Somebody else's ingredient

The press runs until the apples are gone, and what leaves it is a smell before it is a pile. Warm pomace comes off the belt, sweet and sharp at the same time, and by evening it has steamed itself into a heap behind the shed that the wasps find before anyone else does. Leave it a day and the sweetness turns, the heap heats in the middle like a loaf, and somebody has to decide something: tonight it goes over the fence to the pigs, tomorrow it goes on the compost, and next week it is a problem with a truck attached to it.

Every farm carries a second list of customers, and most of them are not people. The pigs will take the pomace, the hens will take the trim, the cattle will take the spent grain, the soil will take what is left if it is spread thin enough, and the compost pile will take what the soil will not. None of that is charity and none of it is new. It is the ordinary arithmetic of a place where everything that eats is standing a few metres away, and nothing in the heap is waste to anyone with a mouth.

The catch is timing, and the catch has always been timing. A glut arrives inside a two-week window and the eaters are hungry all year. What has changed is what happens when the eaters are not on the property: the heap leaves, in somebody else's vehicle, on somebody else's schedule, and increasingly through somebody else's software. That is the subject of this chapter. The farm's leftovers have become an economy, and the economy has an owner.

The oldest outlet is an animal

Start with the stream that is easiest to measure, because it is the one a single product creates. Cheese leaves whey behind in a ratio that surprises people who have never stood in a cheese room: "for every kilogram of cheese produced, nine litres of whey are produced and must be reprocessed or disposed of." That is most of the milk. In Canada the nutrient-rich fraction of it has been "mainly reused as fertilizer in the fields," and researchers have spent years looking for better uses, from protein recovery to fermentation. Pigs drink whey, and always have. The oldest answer to a glut is an animal, and it is also the cheapest answer, because the animal is already there and the truck is not.

Brewing leaves spent grain, and the trade press says the obvious thing plainly: "Spent grains, a by-product from brewing, are a highly valued source of feed for livestock." The reason is in the cattle business's own arithmetic. "Feed is the single biggest cost on most cattle operations," which is why a business exists in British Columbia whose only job is moving brewery grain to small farms at low cost, started by someone who wanted small farmers to have a way to stay in it. Nothing about that arrangement is clever. A brewery has a stream it would otherwise pay to discard, a cattle operation has a feed bill it would like to lower, and the distance between them is the business.

Potatoes take the same idea one step further, into equipment. A Quebec startup looked at the volume of cull potatoes that never leave the field or the shed and built "a mobile starch extraction unit that travels from farm to farm," turning unsold potatoes into starch for bioplastics and paper instead of leaving them to rot. The plant comes to the pile. That is the opposite of the way most processing works, and it is the correct shape for a stream that is too wet, too heavy, and too seasonal to be worth hauling.

The same arithmetic runs through a community's own food. Camden Lawrence, who works with First Nations Agriculture and Finance Ontario, described what a commercial operation leaves behind in terms of who is standing nearby: "the chickens that aren't big enough or the... vegetables that aren't, don't look pretty right? That... could feed that... community or surrounding." The ellipses mark his pauses rather than any tidying. The seconds are real food, the neighbours are close, and the only thing missing is the arrangement.

All four cases share a structure. The leftover is not a product waiting for a use; it is a logistics problem wearing a food's clothes. Which is also why software eventually shows up, because logistics is what software is good at.

The heap leaves the gate

Grocery stores generate the largest single stream of edible surplus in the country, and the surplus is boring and constant: bread, produce, dairy, meat, all of it hitting a date on the same weekly schedule. What happens to that stream is now partly decided by who can be matched to it in time, and Canada has built a small industry of matching.

Loblaw describes the farm end of that pipeline on its own site: "In 2025, more than 20,000 metric tonnes of food was provided to feed animals from our corporate grocery retail stores." That is one retailer's stream, disclosed by the retailer, and it is measured in the tens of thousands of tonnes a year because grocery volume is that large.

Second Harvest, the country's largest food rescue organization, runs the human end of the same problem on a national scale, and its own accounting for 2025 is 95.3 million pounds of food rescued and redistributed, with the greenhouse-gas saving attached. FoodMesh describes itself as trying to be a connected food system, a giant switchboard that asks what you have and who needs it before the food stops being food. Jessica Regan, the company's chief executive and co-founder, says the platform's job is to create the matches and keep them running, and that the last eight years went into grocery retail because that is where surplus arrives every single day.

Regan is unusually clear about why the top of the hierarchy keeps losing to the bottom, and her explanation has nothing to do with technology. She walked through the order the sector itself uses, in which you try to sell what you can, discount what you cannot, "donate it to humans, send it to animals, and then compost," and then she named the thing that fills the landfill anyway: "It's not someone's job to find alternative outlets for the food that they can't sell. Their job is to sell it for a premium." The waste is nobody's assignment. That sentence explains more about the heap behind the shed than any dashboard does, because it explains why the cheap, obvious option, the pigs or the compost or the neighbour, wins whenever nothing pushes back, and why the expensive option, the landfill, wins whenever the alternative costs somebody an hour.

