Two Dublin Surprise Bag listings use the same word and promise very different things.
A Krispy Kreme listing in Tallaght says the bag holds six doughnuts chosen from the cabinet at the end of the day. It does not guarantee the flavours, though the team says it will try to select a mix. An ALDI listing in Nutgrove says the bag is a selection of groceries that could include fruit and vegetables, cakes and pastries, snacks, and fish and meat, depending on what has not sold that day.
Both are Too Good To Go Surprise Bags. One tells you what the bag is and leaves the detail open. The other leaves much more open. Neither is wrong. The difference shows a decision every surplus marketplace has to make: how much a buyer should know before accepting a surprise, and what it may cost the seller to say so.
What Too Good To Go is, and why this matters commercially
Too Good To Go is a marketplace where businesses list unsold stock as discounted Surprise Bags. People reserve a bag in the app and collect it in a set window. In its October 2026 announcement the company reported 150 million registered users and 200,000 active partners worldwide. Registered users are not active purchasers.
The format solves a coordination problem. A business may have small, unpredictable quantities of unsold items, and a bag avoids cataloguing each one. That is our reading of the format's operational advantage, not a measured benefit.
The commercial question follows. A discounted first purchase becomes more useful to the marketplace when the buyer wants to come back and the merchant keeps finding participation worthwhile. Whether better information improves that outcome is a hypothesis, and testing it is the point of this note.
We read public listings, the company's announcements and Irish user terms, and dated reviews. We have no access to Too Good To Go's data and have not spoken to anyone there.
The short answer
Too Good To Go has to balance two needs that pull against each other. The buyer needs enough confidence to judge whether a bag will be useful to them. The merchant needs enough flexibility to put unpredictable surplus into a bag without extra work or a promise they cannot keep.
Improving one side at the other's expense can weaken the marketplace, although some trade-offs may leave the overall outcome better. More certainty for buyers can raise cost and risk for merchants. More flexibility for merchants can leave buyers with bags they cannot use. The useful question is which information most improves a buyer's decision for the least burden on the merchant.
What a listing can say
A listing can carry four kinds of information. They differ in how stable they are, which decides what they cost to keep accurate.
| Kind of information | Example | Stability |
|---|---|---|
| Stable description | Doughnuts, bakery or mixed groceries | Changes rarely |
| Reliable commitment | Six items, where the merchant can consistently deliver it | Holds day to day, or the bag is cancelled |
| Changing information | Today's actual contents, once the merchant knows them | Changes daily |
| Remaining uncertainty | The forecast surplus, an unconfirmed mix | Cannot be settled in advance |
The last row is not "nothing can be said". A merchant can communicate uncertainty, for example that the mix varies or that several of one item are possible, without promising an outcome.
The two Dublin listings spend this differently. Krispy Kreme gives a stable description (doughnuts) and a commitment (six), and states the flavours as a best effort. ALDI gives a broad stable description and describes the contents as depending on unsold items.
This comparison does not show that one format converts or retains better. The categories, customers and inventories differ. It shows that merchants already set the line in different places, and the grocery listing leaves more dimensions open than the doughnut one.
What the buyer needs confidence about
A bag can fail a buyer in two different ways.
- A fulfilment problem: the bag is worth less than the value promised, or the pickup fails.
- A suitability problem: the bag meets its promise but does not suit the person collecting it.
A review on the Irish App Store page shows the second kind. A review dated 3 April 2025 describes repetitive baked goods, quantities unsuitable for one person and a wish for more useful information before choosing. A 2023 review describes broader dissatisfaction and says the reviewer stopped using the app. It does not isolate suitability. The same page shows an overall rating of about 4.9 out of 5.
These are individual, self-selected accounts. They show that a reported experience exists. They do not show how common it is or whether it affects who returns, and the high rating is a necessary counterweight to any claim of widespread dissatisfaction.
The distinction still matters because the fixes differ. Better refunds help when a bag is worth less than promised. They do little for a household that cannot use twelve of the same pastry.
What the merchant needs flexibility for
The Irish user terms, effective 31 July 2024, accept uncertainty as part of the service. They note that stores do not always have the surplus they expected, and that stores can generally cancel up to an hour before pick-up, and sometimes later if they sell out. They say Too Good To Go does not control the contents of a bag, and that the store is responsible for ingredient and allergen information, which buyers should confirm with the store before collecting. They describe the stated value on the app as a minimum value had the bag not been sold as surplus.
Flexibility matters because the stock is genuinely unpredictable. If a bag has to match an exact list, merchants could respond in several ways. They could spend more time describing contents, narrow the offer, adjust how many bags they list, cancel more often when stock falls short, or stop listing. Which of these happens would depend on the business, and we have no merchant data on any of them.
What breaks when either side wins
| If buyer confidence wins | If merchant flexibility wins |
|---|---|
| Merchants may spend more time per bag | Buyers may meet mismatches they could have avoided |
| Merchants may narrow offers or cancel more often | Some buyers may stop reserving, though the size of that effect is unknown |
| Some merchants with very unpredictable stock may leave | Trust in the Surprise Bag label may weaken |
| Total redistributed surplus could fall | Food that is redistributed may go unused |
Both columns are possible consequences, not measured ones.
A rule for deciding what to disclose
Before adding a statement to a listing, ask three questions.
