Own Delivery vs Aggregators: The Commission Math Every Restaurant Should Run

On a 500 rupee order delivered through Swiggy or Zomato, a restaurant can take home as little as 325 rupees. The rest disappears into commission, platform fees, GST on that commission, and the discount the restaurant was told to fund. Run that across a month of aggregator orders and you understand why around 60 percent of Indian restaurants never cross a 10 percent net margin, and why a December 2025 survey found 35 percent of them would leave the aggregators tomorrow if they could.

This is the most viscerally felt problem in the industry. So let us do the math honestly, including the parts that do not favour own delivery, because the answer is not "leave the aggregators." It is "know exactly what they cost, and stop routing orders through them that do not need to be there."

The real commission math

The number restaurants quote each other is the base commission, which in 2026 runs somewhere between 18 and 28 percent depending on city, cuisine, volume, and whether you signed an exclusive deal. But the base is not the real cost. Stack everything on top and the true deduction lands at 25 to 35 percent per order.

Here is what actually comes off a 500 rupee order at a 22 percent commission. The commission itself is 110 rupees. GST on that commission is another 20 rupees. Then there is the per order platform fee, pushed up to 17.58 rupees in March 2026, which is close to pure margin for the platform because it does not scale with your food cost. And if the order came with a "flat 125 off" promo, you likely funded 50 to 100 percent of that discount yourself.

The platform gave you reach. It also quietly took a third of the ticket.

At 100 orders a day on a 500 rupee average, the monthly platform cost runs to roughly 4.4 to 4.5 lakh. The same volume on a direct channel with your own delivery costs a fraction of that. That gap is not a rounding error. It is the difference between a restaurant that survives and one that does not.

What the aggregators are genuinely worth

Before writing them off, be honest about what they do that own delivery cannot.

They are a discovery engine. A new customer who has never heard of you finds you on Swiggy because they were already browsing Swiggy. Your own WhatsApp and app reach only people who already know you. That is the one thing aggregators do that you cannot replicate cheaply, and it is real.

They also handle the logistics you would otherwise have to build: riders, tracking, payments, support, all of it, on demand, with no fixed cost when you are quiet. For a restaurant with unpredictable or low delivery volume, that on demand model is genuinely cheaper than paying riders to stand around.

So the aggregator is not the villain. It is an expensive discovery and logistics service that becomes a bad deal only when you use it for orders that were never discovery in the first place.

The insight: separate discovery from delivery

Here is the mistake almost every restaurant makes. They route a repeat customer, someone who already knows them, already loves them, and would happily order direct, through the aggregator anyway, and pay a third of that ticket for a discovery service that discovered nobody.

That is the money to go after. Not the new customer the aggregator genuinely brought you. The regular you are paying 30 percent to serve for the tenth time.

Own delivery makes sense exactly where discovery is not needed: your repeat customers, your local radius, your loyal base. Aggregators make sense exactly where discovery is the whole point: reaching someone new, in an area you do not cover, at a time your own riders are not working.

The restaurants that win in 2026 do not pick a side. They use the aggregator as a shop window and their own channel as the till.

When own delivery actually works

Own delivery is not free, and it is not right for everyone. It works when a few conditions hold.

You have enough delivery density to keep riders busy. Two or three orders an hour in a tight radius keeps a rider productive. One order every ninety minutes across the whole city does not, and you would be paying idle wages that make the aggregator look cheap.

Your orders are repeat and local. A loyal base ordering within a few kilometres is the ideal own delivery customer, because there is no discovery cost to recover and the distances are short.

You can handle the operational load. Riders, routing, tracking, and the phone call when someone is late all become your job. That is real work, and pretending otherwise is how own delivery projects fail.

We saw this work at Zing Bites in Kozhikode, sitting beside the medical college with four to five riders of their own. Dense area, short radius, repeat student and hospital customers ordering constantly. Every condition for own delivery was met, so they run entirely on their own riders through WhatsApp, with no aggregator and no commission at all. But that is precisely because their situation fit. Drop the same model into a restaurant with sparse, scattered, one time orders and it would bleed money.

The honest recommendation

Do not quit the aggregators. Do audit them.

Look at your aggregator orders and separate the genuinely new customers from the repeat ones. The new ones are the aggregator earning its cut. The repeat ones are the leak. Then build a direct channel, a QR menu, WhatsApp ordering, an app, whatever fits your customers, and give those repeat customers a reason to order there instead: faster, a small loyalty perk, a direct line to you.

Every order you move from aggregator to direct keeps an extra 25 to 35 percent in your kitchen. You do not need to move all of them. Moving a third of them changes your entire margin.

The bottom line

Aggregators are a discovery service priced like a tax. They are worth it for discovery and ruinous for everything else. The skill is not loyalty to one model or the other. It is knowing which of your orders needed the aggregator and which were pure profit you handed away out of habit.

Run the math on your own orders this week. The number will be higher than you think, and it is sitting in your kitchen waiting to be kept.

Want to move your repeat orders off commission? Menuthere gives you WhatsApp ordering, a QR menu, and an app, plus the tools to run your own delivery, so your regulars order direct and your margin stays yours.

Sources: MenuManager and Cybiqon (25 to 35 percent effective deduction, 4 lakh plus monthly cost at 100 orders a day), DineOpen (base commission ranges, GST on commission), Spice Advisors (restaurant funded discounts, sub 10 percent margins), industry survey data (35 percent would leave aggregators, 60 percent under 10 percent net margin).