Food-waste apps promise to “save” meals from the bin — and take a fee on every bag. The framing feels like impact, yet in practice these platforms often act as late-stage clearance channels with opaque effects on actual waste reduction. Too Good To Go’s latest annual report, issued in Danish kroner (DKK), reports 2024 revenue of DKK 725 million and describes a mix of transaction-based and subscription-based pricing with an annual administrative fee for outlets — a model in which “rescue” scales with sales volume.
That creates a built-in tension: genuine prevention would shrink the very activity the platform monetizes. Meanwhile, the global context is stark: countries still lack consistent, audited baselines to track real progress on waste, even as household, food-service, and retail sectors together discarded roughly one-fifth of food available to consumers in 2022.
This report follows the money, interrogates the incentives for overproduction, and sets out what transparent, audited impact would really look like — globally, and with named sources.
| Aspect | Details |
|---|---|
| Trend Name | Food-waste apps as “rescue” platforms |
| Key Components | Surprise/Magic Bags, per-bag commissions, subscription fees, annual admin fee |
| Spread | Global footprint across Europe, North America, selected additional markets |
| Examples | Bakeries, restaurants, supermarkets, quick-service chains |
| Social Media | #FoodWaste, #TooGoodToGo, #Rescue, #Sustainability |
| Demographics | Urban, price- and value-oriented consumers; strong youth adoption |
| Wow Factor | Do-good discount with a gamified surprise effect |
| Trend Phase | Mature growth with regulatory and measurement questions |
The seductive “rescue” narrative
“Rescue” is a powerful word. It signals urgency and moral clarity, implying that each bag represents waste avoided and emissions averted. In reality, a “saved meal” is a marketing claim, not a measured outcome. Without a baseline for how much a given outlet would have discarded absent the app — and without a control group to compare non-app peers — we cannot tell prevention from late-stage discounting. The feel-good framing risks confusing clearance with impact.
Psychology deepens the problem. Moral licensing describes how doing something “good” can license behavior that cuts against the original goal. In the context of food-waste apps, that can show up as consumers buying more than they need because it “helps,” or as operators normalizing the app as part of everyday sales planning because it “rescues.” The surprise mechanic amplifies this: buyers cannot easily compare like-for-like prices, and the treasure-hunt vibe nudges impulse purchases. The result may be higher transaction counts — the very metric the platform monetizes — without clear proof of less waste overall.
Globally, public agencies argue for rigorous baselining precisely to avoid this confusion. The UNEP Food Waste Index 2024 stresses measuring retail, food service, and household waste with consistent methods so progress can be tracked and verified . Until platforms and partners publish comparable, audited before-and-after data, “X million saved meals” remains a volume indicator, not an impact metric.
Follow the money: how the model really works
Too Good To Go’s own disclosures show a business built on transaction volume. In 2024, the company reported revenue of DKK 725 million and a profit of about DKK 9.6 million, while explaining that it earns through a combination of transaction-based and subscription-based pricing, and that each participating outlet pays an annual administrative fee for the license. The marketplace terms further specify a “Reservation Fee” deducted from the merchant’s payout, typically 25% per bag with a minimum fee per transaction.
This model aligns platform revenue with the number and price of “rescued” bags. More bags sold, at higher price points, mean more income for the platform. That is not inherently problematic: two-sided marketplaces routinely monetize transaction flow. The tension arises because true prevention — upstream changes that keep food from becoming surplus at all — would reduce the available bag volume. A platform can tout prevention rhetorically, yet its P&L still depends on transactions happening downstream.
Payment aggregation and visibility give the platform structural power. By handling payments, setting fee rules, and governing discovery in the app, the marketplace can nudge both consumer demand and merchant supply. In practice, that turns “rescue” into an instrument that behaves like a sales channel. As long as fees and growth are tied to downstream transactions, the system has a financial bias toward moving surplus rather than eliminating it.
The overproduction incentive trap
Consider how a predictable residual value changes planning. Without an app, surplus is a sunk cost. With an app, that same surplus acquires a reliable outlet at a known take-rate. Even after the commission, the merchant recovers more than zero — and that matters for week-to-week decisions. When late-day demand is volatile, a rational operator may slightly overproduce to avoid stock-outs, knowing that the app can sweep unspent inventory into surprise bags.
This does not require bad faith. Even if terms forbid producing extra “for the app,” enforcement is difficult because production planning is internal. A bakery that aims to meet the evening rush might add a safety margin; a supermarket might raise order quantities ahead of a busy weekend; a quick-service site might portion generously at lunch knowing unsold items can move in bundles. The line between real rescue and planned surplus blurs not because operators cheat, but because incentives point that way.
To be clear: the presence of an outlet does not prove overproduction. But the incentives are textbook. A consistent, monetizable salvage pathway lowers the cost of forecasting errors and reduces the perceived risk of leftovers. In many contexts, that encourages slightly higher production than a world with no salvage option. Only audited, pre/post data at the outlet level — with explicit baselines for discard, donation, staff meals, and app sales — can confirm or refute this effect. Until then, the trap is a plausible, mechanism-level concern, not an accusation.
Another sales channel, not a waste solution
In daily operations, the app already functions like a sales and discovery channel. Visibility in the marketplace attracts new guests; the surprise mechanic sets a low reference price; repeat usage builds a kind of CRM loop. From a merchant’s perspective, that can be useful. From an impact perspective, it raises three questions: does the channel cannibalize full-price sales, does it displace donations or staff meals, and does it meaningfully reduce net waste?