- Does it materially help the buyer judge whether the bag suits them? A statement that does not change a decision is only cost.
- Can the merchant provide and verify it reliably? An unreliable statement is worse than none, because it turns uncertainty into a broken promise.
- Does the benefit justify the work and the restrictions it creates? Weigh the time, the cancellations it forces and any pressure to over-commit.
Treat low maintenance as a preference, not a requirement. A stable description is cheap and often helpful, but a stable statement can still mislead, and a quick check or an automated one may be worth making daily. The reverse also holds. Some information is not a cost-benefit choice at all. Ingredient and allergen information, for example, is something the terms place on the store, and no argument about merchant effort changes that.
By this rule, product family and reliable minimum quantities are the easiest wins where a merchant can deliver them, such as a doughnut count. A statement of possible repetition or typical size is harder, because it can only be an example and never a guarantee. Item-level detail at closing may cost more than it is worth for some merchants. We have no merchant evidence on how often, and others may find it worthwhile.
Our provisional call
A listing should add a statement only if it helps the buyer judge suitability, the merchant can provide it reliably, and the benefit justifies the work and restrictions. We hold this view provisionally. We have no merchant data and no access to Too Good To Go’s figures.
A pilot with one change
The ALDI listing already names the product family, so asking merchants to confirm a family would add little. A pilot needs a treatment that adds something the current listing lacks.
We would test one change: the contents of the latest completed collection at each scheduled update, shown with both the collection date and the update date and labelled as one past example, not a promise. The selection rule matters. The merchant does not choose which bag to show, because a favourite would turn the example into advertising. The example is refreshed on a fixed schedule, such as weekly, so between updates it can be several days old, and the dates say so. It speaks to the reported mismatch because a buyer can see what one recent bag looked like, but a single bag is one outcome and not a pattern, so the listing should say so. The control is the current listing for comparable stores.
Run it in one category and area, using store-level or clustered groups, since randomising individual buyers alone can miss effects across the marketplace. Audit the existing listings and app first, because the information may already be there.
Measure both sides together:
- Buyer side: completed collections, reported usability, repeat purchase over a period set in advance, refunds and complaint types.
- New mismatches: refunds and complaints that cite the example, since an example can create expectations of its own.
- Merchant side: staff minutes to refresh the example, active-store retention, cancellations and surplus sell-through.
The pilot succeeds only if buyer gains survive the merchant costs. A lower reservation rate can be acceptable if fewer people buy unsuitable stock, but total useful redistribution must be counted too. We offer no sample size or uplift target, because both depend on baseline data we do not have.
Why the decision matters now
On 15 October 2026 the company announced that its marketplace now covers personal care, pet care, flowers and plants, nutrition and supplements, and home products. It said more than 8,500 stores already supply products in these areas, about a third of which already work with it on food, and that the mix available varies by market and store. We have not verified an Irish launch date.
That raises the stakes on the same question. A buyer may accept an unknown doughnut flavour. A pet product has to suit the animal, and some personal care items depend on shade or other personal needs. These are examples of matching constraints, not reports of unsuitable products, and the announcement does not say how the company handles them.
The case against changing anything
- Surprise is part of the appeal. Some people like not knowing, and more detail could put them off. Any change should keep an open option.
- Information may not be the main problem. If cancellations or empty local supply dominate bad experiences, those come first.
- Rare complaints may not justify complexity. The high rating and the many listings that work well mean any change needs evidence for specific segments.
- The company may already be doing this. Public material does not show every experiment or merchant tool, so look at the current product before treating any of this as new.
What we do not know:
- How common suitability complaints are compared with promised-value failures and failed pickups.
- Whether a recent example improves repeat use or creates new expectations, and what it costs merchants in time.
- The Irish fee schedule, merchant economics, cancellation rates or retention.
- How non-food listings handle compatibility. We have not tested the app or made a purchase.
Everything here comes from public pages and listings dated to 28 October 2026 or earlier. Listings, ratings and terms change, so check the live pages before relying on a detail.
Think we have this wrong?
If you have listed on Too Good To Go, run a surplus marketplace or worked on expectation-setting, send us the strongest case against this. With your permission, we will add it to the note. Email hello@productgarage.club.
Next step
Product Garage works through live product and go-to-market decisions with experienced operators. This note takes one of those decisions in public, while the thinking is still provisional.
Subscribe to Garage Dispatch to get the next working note.
FAQ
What is a Too Good To Go Surprise Bag? It is a discounted bag of a business's unsold items that you reserve in the app and collect in a set window. The business chooses the contents, so what you receive depends on what has not sold that day.
Can a store cancel my Surprise Bag? Yes. The Irish user terms say stores can generally cancel up to an hour before pick-up, and sometimes later, for example if they sell out. If you were charged, the terms say the amount is released or refunded.
What does the stated value of a bag mean? The terms describe it as a minimum value had the bag not been sold as surplus. It is a floor, not a promise that the bag will suit you.
Why do some listings say exactly what is in the bag and others do not? Merchants set the boundary themselves. A narrow format such as six doughnuts is easier to describe than a mix of whatever is unsold, and the two Dublin listings we checked show both approaches.
Is Too Good To Go expanding beyond food? On 15 October 2026 it announced products in personal care, pet care, flowers and plants, nutrition and supplements, and home. It said availability varies by market and store. We have not verified an Irish launch date.