Foogly: Waste Is a Market Failure, Not a Logistics One

01Executive Summary

The Argument, Up Front.

The Argument

There is a kind of waste where the goods exist, are perfectly usable, and somebody would happily pay for them, and none of that matters because there is nowhere for the two parties to meet. This whitepaper argues that a surplus marketplace architecture is first an answer to a missing market rather than a missing route, and that no amount of supply chain optimisation substitutes for building the place where a price gets agreed. The Foogly implementation is the evidence: one stream of farm material that previously earned nothing, turned into three saleable products with three different buyers.

Draw the distinction precisely, because the whole paper rests on it. A logistics failure means the transaction is agreed and the goods cannot get there economically. A market failure means the goods and the willing buyer both exist and no transaction is ever attempted, because no venue, no pricing mechanism and no discovery route exist.

The two get confused because waste looks like a movement problem from outside. But the reason nobody collected the material is usually that nobody had sold it, and the reason nobody had sold it is that there was no buyer at a price worth the trouble.

Four patterns follow from that misdiagnosis and this whitepaper rejects all four: treating a market failure as a logistics problem, treating the stream as one product, building a seller console for people who are not sellers, and selling perishable surplus on a catalogue cadence.

Five principles resolve it. Diagnose whether you are missing a market or a route. Segment by buyer rather than by material. Recognise that who sets the price defines what you have built. Match supply-side effort to the supply side's actual job. And treat notification as inventory management where stock is perishable.

NewAgeSysIT built the platform described here under the strategic advisory guidance of Giovanni Livia, as a multi-category marketplace rather than a collection service with a website attached.

One boundary belongs here rather than at the end. This is an economic argument, not an environmental one, and the paper keeps it that way throughout.

02The Market That Was Never Built

The Goods Exist. The Buyer Exists. Nothing Happens.

There is a particular kind of waste where the goods exist, are perfectly usable, and someone somewhere would happily pay for them, and none of that matters because there is nowhere for the two parties to meet. That is not a logistics failure. It is a missing market, and no amount of supply chain optimisation will fix it.

Draw the distinction precisely. A logistics failure means the transaction is agreed and the goods cannot get there economically. A market failure means the goods and the willing buyer both exist and no transaction is ever attempted, because no venue and no pricing mechanism exist. Those need entirely different products.

Why the two get confused. Waste looks like a movement problem from outside. But the reason nobody collected the material is that nobody had sold it, and the reason nobody had sold it is that there was no buyer at a price worth the trouble.

The shape this takes in practice. A producer holds material that falls outside their normal channel, which wants uniform quantities to a standard specification. There is no second channel. So the material is written off, and the write-off is recorded as a cost of doing business rather than as unrealised revenue, which is the accounting decision that makes the whole thing invisible.

Two Failures, One Diagnosis They look the same from outside and need entirely different products

Logistics failure

The transaction is agreed and the goods cannot get there economically.

You are building a route.

Market failure

The goods and the willing buyer both exist and no transaction is ever attempted, because no venue and no pricing mechanism exist.

You are building the venue.

Why the write-off stays invisible

The material is recorded as a cost of doing business rather than as unrealised revenue, so nothing in the accounts ever asks who might have bought it.

↓

Building the venue is the product

The place where a price gets agreed

Not the collection, not the grading, not the transport

The measured research here is unusually good, and it is worth reaching for the measured version rather than the widely circulated one. Johnson and colleagues at North Carolina State did what most food-loss work does not: they went into fields after the commercial harvest had finished and weighed what was left. Across 68 fields of eight vegetable crops on nine commercial farms in eastern North Carolina, published in Agricultural Systems in 2018, samples were sorted using USDA quality indicators into marketable, edible but not marketable, and inedible. Combined, the marketable and edible-but-unmarketable averaged 42% of the yield the growers had actually sold. An earlier field study on a single 121-hectare farm, in Resources, Conservation and Recycling, put the unutilised share at 57% of reported marketed yield. The authors are blunt about why the work was needed: existing estimates rest on assumptions rather than field data. That caution applies to almost every food-waste number in circulation, and it is why this paper cites measurements and stops there.

The observation that makes the rest actionable is buried in that sorting. The largest category was edible but not marketable, meaning produce that does not meet appearance quality standards. Uniformity is what the existing channel requires, not what a buyer requires. Once you stop assuming the buyer must be the same kind of buyer, a stream that looked like one low-grade product often turns out to be several products for several people.

