Make the garment the subject.
Lay it flat or hang it against a simple contrasting background, with sleeves and edges visible. Use soft, even light and check that dark fabrics still show their details. Avoid a pile of clothing, a mirror selfie or a cropped screenshot when you have the actual piece available.
Take one clear front view and one back view. Keep the original files. Filters, heavy smoothing and color changes can obscure the details you later need to compare. If a photo is blurry, retake it rather than trying to infer text that is no longer readable.
Capture five kinds of evidence.
- Whole item: show its overall shape, pockets, sleeves and closures.
- Brand and size labels: capture the text straight on and in focus. Keep the size system as printed.
- Care and product labels: photograph both sides when text is present. A product or style code can be more specific than a logo.
- Distinctive details: record the print placement, embroidery, buttons, zipper pulls or unusual seam construction.
- Condition: show stains, holes, fading, repairs and alterations separately from the identification views.
eBay’s official photo guidance also recommends a clear subject and useful detail views. These are documentation habits, not a promise that any particular app can identify every garment.
Turn the photos into a checkable search.
Transcribe legible brand and product-code text exactly. Search the code with the brand, then compare the candidate’s shape, material description, closures and print placement with your photos. If the label is missing, use descriptive terms such as garment type, color and distinctive motif—but keep the identity tentative.
Do not assign a season, collaboration or rare edition just because an image result looks close. Retailers sometimes reuse photos, and similarly named releases can differ in fabric or construction. A matching logo is evidence about appearance, not proof of authenticity.
Example: two black jackets, one important difference.
Suppose a visual search returns two black zip jackets from the same brand. One has a chest pocket and the other does not. Your full photograph shows a chest pocket, while the label contains a readable style code. Check that code against the pocketed version before accepting its name or comparing prices.
If the code instead belongs to the other version, stop and resolve the mismatch. Possible explanations include a mistaken search result, an altered garment or incomplete information. Choosing whichever candidate has the higher price would turn an identity problem into an unreliable valuation.
Keep fit and identity separate.
A size tag is part of the identity record, but it does not establish the garment’s current fit. When fit matters, measure the actual garment and record your method and units. Alterations or wear may make a tagged size an imperfect comparison. Do not estimate measurements from an angled photograph.
Use the evidence in Pieces.
Pieces photo inventory lets you start an owned-item record from a photo or product link and review the proposed details. Choose a clear whole-item photo for that step. The additional views described here are a manual reference set; this guide does not claim the app processes every view together.
Check the proposed brand, name and category before relying on market evidence. If the identity remains uncertain, do not substitute a guessed rare version. Read why AI collection estimates can be wrong for the next checks. Pieces is awaiting public iPhone launch; you can join the waitlist now.
Building a complete collection record? Start with how to track your closet’s value, then use this guide for the details.