Pakistan wastes 36 million tonnes of food a year.
This is the number you'll see in every World Food Day op-ed, every UN press release that mentions Pakistan, every CSR report from a Pakistani conglomerate trying to look green. It's the number policy people quote, journalists quote, Twitter quotes when somebody wants to start an argument.
It is also a profoundly misleading number. Not because it's false — it's roughly true. Because it's used as if it's one problem with one solution, when it is at least four very different problems with very different solutions.
The number is a lie of omission. Here's what it actually contains.
The 36 million tonnes is a single line on a single chart. To know what's inside it, you have to break it down by where the loss happens in the supply chain. Roughly:
Farm-level losses. Mangoes that rot in transport from Mirpurkhas to Karachi because the cold chain is broken. Wheat lost to pests in storage in Punjab. Tomatoes ploughed back into fields in Sindh because the farmgate price is below cost. Fish that spoil at landing sites because the ice supply is unreliable.
Wholesale and retail losses. Vegetables binned at sabzi mandis. Bread thrown out at neighborhood bakeries that overstocked. Dairy expiring on supermarket shelves.
Restaurant, café, and quick-service losses. End-of-day production that didn't sell. Buffet leftovers. Wedding banquet leftovers, which is a category of waste so specific to South Asia that international researchers don't even have a clean methodology for it.
Household losses. Rice cooked but not eaten. Vegetables that went off in the fridge. Leftovers thrown out after two days.
If you treat "36 million tonnes" as one problem, you end up with one of two unhelpful framings: either we need cold chain infrastructure (which only addresses farm-level loss) or we need awareness campaigns (which only weakly addresses household loss). Neither does anything for the bucket FoodLoot is going after, which is restaurant and retail end-of-day waste.
That bucket is, by any honest accounting, a small share of the total. But it's the share where a software platform actually moves the needle.
Here's the part that frustrates me.
The 36 million tonnes number gets used to justify policy interventions and donor funding for projects that have nothing to do with the bucket they claim to be solving. A cold storage initiative for mangoes in Sindh is a great project. It is also not a food rescue project. It's an infrastructure project. The food it "saves" is mango pulp at the farmgate, which is a different commodity solving a different problem from a focaccia at HobNob at 9 PM.
But all of these get rolled into the same headline number. Then everyone in the room gets to nod about how serious the food waste problem is in Pakistan, and nobody is forced to actually pick which sub-problem they are working on and prove it.
We've been explicit from day one about which bucket we're in. We do not solve farm-level losses. We do not solve household losses. We do not even solve all retail losses — we don't touch supermarkets or sabzi mandis or wholesale supply.
We solve a specific slice of urban prepared-food loss, at end-of-service, in commercial F&B outlets that have a brand to protect and a customer who wants a deal.
That slice, in Karachi alone, is probably tonnes per year. Which is dwarfed by the 36M figure. But it is the slice we can actually do something about with a phone, a sealed paper bag, and a 30-minute pickup window.
The other reason the 36M number is misleading is geographic.
Karachi pulls more than its share. By any reasonable estimate of urban prepared-food waste, the share concentrated in this city is disproportionate, because the share of Pakistan's commercial F&B sector concentrated in this city is disproportionate.
Karachi has more restaurants per capita than any other Pakistani metro. [VERIFY] The middle-class eating-out culture is denser here. The hours run later. The chain density is higher. The dessert sector alone — the donut shops, the cookie places, the gelato counters that have multiplied since 2020 — is a category that barely exists at this scale outside this city.
So when somebody says "Pakistan wastes 36 million tonnes," and you nod along, you are absorbing a number that includes mango losses in Hyderabad and rice spoilage in Multan and roti thrown by aunties in Lahore. None of those are the problem in front of us. The problem in front of us is the focaccia at HobNob and the cookies at Glaze and the half-pizza at Oofy at 9 PM in Karachi specifically.
Smaller number. Different number. Solvable number.
Why do I care about this distinction enough to write a blog post about it?
Because if FoodLoot becomes a real platform in this market, we will eventually be asked — by media, by donors, by investors — to put a number on what we've "rescued." The temptation will be to wave at the 36M figure and say we're working on that. We're not. We are working on a much smaller, much more specific number.
The honest answer is: in our first month, we will rescue meals in the low thousands. In our first year, if everything goes right, we will rescue in the low hundreds of thousands. That is a useful number. It is not a national policy intervention. It is not going to dent the 36M figure in any visible way at the national level for years.
The 36M figure was never the goal. The goal is the bucket we're in.
That bucket gets smaller, day by day, every time someone in Karachi opens a sealed Loot Bag from one of our partners at 10 PM.
That is the lie of omission I want to fix. Not the number. The conflation of buckets that lets people pretend they are solving one problem when they are solving another, or none at all.
We are solving one. We are explicit about which one. The rest is somebody else's work.