What the delivery apps take from a NPR 600 order
It is eight in the evening in Baneshwor. The kitchen is loud, the fryer is going, and the aggregator tablet on the counter dings again. Another order. It feels like a sale, and it is one. But sit down after closing with a calculator and the feeling changes a little.
This post is that calculator session. Not to tell you to quit the apps, because they do real work for you. The argument is narrower: your regulars, the people who order from you every Friday, should not be reaching you through a channel that charges you as if they were strangers.
The maths on a NPR 600 order
Take an illustrative order, say NPR 600: two plates of momo, one chowmein, one Coke. Commission rates vary by platform and by contract, so use a round 20 percent for the sums. Twenty percent of 600 is NPR 120, deducted before you ever see the money.
Now the rest of the plate. Ingredients on a typical food order run somewhere around 35 to 40 percent, call it NPR 230. Packaging, meaning the boxes, the bag, the sauce cups, comes to another NPR 40 or so. That already leaves about NPR 210, and rent, gas, electricity and staff have not taken their share yet. On this order the commission costs more than all the packaging and is close to half the ingredient bill. For many kitchens it is the single largest cost after food itself.
One order does not hurt. Volume does. Say the app brings ten such orders a day. That is NPR 1,200 a day, roughly NPR 36,000 a month, paid quietly through a tablet. In plenty of Kathmandu kitchens that is a cook's salary.
To be fair, the apps earn part of that
Honesty cuts both ways. The commission buys real things.
- Discovery. A student new to Thamel opens the app and your kitchen appears next to fifty others. You could not reach that person any other way for that money.
- Riders. The platform handles delivery, so you do not manage scooters, fuel, or a delivery boy who vanishes during Dashain.
- Trust. People order from an unknown kitchen because the app sits between them and you.
For a brand-new customer, 20 percent can be a fair price. The problem is not the first order. The problem is the fiftieth.
The order you are overpaying for is the repeat one
Think of the family in Patan that orders your sekuwa every Saturday. They found you months ago. They are not being discovered any more. Yet every Saturday the same commission comes off, at the same rate as on day one.
Worse, they are not really your customer while they stay inside the app. You do not have their number. You cannot message them when you add a new item or close for Tihar. And when a rival kitchen two lanes over runs a discount on the platform, your regular sees that banner before they see your menu. You paid to build a habit, and the habit lives in someone else's app.
Most owners can name ten such households from memory. Those ten are exactly the orders that should be coming to you directly.
What an ordering page of your own looks like
An ordering page is simply your menu on your own website: photos, prices in NPR, a cart, and an order button. The customer taps, the order lands in your dashboard, and they get an order email confirming it. No commission on the order, whether it is NPR 600 or NPR 6,000.
This is the part Pasal was built for. You pick a restaurant template, type in your menu, and the page is live in about five minutes, on your own domain if you want, in Nepali and English. Customers can make an account so their details are saved for next time, and you can give loyalty points on direct orders, so the Saturday family has a reason to keep taking the direct road. If you also do dine-in, the same menu works for QR ordering at the table.
Now the honest trade-offs, because there are some.
- There are no riders. You arrange delivery yourself: your own scooter, a cousin, a local courier. Within your own neighbourhood this is usually manageable. Across the whole valley it is not.
- Online payment through eSewa and Khalti is coming to Pasal but is not live yet, so for now direct orders are cash on delivery or whatever arrangement you already use with regulars.
- Nobody stumbles onto your page. The app has crowds; your page has only the people you send there. Moving them is your job, which brings us to the practical part.
How to move regulars across
You do not need a marketing budget. You need the delivery bag.
- Put a small card in every aggregator order: your page address, a QR code, and one line saying order direct next time and pay less. Printing a thousand cards costs very little against one month of commission.
- Offer a promo code for the first direct order, say 10 percent off. On a NPR 600 order you give up NPR 60 instead of paying NPR 120 in commission, and the customer got a better deal than the app gave them. Both sides win.
- Give loyalty points only on direct orders. The app will never let you reward your own regulars. Your page can.
- Put the link everywhere that is already yours: the Instagram bio, the signboard, the till counter, the WhatsApp status before Dashain when families order in bulk.
- Say it at the counter. Shopkeepers from New Road to Patan underestimate this every time. Next time order from our own site, it is cheaper for you and better for us. It works because it is true.
Do not expect everyone to move. Some customers live inside the app and will not leave it. That is fine. You are not chasing all of them. You are chasing your top thirty households.
Keep the apps, change their job
The end state is not app-free. It is a split that matches what each channel is good at.
The aggregator's job is discovery: new customers, new areas, slow weekday nights, the student who just moved to Thamel. Treat that commission as a marketing cost and judge it as one.
Your own page's job is the relationship: repeat orders, bulk festival orders, the customers who already know your taste. No commission, your customer list, your rules.
Run both for three months and count. If even eight orders a day shift to direct at our illustrative NPR 120 each, that is around NPR 28,000 a month staying in the till instead of leaving it. Your food did not change. Your customers did not change. Only the road they took to reach you changed.
If you want to try the direct road, you can set up your restaurant's own ordering page at https://pasal.studio/start and see how it feels before you print a single card.