The Reframe

Building the venue is the product. Not the collection, not the grading, not the transport. The thing that does not exist is the place where a price gets agreed, and everything else is downstream of it existing.

What Follows

If the missing thing is a market, the design questions are who the buyers actually are, who sets the price, and how much the supply side is willing to do. Section 04 answers all three.

Sources

Johnson, L.K., Dunning, R.D., Gunter, C.C., Bloom, J.D., Boyette, M.D. and Creamer, N.G. (2018), "Field Measurement in Vegetable Crops Indicates Need for Reevaluation of On-Farm Food Loss Estimates in North America," Agricultural Systems 167:136-142. 68 fields, eight vegetable crops, nine commercial farms in eastern North Carolina. Samples sorted using USDA quality indicators into marketable, edible but not marketable, and inedible. Marketable plus edible-but-unmarketable averaged 42% of marketed yield.
Johnson, L.K. et al. (2018), "Estimating On-Farm Food Loss at the Field Level: A Methodology and Applied Case Study on a North Carolina Farm," Resources, Conservation and Recycling 137:243-250. 13 fields on one 121-hectare farm. Unutilised share averaged 57% of reported marketed yield.
Both studies measure unsold output. They do not measure waste avoided, emissions or farmer income, and they are not presented here as doing so.

03Why Conventional Approaches Fail

Four Patterns, and One Diagnosis Error Behind Them.

Four patterns explain why surplus keeps being written off in categories where somebody would clearly have bought it, and the first one is a diagnosis error that makes the other three inevitable.

01

Failure Mode 1

Treating a Market Failure as a Logistics Problem

The surplus is understood as a handling problem, so the effort goes into collection, grading, storage and routing.

Every one of those is real work and none of it produces a buyer. A well-optimised route to nobody is still a write-off, now with an operating cost attached.

Name why the misdiagnosis is natural rather than careless. Logistics problems are legible. They have owners, budgets and known solutions, and somebody's job description covers them. A missing market has none of that, so nobody is responsible for noticing it is missing.

The diagnostic fits in one line, since Principle 1 argues it properly. If you can name a buyer who would pay and the sale still does not happen, it is a venue problem. If you cannot name the buyer at all, building the venue is the work.

02

Failure Mode 2

Treating the Stream as One Product

The surplus is characterised by what it is made of rather than by who might want it, so it goes to market as a single low-grade category at a single low price.

Mixed material sold as one thing is priced at the value of its least valuable component, and the buyers who would have paid properly for one part never see that part on its own.

The reframe is the most portable idea here: segment by buyer, not by material. The question is not what this stuff is. It is who wants which part and what for. Only the second question produces revenue.

There is a consequence for the platform rather than the business. Once the stream is several products, they share a checkout and very little else. Description, pricing, storage, shelf life and delivery all differ, and a data model assuming one shape fights every difference.

03

Failure Mode 3

Building a Seller Console for People Who Are Not Sellers

The supply side is given the tools a merchant would want, because that is what marketplace software looks like: catalogue management, listing optimisation, inventory levels, pricing controls.

The supply side in a surplus market is usually somebody whose actual job is production. They did not want to become a retailer. They wanted the material gone at a fair price, and every extra task is a reason to go back to writing it off.

Name the misjudgement precisely. The platform has assumed its supply side shares its enthusiasm for the platform. They do not. They have a day job, and the surplus is a side effect of it.

Where a supply side does think of itself as sellers, the opposite applies and they need more product rather than less. That case is argued in the Honu whitepaper in this series. The variable is self-perception, not size.

04

Failure Mode 4

Selling Perishable Surplus on a Catalogue Cadence

Standard commerce rhythms are applied to stock that is unpredictable and short-lived: a browsed catalogue, a campaign calendar, a weekly newsletter.

The stock that arrived this morning may be gone or past its window by the time the campaign goes out, and the customer who would have bought it never knew it existed. The catalogue is describing a shelf that has already changed.

The reframe: in a surplus category, telling customers what has just arrived is not marketing. It is how the inventory clears, and it belongs in the operational design rather than in the growth plan.

The second-order consequence is the interesting one. Because the notification is doing inventory work, what determines its value is timeliness rather than persuasiveness. That is an unusual position for a marketing channel to occupy, and it changes who inside the business should own it.

