Ideal Clothes Shopping
A purchase scorer is not enough for something that outlives its first owner.
premise
Groceries are consumed. You eat them and the decision is spent. Clothes wear down slowly, and for most of that time they remain perfectly re-buyable by somebody else. A tool that only helps you buy well therefore captures maybe a fifth of the available value. The rest sits in the years after the purchase, in whether the garment keeps moving or ends up in a bag in a cupboard.
So this holon has two halves that have to work together. The first is getting the purchase right, which mostly means getting fit and colour right, because a garment that does not suit you is the one that stops being worn first. The second is making the handoff to the next owner cheap enough that it actually happens.
confirmed gap
No consumer resale platform has photo-to-measurement, and no fibre scanner has a consumer price point. Both halves were checked, and both held.
| finding | detail | basis |
|---|---|---|
| Measurements are still typed in by hand, everywhere | ThredUp, Vinted, Depop, Vestiaire Collective and eBay were all checked. Every one of them relies on the seller entering measurements manually. Depop and eBay do generate a title, a category and item specifics from a single photo, so the photo-understanding step is clearly commercially viable. Neither derives a dimension from it. This is a confirmed gap across all five, not an inference drawn from one. | A and B sources, platform documentation and trade press, September 2026 |
| Fibre identification works, and not at household prices | Near-infrared fibre identification is real and deployed at industrial scale, so the question is cost rather than feasibility. The cheapest confirmed route, the Matoha handheld, starts at GBP 1,149 of hardware plus GBP 199 a month for API access. That is a small-business tool. trinamiX, Picvisa, Fibersort and the Refashion sorting line are all enterprise, quote-only, and physically installed inside a sorting facility. Nothing in this space can be called per item by a listing app at low marginal cost today. | B and C sources, vendor pricing pages and trade press, September 2026 |
| Nobody documents automated condition grading | No platform describes autonomous defect grading in its own materials. The ThredUp 12-point inspection is explicitly human-driven in the source found, and the machine-learning layer at Vestiaire Collective flags anomalies rather than grading condition. Condition remains a human judgement expressed on a scale that does not transfer between platforms. | B and C sources, September 2026 |
Read together, those three say the same thing from different directions. The information a second-hand garment most needs to carry, its real dimensions, what it is made of and what state it is in, is exactly the information no existing system produces automatically. That is what the device and the fabric score are for, and it is why the honest framing is an open window rather than a crowded market.
Full survey with sources and confidence values on the prior art page.
type analysis
Two questions decide whether a garment gets worn: does the colour work against your face, and does the cut work with your proportions. Both have an established practitioner vocabulary and both are usually answered by a professional in a paid session.
Heuristic, not science
Seasonal colour analysis is a professional-shopper practice, not a validated instrument. A targeted literature search for this project found no peer-reviewed controlled study validating the four-season or twelve-season typing system against clothing-satisfaction outcomes. That is a reported absence of evidence rather than evidence that the practice does not work, and it is the specific reason for the label rather than a general hedge. The vocabulary is internally consistent and widely taught, there is no agreed measurement protocol, and different schools place the same person in different seasons. Cut advice is styling convention: it encodes one aesthetic goal, a balanced vertical line, which is a taste rather than a fact. Treat everything below as a starting point that a mirror can overrule.
Searched September 2026. The nearest peer-reviewed work found addresses skin-tone tools in general, not season typing.
Colour
Skin colour and hair colour carry the decision here. Eyes are a minor input, weighted accordingly.
Carries the most weight. The usual test is whether gold or silver jewellery sits better against the skin.
How light or deep the skin reads, independently of its undertone.
Natural colour rather than current dye, since the model is about contrast against the skin.
A minor input, weighted at a sixth of the decision. It mostly affects the clear or soft reading.
Colour result:Soft Autumn
Warm but low in saturation. Dusty, blended, earthy.