Regan also put a number on the reason farmers want the stream at all. "Feed is one of the largest variable costs for a farmer. It's gone up 50% since pre COVID." A farm that can take somebody else's surplus is a farm buying down its biggest bill, which is why the grocery-to-farm pipeline exists at all, and why it grew.

Who pays for the surplus

Follow the money far enough and a question about groceries turns into a question about accounting. Every large Canadian grocer now publishes a food-waste commitment, and the language repays reading closely. Loblaw's targets include sending "zero food to landfill by 2030," and its own page counts the progress in two numbers: every food retail store partnered with a food waste diversion partner, and more than 83,000 metric tonnes of food donated or diverted in 2025. Empire, which owns Sobeys, announced in 2026 that it had "surpassed its commitment to reduce food loss and waste by 50 per cent," measured against a 2016 baseline. Those are real tonnages and real reductions. They are also achievements counted at the loading dock, which is a different measurement from tonnage that was never unsold, and the difference between the two is where the interesting questions live.

The mechanisms that produce those numbers are documented, and some of the documents are written to encourage them. Public health guidance for food businesses describes donation as one way to "offset costs associated with surplus and un-sellable inventory," alongside reducing the handling costs of disposing of edible but unsellable food and claiming the value as a deduction or a charitable receipt. Ontario's own food and organic waste policy statement reports that the industrial, commercial and institutional sector "sends an estimated 75% of its organic waste to disposal, mainly to landfills," and sets a 70 per cent diversion target for its largest facilities. Metro Vancouver bans food scraps from disposal outright and supports what it calls "a price differential for green vs solid waste tipping fees," so that composting costs a business less than landfilling does.

Put those three things together and the incentive is plain. A store holding a pallet of yogurt it cannot sell has to do something with it, the rules make throwing it out expensive or illegal, and the cheapest remaining move is to hand it to somebody who will take it. Who sorts it, who hauls it, who keeps it cold, and who composts the part that nobody could use? Those costs land on the receiver, which is often a charity, a food bank, a farmer, or a municipal organics program funded by ratepayers. The store gets a smaller disposal bill, a compliance number, and a paragraph in an annual report. Where a city runs or regulates the organics system, part of that public infrastructure absorbs commercial waste at tipping fees that were set to make diversion cheaper than landfill, which moves the cost from a corporate ledger to a public one.

Which is why the question to ask of any grocery diversion programme is not how many tonnes it moved, but what it changed upstream. A pledge measured in diverted tonnage can be met without ordering one fewer crate of strawberries, and the arrangements that manufacture the surplus, the private cosmetic specifications and buyer grading rules that decide what counts as sellable, are the part nobody outside the industry gets to read. Whether any of this amounts to laundering, in the sense of routing a disposal cost through a charity, a farm, or a municipal system so it can be counted as a contribution, is not a question the tonnage can settle. What the tonnage does show is where the bill lands: not on the store. Small farms sit at the far end of that arrangement, taking feed they did not choose, on a schedule they do not set, at a quality they cannot check before the truck arrives, which is a strange kind of help to be offered by the company that created the surplus.

The camera over the bin

The food-waste technology that gets covered is aimed somewhere else: at commercial kitchens, where a scale and a camera now watch what gets thrown away. LeanPath has sold that system into thousands of commercial kitchens and made the tracking touchless, so nobody has to log anything and the software identifies the food as it is dumped. The company reports an average 50 per cent reduction in food waste across its deployments, along with a smaller cut in purchasing costs, and it is not the only vendor claiming a figure in that range. Those numbers come from the vendors. Independent verification at that level has not surfaced, and a camera over a bin is excellent at telling a kitchen manager what the kitchen already wastes, and useless for producing less of it upstream.

Canada's own record in this area is the corrective the marketing never offers. The substantive food-waste work here is operational, not artificial intelligence: the platforms move food with pickups, contracts, and phone calls, and the largest Canadian diversions on record were achieved with logistics, not models. When the research behind this book went looking for Canadian food-waste deployment built on artificial intelligence, it found that layer living almost entirely in the United States, and Canadian retail artificial intelligence pointed at supply chain forecasting rather than at waste. That is worth saying plainly, because the subject here is not the sensor. The sensor is the smallest part of the story and the easiest part to sell.

Two things not to merge

There is a distinction in this sector that the people who do the work insist on, and it deserves the space. Lori Nikkel, the chief executive of Second Harvest, put the two halves apart in one sentence: "supporting people with food is an outcome of managing food waste," and she was direct that feeding people does not, in any way, stop them from being food insecure systematically or sustainably. Feeding people and managing waste are different problems that happen to share a truck. Her organisation made the same shift deliberately when she took over, from a food charity to what she calls a food waste organisation, which is a strange thing for a rescue network to call itself and a clarifying one.