Pattern	What it assumes	Where it breaks	What it costs
Market failure as logistics	The buyer exists and cannot be reached	A well-optimised route to nobody	A write-off with an operating cost attached
One product	The stream is defined by its material	Priced at its least valuable component	Every buyer who would have paid properly
Seller console	The supply side wants to be merchants	Their day job is production	The supply side, back to writing it off
Catalogue cadence	Stock is stable enough to describe in advance	The shelf changed before the campaign shipped	The stock clears too late to be worth anything

04Foogly: Five Principles for Trading Surplus

Building a Market Where None Existed.

Building a market for material nobody currently trades means answering three questions an existing market would have answered for you: who the buyers are, who sets the price, and how much the supply side is prepared to do. Nothing can be copied, because there is nothing to copy from.

Sector-Agnostic

What follows is sector-agnostic. Someone holding manufacturing offcuts, returned goods, short-dated stock or industrial by-product could apply it without ever seeing Foogly.

Principle 01

Diagnose Whether You Are Missing a Market or Missing a Route

Before building anything, establish whether the transaction fails because it cannot be fulfilled or because it is never attempted.

Implementing it is concrete. Try to name a specific buyer who would pay a specific price for a specific part of the stream. If you can and the sale still does not happen, the gap is discovery or fulfilment and you are building a route. If you cannot name them at all, the gap is the market itself, and the first work is finding those people rather than writing software.

What it buys is that it stops you optimising a chain with nothing at the end, which is the most expensive way to fail here because it looks like progress throughout.

Name the honest outcome, because this paper needs one. Some surplus really is a logistics problem, and some has no buyer at any price. A stream failing both tests should be written off, and finding that out cheaply is a good result rather than a disappointing one.

Principle 02

Segment by Buyer, Not by Material

A waste stream is a portfolio. Split it by who would want each part and what for, rather than by what it is made of.

Implementing it means working backwards from candidate buyers to the parts of the stream that serve them, then carrying those as genuinely separate product types that happen to share a checkout. They differ in description, pricing, storage, shelf life and delivery, and a data model assuming one shape will fight every difference.

What it buys is that each part is priced at what its own buyer will pay rather than at the value of the least valuable thing in the mix.

The diagnostic question is the sentence to take away: what would somebody have to want in order to pay properly for this specific part? Ask it once per part, not once per stream.

Principle 03

Who Sets the Price Defines What You Have Built

Pricing authority determines whether you are operating a marketplace or a storefront, and it is routinely treated as an implementation detail when it defines the business.

Set out the two options honestly, because this is a genuine fork rather than a recommendation with a caveat. In a managed model the platform prices the goods, so it carries the pricing risk, needs category expertise it must acquire and keep, and offers the supply side a very simple proposition. In a direct model the supply side prices, which scales considerably further, costs the platform no pricing judgement, and requires sellers willing and able to price.

What the managed model buys is the strongest argument for it. A producer who has never sold this material before, and by definition nobody has, faces the hardest part of the transaction first: discovering what it is worth. Handing that to the platform removes the main reason not to participate. What it costs is that the platform is now in the pricing business, which needs judgement, carries margin risk, and does not scale the way a self-serve listing model does.

So the principle is not that one model is correct. The choice is architectural and largely unrecoverable, because margin, dispute handling and the whole supply-side experience sit on top of it. Treating it as a feature toggle means deciding it without anyone weighing the trade.

Principle 04

Match Supply-Side Effort to the Supply Side's Actual Job

Ask the supply side for the least that makes the transaction possible, and decide what that is by looking at what they do all day rather than at what a merchant would tolerate.

Implementing it means a flow consisting of describing what you have and getting an answer, with everything else absorbed by the platform. No catalogue, no inventory management, no listing optimisation.

What it buys is participation. A supply side whose main work is production will use a platform that costs ten minutes and abandon one that costs them a role they did not apply for.

Now the distinction that matters most here, and it is a disagreement with another paper in this series. The Honu whitepaper argues the opposite: that a marketplace's supply side needs a real product rather than a listing, with calendar control and accumulating reputation. Both are correct, and the variable is self-perception rather than size. A swimming instructor is running a business and wants control. A farmer is growing food and wants a number. Ask which one you have before deciding how much to build them, because building the wrong one loses the supply side either way.

Principle 05

When Stock Is Perishable and Unpredictable, Notification Is Inventory Management

Where availability changes daily and does not last, telling buyers what has just arrived is an operational function rather than a marketing one.

Implementing it requires notification wired to the moment stock enters the system rather than to a campaign calendar, and segmented enough to reach the people who want that specific category.

