GoFood Skims 20 Percent Off a Rp15,000 Plate, So Jakarta's Wartegs Went Back to the Counter

A plate of rice, tempe, and sambal at your regular warteg runs about Rp15,000. Do the math GoFood does. Twenty percent commission gone, then the platform asks the owner to fund a promo just to stay visible in the app. What's left after the courier and the packaging is not a margin. It's a rounding error the owner eats.

So the stalls are pulling out. The listing goes dark, and if you want the food you walk over and eat it at the counter like your uncle did.

The cheap plate was never built for a middleman

Warteg economics run on volume and thin edges. The owner buys ingredients in the morning, cooks once, sells all day, and clears enough to restock tomorrow and pay the BPJS premium in installments. A 20 percent cut off the top does not scale down for a Rp15,000 order the way it does for a Rp150,000 one.

Platforms price commission as a flat percentage, which quietly punishes the cheapest food hardest. The mall tenant selling a Rp90,000 bowl can absorb the same rate and still breathe. The warteg cannot. The promo buy-in makes it worse: pay to discount your own plate, then pay the courier, then hand over the commission, then wonder why the till is lighter at closing than it was before you signed up.

The regulars were the business all along

What keeps the gas burning is the crowd that shows up on foot. The ojol driver killing three idle hours between orders. The construction crew that eats at the same table every noon. The office worker who has budgeted Rp15,000 and not one rupiah more for lunch.

Those people never needed an app. They needed a stall that stayed open and a price that didn't move. Delivery promised the warteg a bigger market and delivered a smaller cut of a plate that was already priced to the bone.

Some owners tried to hold both. List on GoFood for the lunch traffic, keep the counter for everyone else, split the difference. The counter kept winning because the counter kept all the money. Every plate sold through the app was a plate sold at a discount to a company in an office tower that has never smelled the sambal.

What the pullout actually costs

You lose the convenience. The owner loses the app's reach, which sounds worse than it is, because that reach was renting the owner's kitchen at a rate the kitchen couldn't afford. Delivery didn't add customers so much as tax the ones already coming.

The regulars gain nothing and lose nothing. They were always going to walk over. Whether the stall survives the year still comes down to the same numbers it always did: cost of rice this morning, plates sold by dark, premium paid in pieces so one bill doesn't swallow a day's takings.

The app was optional. The Rp15,000 ceiling is not. And a platform that takes a fifth of a plate priced that low was never a partner. It was a cost the burner couldn't carry, so the burner stopped carrying it.