Elaine Power, a food systems researcher at Queen's University, traced the same confusion back to the beginning, noting that "the origins of food banks were about redistributing food waste." The institution that now feeds millions of Canadians started as a way to move surplus, and the surplus moved because it was expensive to throw away, not because anyone had decided that hunger was the emergency. That history is why the sector's own people separate the two halves so carefully, and why a chapter about leftovers has to keep them separate too. A heap turned into an ingredient is a logistics success and it is not a food-security policy.

Who owns the match

Lots of farms can each take a pallet of unsold bread, but only if somebody tells them the bread is available, and that somebody has become the most important figure in the leftovers economy. The match is the product. It is also, in Canada, almost always private.

FoodMesh is a certified B Corporation running a proprietary marketplace on a commercial platform, and its growth has come substantially from public contracts, with Metro Vancouver paying it to run the region's food recovery network. Public waste-reduction policy is being delivered on private, closed software, which is the same enclosure pattern a reader will meet again in the data chapters, arriving here through the compost. LOOP Mission is the consumer-facing version of the same story: the country's most visible upcycling company takes cosmetically rejected produce and turns it into cold-pressed juice, and its scale depends on high-pressure processing machines it owns, described in its own interviews as the biggest purchase the company ever made. Access to that machine is sold by the hour to anyone who can pay the toll. The category's own standards are behind paywalls as well, so even the definition of upcycled is a product.

Against that, the open alternative in this corner of the economy is thin in an instructive way. The clearest example in the research of a surplus platform that calls itself fully open source publishes repositories with no licence attached to either of them, which means it has published nothing anyone is legally allowed to reuse. Community canneries, shared kitchens, and co-operative processors are the genuinely open layer here, and they are open in the oldest way rather than the software way: shared equipment, shared inspection costs, shared season. A pig is still a pig, and a compost pile still asks for nothing.

If the matching layer is where the value is created, and the matching layer is proprietary software operated by a handful of firms, then the leftovers economy gets more efficient without getting more open, and the farms and charities inside it end up as customers of the match. The physical layer of reuse is as open as it has ever been. The layer that decides what moves is not.

Open questions

What does the extra trip cost, and who pays for it? A heap that stays on the farm costs nothing to move. A heap that leaves costs fuel, labour, refrigeration, and a margin for whoever arranged the match, and in a tight year that arithmetic decides more streams than any standard does.

Does a market for waste change what gets grown? Appearance standards manufacture the waste stream in the first place, and a market that reliably pays for seconds could make seconds worth producing on purpose, which is either the best news in this chapter or the strangest.

Who checks the numbers? Almost every figure here, from the tonnes diverted to the pounds rescued to the vendor's 50 per cent, is reported by the organisation whose business depends on the figure. That is a gap rather than an accusation, and it is the same gap the data chapters describe in a different currency.

Does the rescue layer relieve pressure to fix the cause? It moves a great deal of food and it also makes the same food system look like it is working, and the people quoted above disagree about how much weight that should carry.

Moves you can make

Turn one waste into one ingredient this week. Bread ends become breadcrumbs or croutons. Whey from making yogurt becomes ricotta or the liquid in a soup. Coffee grounds go to a neighbour's garden or a window box. Apple cores and peels become vinegar. Vegetable trim becomes stock, and the stock's solids become compost. Keep a note of what it took: time, heat, a jar, a strainer, a trip to the shop for something you did not have. You are measuring what a commercial kitchen measures, with a smaller bin.

Follow one stream you generate to its first exit. Culls, manure, trim, surplus dough, spent grain, packaging, coffee grounds at work, whatever your household or your market stall produces in quantity. Find out who takes it, whether they pay you or charge you, and how far it travels. Most people have never asked, and the answer is usually a business they had never heard of.

Ask three questions of whoever takes it. Where does it actually go. Who owns the match between what you have and who wants it. What is the number, and who reported it. Those three questions will sort genuine infrastructure from a green logo, and they work the same way whether the answer is a farmer, a charity, a platform, or a municipality.

The situation: one waste, one week

Pick one thing your kitchen throws away and give it a second life before the week is out, working at whatever scale you can manage. You need something in surplus, one process you have not tried, and a note of what it cost you: twenty minutes, a stovetop, a trip across town. The point is not the recipe. The point is that you will discover, in your own hands, what every processor in this chapter already knows: the leftover arrives faster than you can use it, the process has a fixed cost that only makes sense if you have a lot of the thing, and the result is only an ingredient if somebody wants to eat it.

Then take the version you can actually reach, which may be the one above and may be this one: a stream that leaves your household, your block, or your operation on a schedule. Find its first exit and stay with it until it becomes something. If it goes to an animal, go and see the animal. If it goes to a charity, ask what happens to what the charity cannot move. If it goes into a bin, find out where the bin empties. Write down the chain, and mark the point where the value is claimed.

Run this in a room and the chains will look nothing like each other, and the point where value gets claimed will be the same. In every version, somebody arranged the match, and that somebody decided what the waste was worth. The heap behind the shed still answers to anyone who owns a pig, which is the good news in this chapter, and some of the machinery that moves it now answers to somebody who owns a subscription, which is the part the rest of this book has to keep asking about. Some of that machinery has a locked door on it as well, and the next chapter starts there.