- #C2A87C
- #B58C74
- #8E9B72
- #7F9A95
- #C69A8C
- #6B5F4E
- Warmth
- cool warm 0.15
- Value
- light deep 0.17
- Chroma
- soft clear -0.47
Confidence 38 percent. Subtype named by the chroma axis. These inputs sit close to the warm and cool boundary, so treat this as a coin flip rather than an answer: winter fits about as well, and the two palettes are worth trying side by side in a mirror.
Cut
Proportion, not measurement. Brand sizing varies so widely that no recommendation survives a fitting room unaided.
Which reads wider, standing relaxed in front of a mirror.
Whether the waist is visibly marked or the line runs fairly straight.
Decides the trouser rise. Compare the distance from shoulder to hip against hip to floor.
Decides hem and jacket length, because standard sizing is cut for the middle band.
Cut result:Balanced with a marked waist
Shoulders and hips read level and the waist is already marked.
- Waist-marking seams and belts work with the existing line rather than imposing one.
- Wrap and princess seams follow the shape without extra structure.
- Rise is a free choice here; nothing in the proportions pushes it either way.
- Most standard hem lengths work without alteration.
Answers stay in this browser under the key prefix ideal-consumption:clothes: and are never sent anywhere.
collaboration slot: stub, by design
What you see above is a deliberately thin implementation. It reduces four inputs to three axes, picks a season and names a subtype. That is enough to show the shape of the interface and to prove the rest of the page can consume the result. It is not enough to be good.
A real type analysis needs things this project has no business building: calibrated colour capture from a phone camera under uncontrolled lighting, a drape simulation that reflects how a fabric actually behaves against a face, a practitioner-grade subtype model, and a body-proportion estimate that does not require the user to self-assess. That is a product in its own right.
So this is an open slot, not a placeholder we intend to fill. If you are building a type analysis app, this is exactly the part you would own. The interface contract is small and stable.
What the slot needs from a type analysis partner
in
- A colour result: a season, a subtype from the twelve-system, and a palette of hex swatches.
- A cut result: a silhouette family and a short list of concrete garment notes.
- A confidence value, so the rest of the interface can say honestly when the answer is ambiguous.
out
- The garment scorer filters listings by palette match rather than by a colour name.
- The circulation loop attaches a colour-type match to every resold item, so a garment finds the next person it suits instead of the next person browsing.
- The fit record carries forward, so a second-hand purchase can be filtered by measured dimensions and not by a brand size label.
The value of plugging in here is the second half: a type analysis that only helps at the point of purchase is a one-off consultation, while one wired into a circulation loop keeps paying out every time a garment changes hands.
fabric score
The score a garment carries through its whole life. It has to be computable from the garment itself, because by the third owner nobody has the label, the receipt or the brand claim any more.
The intent is an open standard held by a foundation rather than a rating owned by a company, for the same reason the grocery work insists on it: a score whose owner also sells placement in it stops being a score. That governance does not exist yet: a survey of eight standards and registries in September 2026 found no open, foundation-governed fabric quality score. Naming it as an intent is not the same as having it.
