

You are not choosing between an app and nothing — you are choosing between a custom build and a tool you already know. Here is the honest comparison, and sometimes the answer is to stay where you are.

Talabat / Deliveroo
Keep it ifYou are new, or a branch is new, and you need discovery. They are the cheapest customer-acquisition channel in Kuwait food.
Build ifA meaningful share of the orders are repeat customers. Multiply that share by the commission — that is what your own app competes against every month.
Instagram menus
Keep it ifYou are testing a concept or running a pop-up and volume is low.
Build ifSomeone is reading screenshots to the kitchen, or you cannot tell which customer ordered twice.
Off-the-shelf ordering plugins
Keep it ifYour menu is simple, has no modifiers, and delivery is a flat fee everywhere.
Build ifYou need combos, per-area fees, time slots, branch routing, or the kitchen to be notified without someone watching a screen.

the closed loop

Software and marketing, one system.

The ad shows a dish, lands on that dish, and the paid order fires back as a server-side conversion so the campaign optimises on revenue rather than menu views. Every direct customer enters the reorder flow timed to their own pattern, and the dashboard shows direct-order share and commission saved next to cost per order — which is how you know whether moving off the aggregator is actually working.

What we wire in

  • KNET, cards and cash on delivery
  • Delivery fleet APIs or your own drivers
  • Governorate, block and time-slot rules
  • WhatsApp Business API (updates, reorders)
  • Kitchen / branch order notifications
  • POS integration ‹on request›
30

days to launch

0%

commission on direct orders

2

weeks to a test order

1

message that wins the second order

Figures shown are illustrative examples.

Get the 30/60-day plan

Tell us about the business and what you want a customer to be able to do. You get back a first-version scope, a day count and a cost range — and if we think you should not build yet, we will say so.

Prefer to talk now?

Based in Kuwait. Working across the GCC.

Or ask right now

Ask about cost, timeline, or whether an app is even the right answer for your business.

You're on the restaurant ordering app page. How are orders reaching you today — aggregators, DMs, phone? And roughly what share is repeat customers?

AI assistant · answers checked by a human before anything is promised

Or read first

Everything we would do is explained in these. Take it and run it without us.

Frequently asked questions

Should I leave Talabat if I have my own app?
No — keep it for discovery and move the repeat orders. Pull one month of aggregator orders, count how many are repeat customers, multiply by the commission. That monthly number is what your own app competes against, and it usually clears the build cost in months rather than years.
How do I get customers to order directly instead?
Give the first direct order something the aggregator never did, put an insert in every aggregator bag, a QR on the table, and retarget everyone who viewed your menu. Then send the reorder message on the day and hour that customer usually orders. Launch is a campaign, not an announcement — we run it.
Can the app handle modifiers, combos and multiple branches?
Yes — modifiers, combos, per-area delivery fees, time slots and branch routing are what separate a real ordering app from a form. Get your menu into a spreadsheet with modifiers and prices before the build starts; it is the single biggest cause of delay.
Do I need my own drivers?
Not necessarily. We can wire a delivery fleet API so orders are dispatched without you hiring, or route to your own drivers, or both by area. Price a fleet API per order against your aggregator commission on your average basket — the crossover is usually lower than owners expect.
Can we take table reservations too?
Yes, as a module on the same app. The part worth building is the deposit above roughly eight covers and the morning-of WhatsApp confirmation — large parties that do not arrive are where reservations actually cost you.