What it buys is that the material clears while it is still worth something. That is the entire economics of a perishable surplus business and there is no second chance at it.

The organisational consequence is worth acting on. Because the value of the message is its timeliness rather than its persuasiveness, this channel should not be owned by whoever owns campaigns. It belongs with whoever owns stock. The same logic covers any inventory with a hard expiry, including unsold capacity and time-limited access.

The Ordering Matters

Principle 1 tells you whether there is a product here at all. Principle 2 decides how many. Principles 3 and 4 together decide what kind of business you are running, and they are linked, because how much the supply side will do depends on whether you are willing to price for them. Principle 5 keeps the goods moving once the market exists.

05Evidence: Foogly in Practice

One Stream and Three Buyers.

NewAgeSysIT implemented this approach for Foogly, a platform that takes one stream of farm material which previously earned nothing and gives it three routes to a buyer, under the strategic advisory guidance of Giovanni Livia.

The starting point, led by the market failure rather than the scope. A significant share of what a farm grows never reaches anyone. Some of it fails retail appearance standards, which has nothing to do with whether it is good to eat. Some is by-product for which no buyer exists at all. The growers' normal channel wants uniform produce in predictable quantities, so there is no second channel and the material is written off.

Principle 2 in practice, and the core evidence

Three categories, each defined by its buyer rather than by its contents. Discounted fresh produce, for someone who wants cheap good food. Compost and fertiliser made from by-product, for someone who wants to improve their soil. Pet food made from surplus fruit and vegetables, for someone who wants to feed their dog better than kibble. The same material, three buyers, three prices. The by-product line is the cleanest demonstration of the whole thesis, because that material had no market at any price before the platform gave it one, and nothing about the material itself changed.

Principle 5 in practice

Real-time notifications carry new availability out to customers as stock arrives rather than on a campaign schedule, because in this category availability changes daily and is often short-lived. The notification is doing inventory work, not promotional work.

The supply side

A farmer registers and describes what they have by type and quantity, and the platform takes it from there. That is deliberately as far as this document goes, because who sets the price is a question the source material does not settle, and Principle 3 explains why guessing at it would misdescribe the business rather than just the feature.

The customer side

Two-field registration, browsing across all three categories, one cart, and a choice of store pickup or delivery at checkout, with Stripe handling payment. The short registration matters more here than it looks: someone arriving to buy discounted vegetables will not work through a long sign-up in order to do it, and a market nobody can enter easily is still not much of a market.

One thing the three categories share

And almost nothing else. They sit behind a single checkout and a single account, and that is the extent of it. Their descriptions, price logic, storage assumptions, shelf lives and delivery expectations all differ, which is the shape Principle 2 predicts and the reason a data model built for one product type would have been fought at every one of those points.

The stack, named narrowly

React Native, NestJS, MongoDB and Stripe. The relational database and the web framework are not named, because the sources conflict on both and this document's reader is the kind who checks.

Where the detail lives

Full delivery detail and all five components are published in the Foogly case study.

06What the Evidence Shows

Structural Evidence, and No Invented Impact.

This implementation published no performance figures, and in a sustainability-adjacent category the temptation to estimate some should be resisted rather than managed. What follows is structural evidence: what the platform made possible for material that previously had no buyer.

Evidence	Before	After	What it evidences
Routes to a buyer	One channel wanting uniform produce in predictable quantities, and nothing else	Three product categories, each defined by a different buyer	Principles 1 and 2 together. The most consequential decision in the build
By-product destination	Material with no market at all, written off	Compost and fertiliser sold as a product in its own right	The clearest single demonstration of the thesis: the material did not change, the market did
Stock visibility	Availability changing daily with no way to tell anyone	Real-time notification as stock enters the system	Principle 5. Operational rather than promotional
Buyer access	No consumer route to farm surplus at all	Mobile and web, two-field registration, pickup or delivery	Not a principle, and it is what makes the rest reachable
Supply-side workload	⚠ Held. Depends on the pricing model	⚠ Held. See block C	Principles 3 and 4. Not written either way until the question is settled

Address the absence of figures directly rather than apologising for it. In a sustainability-adjacent category, a paper that invents an impact number is doing exactly what evidence-light environmental claims have always done, and a reader here has seen it many times. The restraint is not modesty. It is the only position from which the rest of the document is believable.

Each row traces to a decision rather than to general competence. The three categories exist because somebody asked who would want each part, rather than what the material was. The by-product line exists because that question was asked about a part of the stream nobody had ever tried to sell. The notification is timely because it was wired to stock rather than to a calendar.

