The Restaurants With the Most Customers All Have One Thing in Common
Most operators know this number. It gets quoted in every WhatsApp group in the industry. What gets discussed far less is the second cost, the one that never appears on the invoice.
On a ₹500 delivery order in India, a restaurant can take home as little as ₹325. Base commissions on Zomato and Swiggy run somewhere between 15% and 30%. Add GST on the commission, add the platform fee that both apps pushed to roughly ₹17.58 in March 2026, add the discount you are expected to fund yourself, and the effective deduction lands closer to 25% to 35% of the order value.
You do not get the customer.
What we noticed by accident
We did not set out to study delivery models. We build digital menus. But when we looked at which restaurants get the most out of Menuthere, a pattern showed up that we were not looking for.
The ones running their own delivery had, by a clear margin, the highest number of customers.
Not the largest kitchens. Not the ones listed on the most platforms. Not the ones with the biggest ad spends. The ones with their own boys on their own bikes, delivering to addresses they had in their own records.
Our first instinct was that this was about margin. It isn't, or at least not only. Own delivery is not automatically cheaper. Operators consistently underestimate their true per order fleet cost, often by 40% or more, because they count rider salary and fuel and forget maintenance, insurance, idle hours and the management time the fleet quietly eats. Industry voices have argued for years that if delivery is under roughly 20% of your sales, running a fleet does not pay for itself. That argument is sound.
So the advantage was not the bike. It was what the bike leaves behind.
Orders versus customers
An aggregator gives you orders. Your own delivery gives you customers. These sound like the same thing on a busy Friday night. They are not the same thing at all when you look at a twelve month view.
An order is a transaction. It arrives, you cook it, it leaves, and it is gone. A customer is a record: a name, a number, a building, a preference, a frequency. The first can be bought over and over. The second compounds.
The uncomfortable arithmetic is that on an aggregator you pay full acquisition cost on the fiftieth order from the same person that you paid on the first. You are not buying reach at that point. You are renting back a customer you already earned, at 25% to 35%, forever.
A restaurant with its own delivery breaks that loop by default. Not through strategy, usually. Just because the order came to them, so the details stayed with them.
Why the relationship is worth more in 2026 than it was in 2022
Two shifts made customer ownership a survival issue rather than a nice to have.
First, diner patience with fees has run out. Recent consumer research found that 61% of diners have abandoned a delivery order because the service fees felt too high. That customer did not stop wanting your food. They stopped wanting the checkout page. If you have a direct channel, that abandoned order is recoverable. If you don't, it simply vanishes.
Second, loyalty itself has loosened. In the same body of research, 45% of diners said their favourite restaurant had changed within the past year. The old assumption that a good meal buys you a regular no longer holds on its own. Something has to bring them back, and an algorithm you do not control is a poor candidate for that job.
Meanwhile the platform economics keep moving in one direction. Around 60% of Indian restaurants never cross a 10% net margin. A December 2025 industry survey found 35% of restaurants would leave Zomato and Swiggy tomorrow if they had somewhere to go. That is not a loyalty problem. That is a dependency problem.
The part most own fleet restaurants get wrong
Here is where the pattern gets genuinely useful, and where it stops being a story about delivery at all.
Owning the customer is not the same as using the relationship. Most restaurants with their own delivery are sitting on the raw material and doing nothing with it. The numbers live in a WhatsApp thread, or a diary near the billing counter, or in one senior boy's head. Nothing gets segmented. Nothing gets brought back. The advantage exists on paper and expires unused.
This is the gap. Own delivery gets you the relationship. What you do next decides whether it becomes revenue.
That "next" is almost always the ordering surface. If the customer's second, fifth and twentieth order all run through a menu you control, three things become possible that were impossible before:
You can change what they see. Prices, combos, availability, sequence. A slow item gets moved up. A dead item comes off. The change is live in seconds and costs nothing, which means you will actually make it, rather than waiting until the next reprint.
You can merchandise properly. Photographs, descriptions, tags for the dishes you make the most money on. On an aggregator you are one tile among forty and every tile looks the same. On your own menu you are the only restaurant on the page.
You can run different menus for different moments. Breakfast, lunch, late night, weekend. The same customer sees the right menu at the right hour instead of one flat list built for the average, which fits nobody.
None of this is exotic. It is basic retail merchandising that restaurants have been unable to do because someone else owned the shelf.
What to actually do this quarter
If you already run your own delivery, you are further ahead than you think. The work is to convert the relationship you already have.
Get the numbers into one place. Whatever exists in diaries, phones and WhatsApp threads should be in a single list this month. Nothing else on this page works without it.
Put your own menu on every physical surface. The bill, the packaging, the sticker on the bag, the QR at the table. Every touchpoint is a chance to make the next order direct.
Give the direct customer a reason. Not a deep discount. Something the apps cannot match: a free add on, priority on busy nights, a dish only the direct list gets to order.
Price the aggregator menu separately. Most successful operators list 10% to 25% higher on platforms to absorb commission. Running one price everywhere means the direct customer is subsidising the platform.
Keep the aggregators for discovery. This is not a delisting argument. Treat platforms as paid acquisition, then move the repeat order home. The expensive mistake is paying commission on the fiftieth order, not the first.
Fix the menu itself. Once ordering is direct, the menu becomes the highest leverage surface in the business, and the only one you can change daily.
The pattern, stated plainly
The restaurants winning on customer count are not winning because they own bikes. They are winning because every order they take adds to something they keep.
The fleet is one way to get there. The direct ordering channel is another. The menu is where both of them turn into money.
Menuthere is built for exactly this stretch of the journey: restaurants who already have the relationship and want the ordering surface to work as hard as the kitchen does. If you run your own delivery and your menu is still a static PDF or a printed card, that is the cheapest gap in your business to close.
Want to see what your menu would look like as a live, editable ordering surface? Start with Menuthere.