established: 1 of 5
| criterion | state | what is or is not established | sources |
|---|---|---|---|
| Fibre composition What the garment is actually made of, verified rather than declared. | known | EU textile labelling has been mandatory and harmonised since Regulation 1007/2011, so a fibre composition label is legally required and standardised across the single market. That gives a reliable starting input for any garment sold in the EU, though it covers composition only, not durability or origin. | |
| Expected wear life How many wears and washes the garment is likely to survive. | not established | No open, garment-level durability standard is in force. The EU Ecodesign for Sustainable Products Regulation entered into force in 2024 and names textiles as a priority group, with a Digital Product Passport as the intended carrier, and the acts banning the destruction of unsold apparel were adopted in February 2026. The general textiles act that would define measurable durability requirements and the passport fields themselves had not been adopted as of September 2026, and the widely cited 2027 date could not be confirmed from Commission primary text. Until it lands there is no number here that is not a guess. |
|
| Construction quality Seam type, stitch density, lining, hem allowance, whether it can be repaired. | not established | Assessed by tailors and vintage buyers with a trained eye and no published rubric. The nearest open scheme found is circularity.ID by circular.fashion, a real GPL-3.0 garment-metadata standard, but it is company-led rather than foundation-governed and it carries circularity metadata rather than a construction grade. No platform documents automated defect grading either. Marked as not established rather than approximated. | |
| Current condition Pilling, thinning, colour loss, stains, repairs already made. | not established | Second-hand platforms each use their own condition ladder, which is why a rating does not transfer between them, and none of the five checked documents automated grading. A shared open condition scale would be the single cheapest interoperability win in this whole design, and it does not exist. | none |
| Supply chain provenance Who made it, under what conditions, and how far it travelled. | not established | The EU Corporate Sustainability Due Diligence Directive was adopted in 2024 and will oblige large firms to identify adverse impacts in their chains, but it applies above high employee and turnover thresholds, phases in over several years, and produces company-level disclosure rather than a garment-level record. It will not tell you who sewed the shirt in your hand. The Open Supply Hub, formerly the Open Apparel Registry, is genuinely open data with a public API and standardised facility identifiers, and it maps facilities rather than garments, so it cannot close this by itself. |
circulation loop
The half a purchase scorer cannot reach. Every step here is a place where a garment currently stops moving.
- 00 Home
- A garment stops being worn. Today this is where it stays, for years, because every onward path costs more effort than doing nothing.
- 01 Device
- Measured, assessed, folded. The output is a parcel that already knows what is in it.
- 02 Doorstep
- The box goes outside. That is the entire effort asked of the household.
- 03 Postal pickup
- Collected on a round that already passes the door. Österreichische Post already collects parcels from addresses, so no new logistics network is needed.
- 04 Local station
- A vintage store, repair shop or neighbourhood collective. Regional rather than central, which is what makes the resilience argument work.
- 05 Listing
- Measured dimensions, fabric score, condition, and a colour-type match. Not a brand size label, which means nothing across brands or decades.
- 06 Sale
- Through whatever marketplace already has the buyers, including eBay, willhaben and Kleiderkreisel. The seller takes a cut.
- 07 Textile recovery
- Unsold after an agreed window, the garment goes to Humana or an equivalent recovery route. This is the fallback, not the default, which is the whole difference from donating blind.
A sold garment re-enters at Home under a new owner, and the record it carries goes with it. That is the part that compounds: by the third owner the measured dimensions and the fabric score are the only information anyone has, and they are worth more than the original label ever was.
access policy
The tool is offered to vintage and second-hand stores, repair businesses, and regenerative garment makers. It is not offered to fast fashion.
This is a stance, not a technical limitation, and it should be argued rather than asserted. A fit and colour scorer reduces returns and raises the share of garments that get worn instead of shelved. In a second-hand shop that is straightforwardly good: the stock already exists, and better matching means more of it stays in use.
Handed to a fast fashion retailer the same tool does something different. It lowers the friction on the volume that the model depends on, and it lends an independent-looking quality signal to garments the score itself would rate poorly. The tool would be doing marketing work for the thing it was built to argue against.
The honest counter-argument, kept here rather than dropped: a better-fitting fast fashion purchase is still worn longer than a badly fitting one, so exclusion has a real cost. The judgement is that the volume effect outweighs the fit effect, and that judgement could turn out to be wrong.
enforceability: Enforceable only weakly. An open standard cannot police who reads it, and should not try. The lever that exists is the partner listing and the circulation loop, which are granted rather than open, and the reputational cost of appearing on a list you were not admitted to.
open questions
- What is the shortest useful condition scale that a shop, a household and a marketplace would all actually use?
- Does a colour-type match raise second-hand sell-through enough to pay for the analysis, or is it a nice-to-have?
- Who holds the garment record between owners without becoming a platform that can later charge rent on it?