The first question — cannibalization — is often context-dependent. If an app bag substitutes for a full-price purchase the buyer would have made later that week, then the platform has moved revenue timing rather than eliminated waste. If, instead, the bag moves items that would otherwise fail a freshness window and be discarded, the net effect can be positive. Yet the surprise format makes substitution hard to measure because contents vary, and price comparisons are murky. Absent product-level tracking, merchant anecdotes will dominate.
Second, displacement matters for equity. If an outlet historically donated its surplus or fed staff after service, monetizing that surplus through a platform could shift food away from recipients who paid nothing to recipients who can afford a discount. That change may be rational for the outlet and perfectly legal. But it is not the same as preventing waste. The app’s narrative stresses climate impact; the social redistribution effects deserve equal scrutiny. Transparent accounting of where surplus would have gone without the app is essential if we want to claim net benefit rather than mere reallocation.
Displacement and the hidden costs
Waste happens across the chain. Some losses are upstream (field, storage, logistics), others downstream (retail, food service, households). Platforms deal mostly with the downstream end, so their frame of reference is inherently narrow. They can shine at last-mile matching — the right place and time to move perishable items. But without guardrails, the same system can re-route surplus from free channels to paid ones, or normalize “planned surplus” as part of revenue management.
For households, the surprise format can lead to mismatches between what buyers receive and what they actually cook. A bag of discount pastries solves tonight’s craving but may crowd out planned meals, raising the risk of later spoilage at home. Again, we need data: Do frequent app users throw away more or less at home? Do they shift purchases away from fresh items with longer shelf lives? The platform could answer this with opt-in, privacy-safe research; until then, the rebound risk is conceptually strong, empirically weak.
At a societal level, cheap “rescued” food can feel like a win, yet it may externalize costs. If platforms capture value that previously flowed to food banks or community fridges, public budgets and volunteers shoulder more of the burden to feed those in need. None of this is inevitable. It depends on design choices: donation-first routing, caps on planned app volumes, or rules that prioritize truly unsellable items. But the hidden costs remain hidden until someone measures them.
The evidence gap — and a path to real impact
The strongest claims require the strongest evidence. Today, the platform’s most visible headline — “meals saved” — is a throughput metric. It reflects activity, not net prevention. What’s missing are three layers of proof: (1) outlet-level baselines of discard, donations, staff meals, and app sales; (2) pre/post changes after adoption; and (3) displacement analysis to show that app transactions replace waste rather than full-price sales or free redistribution. Without those, the climate and social claims are suggestive, not conclusive.
Fortunately, we do not need to reinvent methodology. UNEP’s Food Waste Index provides a transparent, internationally recognized framework for measuring retail, food-service, and household waste — establishing the exact baselines that impact claims require . A credible platform could align merchant reporting to these categories, audit a representative sample with independent verifiers, and publish longitudinal results. That would transform “we saved X meals” into “we reduced net waste by Y% at participating sites,” which is what matters.
Company disclosures show that the financial plumbing is already sophisticated. The 2024 annual report states that revenue comes from transaction- and subscription-based pricing and that outlets pay an annual administrative fee; it also reports geographic revenue splits and a modest profit. The marketplace terms explain the 25% Reservation Fee (with a minimum) deducted from merchant payouts Together, those sources invite a simple governance fix: pair financial KPIs with prevention KPIs, and make the latter the north star for product decisions. For example:
- Baseline-first onboarding. Before the first bag sells, capture three months of discard/donation/staff-meal data per outlet, using UNEP-aligned categories.
- Prevention-weighted fees. Reduce take-rates for outlets that demonstrate year-over-year reductions in net surplus, verified by third parties.
- Donation-first routing. If an item meets donation criteria and a charity pickup is feasible, the app should default to donation before listing.
- Planned-surplus caps. Set a rolling cap on predictable daily volumes to deter routine overproduction, with exceptions for genuine high-variance days (storms, strikes).
- Product-level transparency. Disclose typical bag composition ranges so buyers can plan meals and reduce home waste.
None of this kills the business; it future-proofs it. It also aligns the platform’s incentives with society’s interest in prevention over clearance.
What consumers, operators, and policymakers can do now
Consumers. Treat “rescue” like a tool, not a virtue signal. Buy only what you can cook this week. If a bag’s contents don’t fit your meals, skip it. Prefer outlets that disclose bag composition and participate in donation programs. Reward transparency with repeat business.
Operators. Before using an app, define your baseline: what is typically discarded, what is donated, and what staff consume. Publish your internal goal to shrink surplus over time. Use the app as a safety valve for genuine leftovers, not as a crutch for loose forecasting. Communicate clearly with your team: the goal is less waste, not more bags sold.
Policymakers and funders. Require standardized reporting for platforms claiming climate impact. Tie public endorsements or procurement access to evidence of net prevention, not volume. Encourage data-sharing pilots between platforms and food banks to minimize displacement. Fund independent audits that align to UNEP’s measurement categories and make the results public.
Company perspective, fairly stated
Platforms often reply that they do not create waste; they redirect it. They note that their terms prohibit producing extra “for the app,” that they facilitate millions of meals that would otherwise be discarded, and that they invest in awareness and labeling initiatives. They may also argue that any channel that makes it easier to move short-dated food lowers the probability of disposal. Those points deserve space in the debate.
Yet the core issue remains unresolved without evidence. If app adoption coincides with lower net waste per outlet — after accounting for displacement and cannibalization — that is a success worth celebrating. If not, then the app is primarily a smart clearance mechanism with mixed equity effects. Measuring the difference is not an attack; it is how a sector that invokes climate and social benefits earns trust.
Ready to go beyond feel-good “rescue” and tackle root causes—start here: Are restaurants fighting the wrong enemy? The trouble with zero-waste trends.