Ideal Consumption

Doorstep Handoff

The least-effort path from a cupboard nobody opens to a person who wants the thing.

Österreichische Post already collects parcels from addresses. Its Abholservice is live and costs a flat EUR 3.48 for up to five parcels, nationwide, Monday to Friday. That single existing fact carries most of this design: the collection network does not have to be built, only used and paid for at a known price. A household sets a box outside, the round that was already passing takes it, and a local station lists what is inside.

The reason to prefer this to donating is not sentiment about waste. It is that the seller gets a cut. In Austria the three major charity channels, Humana, Carla and Volkshilfe, are all donation-only with no payout to the donor: the garment leaves and its fate becomes unknown, which is fine if you did not want it and unsatisfying if you did. A share of the sale is a real incentive, and it works especially where no charity-resale equivalent exists at all.

Load-bearing sources

  • Österreichische Post, Abholservice, 2026 [A]

    Confirmed live and current on 4 September 2026. Flat EUR 3.48 for up to five parcels, nationwide, Monday to Friday. Generic parcel logistics rather than a resale service, which is precisely why it is usable as a building block.

  • Humana Austria, Carla by Caritas Wien, and Volkshilfe, 2026 [B]

    All three confirmed as donation-only, with no revenue share back to the donor. This is the gap the seller cut is aimed at, and it was checked rather than assumed.

How the loop works

Named safeguard

Distress-selling safeguard

Nobody should sell their last shirt. A system that pays per item and asks for nothing back has an obvious failure mode: it becomes most attractive to the person who can least afford to use it. So the pattern is watched for, and the response is not a warning banner. If the selling pattern suggests real need, and depleting a wardrobe rather than clearing a surplus is what that looks like, the flow routes to livelihood resources instead of to another listing. This is a named part of the design with a place in the interface, not a line in a policy document.

Anti-dumping: one box until reputation is earned

The obvious abuse is people posting rubbish, because a free collection is a free disposal. The first mitigation is a hard cap: one box per address until reputation exists. The second is vouching, where somebody with standing in the Web of Trust puts their own standing behind you and shortens your path to a higher limit.

A lock only the courier can open

The box has a lock the postal courier alone can open, so the contents can be checked against the declaration at the point of collection without being exposed to anyone else in the chain. Verification and privacy usually trade against each other; here they do not, because the courier is already a trusted party in a way no platform is.

The seller takes a cut

The financial incentive is the mechanism, not a sweetener. It is what makes the effort of putting a box outside worth more than the effort of doing nothing, and it is what distinguishes this from a donation bin. It also gives the household a reason to declare the contents accurately, since an accurate declaration sells faster.

Not only clothes

The same loop works for any appliance or item that has resale value and no easy route to a buyer. Clothes are the first case because the volume is high and the emotional cost of discarding is low, not because the mechanism is clothes-specific.

This is not a new marketplace

Austria already has willhaben, and the wider market already has eBay and Kleiderkreisel. Re-doing any of them would be a waste of effort and would compete for the same listings that are already getting listed.

Commercial consignment already exists here too, and the design has to be honest about it rather than pretend to an empty field. Sellpy has operated in Austria since 2021, backed by the IKEA and H&M groups, paying 40 percent up to EUR 50 and 70 percent above it, with a EUR 1.20 listing fee. Rebelle covers the luxury end. Momox Fashion buys outright at a fixed price rather than sharing revenue. What none of them do is start from a measurement the household did not have to produce, and none is a neighbourhood-scale operation. The distinguishing claim is the measured record and the local station, not the existence of a payout.

The gap is one level upstream. The items this is for are the ones that never get listed anywhere at all, because the fifteen minutes of photographing, measuring and describing is more than the item is worth to the person holding it. Feeding those items into marketplaces that already have the buyers is the design, and reuse of those marketplaces is a feature rather than a compromise.

The closest existing practice is sight-unseen collection by antique dealers, who will take a whole lot without inspecting each piece. That works, which is encouraging, but it only operates at the scale of an entire household clearance. Getting the same convenience down to three garments is the actual problem.

On generating the listing photo

A listing sells better with a garment shown on a body, and a household box cannot produce that. The obvious answer is a generated on-model image, and the industry has already gone there: roughly 70 percent of Zalando editorial imagery in the fourth quarter of 2024 was AI generated, and Google ships virtual try-on as a free consumer feature in Shopping.

For this project the route is blocked, and blocked on licensing rather than capability. The three leading open virtual try-on models, IDM-VTON, OOTDiffusion and CatVTON, all carry CC BY-NC-SA 4.0 on both their code and their weights. They are self-hostable in the technical sense and non-commercial in the legal sense. Under a standing rule of permissive licences only, that rules them out of anything with a revenue share attached, which is the entire point of this loop.

The remaining options are a paid enterprise API, which puts a per-image cost on every listing and a dependency at the centre of a local loop, or no on-model image at all. For now the honest answer is the third one: a flat photograph plus measured dimensions, and a note that a permissively licensed try-on model would be a genuinely useful thing for somebody to release.

OOTDiffusion licence file on GitHub, CC BY-NC-SA 4.0, verified alongside IDM-VTON and CatVTON, 2026 [A]

Earmarked machine tax

Jakob's position, stated as a position rather than a finding.

The long-run direction for this loop is automated: robot collection, automated sorting, automated listing. That is desirable, because the friction being removed is exactly the friction that keeps items out of circulation. It also removes jobs, and the sorting and logistics work in question is disproportionately done by people with few alternatives.

The proposal is a levy that is specific and directed rather than general: a share of the profit from a given automation funds education and retraining for the people whose work that automation displaced. Not a general robot tax paid into a common budget, where it becomes indistinguishable from any other revenue and funds whatever is politically convenient.

The contrast is the whole point. General taxation redistributes without acknowledging a relationship. An earmarked levy states that this specific gain came from this specific displacement, and directs the proceeds accordingly. It is harder to administer and easier to defend.

Reputation via the Web of Trust

Vouching, standing and the gating of higher limits all run on the existing EvoBioSys Web of Trust work rather than a reputation score invented here. A per-marketplace star rating is exactly the kind of thing that should not be rebuilt for the fifth time, and a reputation that only exists inside one loop is worth very little to the person who earned it.

Web of Trust

Open questions

  • What signal actually distinguishes clearing a surplus from depleting a wardrobe, without being invasive to collect?
  • At EUR 3.48 per pickup of up to five parcels, does a single-box collection pay for itself, or does the local station need to aggregate a street before calling the courier?
  • Who operates the local station in a place that has no vintage store, and does the loop still function without one?
  • Sellpy already pays 40 to 70 percent and operates in Austria. What does a neighbourhood loop offer a household that a national consignment service does not, beyond keeping the margin local?