See it, want it: how to request a product from a single photo

A photo is often all we need. Here is how photo requests work, how to take a photo we can act on, what to add, and what a good request looks like next to a vague one.

A person holding up a phone to photograph a product in a shop

"See it. Want it. We'll find it." is the shortest description of what we do, and it is literal. Many requests reach us as a single photo: a jacket on someone in a café, a sauce bottle in a friend's kitchen, a lamp in a hotel lobby. You do not need to know the brand or the name. You need a photo we can work with, and a few words about what you want. This guide covers both.

How a photo request works

Upload the photo on the shop's request page or in the composer at feelbetter.agency, add a line or two, and leave an email address. There is no account. Your request gets a reference number like FB-00012 and a status page at feelbetter.agency/r/ followed by the number.

From there, AI reads the image and your note and drafts a brief: what the item appears to be, the likely brand, the variant it can see. A concierge reads the brief, corrects it where needed, and searches. When a candidate passes our seller checks, we build a private product page on shop.feelbetter.agency and email you one link, usually within the hour. The full sequence is in inside the hour.

How to take a photo we can act on

  • The label or tag. A brand label, a care tag, a sticker on the sole, a stamp on the base. One clear shot of a label is worth five of the whole item.
  • The barcode or model number. If there is packaging, photograph the barcode and any code like a model or style number. These turn a guess into an exact match.
  • The packaging itself. For food, cosmetics and household items, the front of the box or bottle is usually enough to identify it, and the back tells us size and variant.
  • Context. Where did you see it? A shop name, a city, a screenshot of the post, a hotel. Context narrows the search when the item itself is generic.
  • Light and focus. Daylight, no flash glare on glossy packaging, text readable when you zoom in. Two or three photos from different angles are better than one blurry one.

Screenshots count. A screenshot of a post, a video frame or a product page in another language is as useful as a photo, and often more, because the text is already there.

What else to add

A photo tells us what. Your note tells us which, how many, by when and for how much. The more of these you include, the fewer questions we need to send back:

  • Variant: size, colour, edition, region or plug type.
  • Quantity: one, or a set.
  • Budget: a rough ceiling in any currency. It helps us choose between an official store and an authorised retailer.
  • Deadline: a date you need it by, or "no rush".
  • Delivery country: we check that the seller ships there, or to a forwarder we can use.

Good request, vague request

Vague Good
"This bag" with a photo of someone across the street. Photo of the bag plus a photo of its tag, with "tan, medium size, seen at a shop in Hongdae, budget around ₩300,000, need it before 20 October, deliver to Seoul".
"Sneakers like these" with a blurry side view. Photo of the shoe and a photo of the box label with the style number, "US 9, the white and grey colourway, one pair, international delivery to Tokyo".
"That snack from Japan". Photo of the front and back of the packaging with the barcode, "3 bags, the spicy one, to Busan, no rush".

The vague versions are not wrong, and we will still try. They just cost a round of questions, and the hour starts again once you reply. The good versions usually go from photo to link without us needing anything more.

FAQ

What if I only have a screenshot, not a photo?

A screenshot works just as well. Post, video frame, product page in any language: upload it the same way and add your note.

Can I send more than one photo?

Yes, and please do when it helps: the item, the label, the packaging. Different angles of the same item make identification faster.

What if you can't identify the item?

We write back with what we think it is and what would help, such as a shot of the label or where you saw it. We do not guess and send a link for the wrong thing.

Back to the journal