The second row is the one to lead on, and it is worth separating from the first. Three categories is a segmentation decision and a reader could reasonably ask whether the market was already there. The by-product line cannot be read that way. Nobody was buying that material at any price, in any channel, before there was somewhere to sell it. The material did not change, the processing did not make it into something else, and a market appeared where there had been none. That is the thesis in its narrowest and most defensible form, and it is the row a sceptical reader should be pointed at first.

State plainly what the evidence does not support. No volume traded. No farmers onboarded. No split of sales across the three categories. No repeat purchase rate. The architecture is evidenced. The size of the market it created is not, and this document does not claim it.

The figure worth chasing, and why this one. The split of sales across the three categories. It would show whether the segmentation actually worked, which is the paper's central claim, and it carries no environmental assertion whatsoever. It is a revenue breakdown.

Boundary Condition

This approach earns its cost where the stream genuinely contains several buyers. Where it does not, splitting it produces three thin catalogues instead of one good one, and the single-category route is correct.

07Implementation Considerations

Six Decisions Before the First Listing.

Six decisions determine whether a surplus platform creates a market or simply moves material around more efficiently. The first one costs nothing and is almost never made deliberately.

#	Decision	What to weigh
1	Market failure or logistics failure	Try to name a buyer and a price
2	How many buyer types the stream contains	Each additional buyer is a revenue line and a product surface
3	Who sets the price	Not a feature toggle. A different business
4	How much the supply side will actually do	Look at their day job, not at what a merchant would tolerate
5	How buyers learn what just arrived	Wired to stock, or wired to a calendar
6	What you claim about impact	Describe the mechanism and let the reader conclude

1. Market failure or logistics failure.

Try to name a buyer and a price. If you can and the sale still does not happen, build a route. If you cannot, the market is the missing thing and finding the buyers comes before any software. Skipping this means optimising a chain with nothing at the end, which looks like progress the entire time and is usually discovered only when the finished system produces no transactions.

2. How many buyer types the stream contains.

Ask what someone would have to want in order to pay properly for each part. The answer is often more than one, and each additional buyer is a revenue line. It is also more product surface, so a stream that genuinely has one buyer should stay one category. This one is close to unrecoverable: the number of categories shapes the catalogue, the pricing model and the fulfilment logic, and adding a second later touches all three.

3. Who sets the price.

Managed means you carry the pricing risk and need category judgement, and it gives the supply side almost nothing to learn. Direct scales further and requires sellers who can price. This is the other close-to-unrecoverable decision, and it is not a feature toggle. It is a different business, and it should be settled before anything is built, because margin, disputes and the whole supply-side experience sit on top of it.

4. How much the supply side will actually do.

Look at their day job rather than at what a merchant would tolerate. Ask for less than you think, then ask whether even that is too much. The exception matters: where the supply side sees itself as sellers, they need more rather than less, and building them a thin experience loses them just as surely as overloading a producer does.

5. How buyers learn what just arrived.

With perishable or unpredictable stock, notification is inventory management and belongs with whoever owns stock. On a campaign calendar it is describing a shelf that has already changed, and the cost of deferring this is measured in material that cleared too late to be worth anything. It is also the cheapest of the six to get right, which is why it is worth checking early rather than treating as a growth-phase concern.

6. What you claim about impact.

Describe the mechanism and let the reader conclude. In this category an unsupported impact claim costs more credibility than it gains, because the audience has been trained by years of them. The last clause is the useful one, and it is architectural rather than editorial: if you intend to claim impact later, you have to instrument it now. Most platforms discover that after somebody has already made the claim, at which point the measurement cannot be backdated.

Two Can Conclude Against Building

Two of those six can legitimately conclude against building anything. Decision 1 will sometimes find the surplus should be written off, and Decision 2 will sometimes find there is only one buyer. Both are correct outcomes, and a paper that cannot reach them is a brochure.

08Key Takeaways

Five Insights, Each Independently Quotable.

Editorial Note

⚠ These five travel without their context, and this is the category where an extracted sentence is most likely to be read as an environmental claim. None asserts impact, quantifies waste avoided, or implies certification. Re-check any rewrite the same way.

01 The Diagnosis Claim

Waste is a market failure before it is a logistics one. The goods exist, they are usable, and somebody would pay for them, and none of that produces a transaction because there is nowhere for the two parties to meet. Building that place is the product.