The Facility Was Right. The Product Was Wrong.
Thirty kitchens, six months empty, and what it cost to find out what we were selling.
The first real decision was a building in Gurugram — ten thousand square feet, thirty kitchens, the biggest site we could sign. Correct on every dimension I knew how to measure. We built it, turned it on, and sat with thirty empty kitchens for six months, during the most bullish period the category has ever seen.
We thought we’d built delivery infrastructure — a brand plugs in and starts selling in thirty days, no site search, no build-out. Good product. Brands liked it. Almost none of them signed.
Here’s the part the sector doesn’t say out loud: a cloud kitchen isn’t structurally superior, it’s a different trade. You remove the room — capex drops, rent stops being a fixed cost. Then you hand the customer relationship to an aggregator, who takes twenty-plus per cent of gross. Fixed cost down, commission up. That trade only works above a certain volume, and in 2020, volume wasn’t our constraint — the sales cycle was. Delivery habits were changing exponentially, but no capex committee moves at exponential speed. A market moving is not a business winning.
The fix came from a burger brand that didn’t want a kitchen — they wanted a commissary. Central production for outlets across the city, not one kitchen selling into a delivery radius. Same square footage, same infrastructure, completely different buyer. A cloud kitchen sells to whoever owns the delivery P&L. A commissary sells to whoever plans capacity and already had “build a central kitchen” on a five-year plan they’d rather not fund themselves.
We’d built the second thing and spent six months selling it as the first. Once we fixed that, we went on to build 122 kitchens across three cities in under two years. But those numbers aren’t the ones worth learning from. The number that taught me something was six empty months in Gurugram — a facility that was correct in every way I’d tested, because I’d never tested what the buyer thought they were purchasing. You can test a business exhaustively and still be testing the wrong thing. The only thing that catches it is a customer telling you, plainly, that you’ve misunderstood your own asset.