Listing Image Types: Ten Groups of Shots, Explained

A shot type names the job one image does and whether words are allowed on it. A hero shot proves what the thing is. A size guide proves how big it is. A nutrition panel copies what is printed on the pack, row by row. Once you plan in jobs, a gallery stops being a pile of pictures. It becomes a short list of questions you answer in order, with the most expensive one first.

The types sit in ten groups. Plan with the group. Noticing that you have nothing at all from Trust is quicker than reading every row and guessing.

None of this is an official taxonomy. Amazon and Walmart publish file rules rather than shot-type lists, and Flipkart and Meesho publish less than that, so nobody can hand you a canonical list. The names here come from that catalogue rather than from any public page, so treat them as one tool's inventory on the date of writing. The names will not match another tool's names. The jobs behind them will.

Here is every type in the list, grouped by the job it does. Read the last two columns first. Text says what words suit the image. The final column says whether the type needs a photograph taken of that specific face, view or panel. Do not make one of those without it. Screenshot the table and plan against it.

Some types do nothing but show the product clearly, with no text, and they are the shots a thin gallery is usually missing. Hero Image is the one built for the main slot, and in Picmato the main image is directed to a plain white studio background. AI renderers do not guarantee an exact RGB value, so open the finished file and check the background with a colour picker before you upload it anywhere.

Seven of these types need a photograph of a specific face, view or printed panel: Packaging, Gift Packaging, Nutrition Facts, Consumer Study, Setup Guide, Rear View and Close-up Angles. No product text tells anyone what the box or the back of a product looks like, so a pack shot without a pack reference would be invented. Without that photograph, skip the type.

The viewpoint shots are where sellers expect a shortcut. Rear View needs a photograph shot from behind. Close-up Angles needs photographs of the product from at least two different camera positions, because ten frames shot front-on are still one angle. What sits on a face you never photographed cannot be reasoned out from the faces you did. Absence has to be stated to be honoured. Asked for the back of a product, a model fills the gap with something plausible unless it is told the reference does not show it.

In Picmato every image starts from photographs of your product. Nothing is produced from a description alone.

Whether a shot carries words follows from the job it does. Ten types in the table carry none at all, and most of a strong gallery is photographs. Words belong on the few shots whose job is to state something a picture cannot state alone. In Picmato the main image is briefed with no text, and each other image's brief decides whether it needs any.

Keep the words short. A headline of six words is readable at thumbnail size on a phone. A paragraph is not readable anywhere, and a shot that seems to need one is usually a shot doing two jobs. Split it into two images and both get read. A data panel, such as nutrition facts, should copy the rows off the product's own label rather than author anything, which is why it needs a photograph of that label.

Some types want less than their text load allows. Hero Badge carries one badge and nothing else. Ingredients works as labels with no headline. Size Guide drops the headline so the measurements get the room. Tech Specs and Warranty & Support read best as rows in a table.

Seven of these types suit a person in frame: Lifestyle, Problem vs Solution, Action Shot, On Model, Before & After, Scale Reference and Claim On Skin. A feature shot can go either way: a feature proven in use can show hands, and one proven by a close-up cannot. Keep everything else product-only, which stops a model wandering into a frame whose job is the product.

If you use a brand character in Picmato, pick one for a listing set, or up to 2 when you regenerate a single image. In a listing set it goes into every image whose brief plans a person, so how many images carry it depends on the set planned for your product, and a set can plan none. You can add one to any single image when you regenerate it. Each image with your character is rendered from its portrait, so look at every face before you upload.

If you generate a photorealistic person, Amazon has a metadata rule as of late July 2026. Images and video containing AI-generated people must carry the IPTC keyword contains-synthetic-performer in the dc:subject XMP field. Amazon says it will add a customer-facing indicator. Photographs of real people that were altered with AI tools are exempt.

No. A gallery has room for a handful of images, not the whole catalogue. Treating the catalogue as a checklist produces a long run of near-duplicates, which is worse than a short gallery. The shopper scrolls past six angles of the same face. They give up before reaching the shot that would have answered the question holding them back.

Only some types apply to every product. The rest depend on what the product is, what the listing states, or what your photographs show. On Model needs something wearable, and Setup Guide something that assembles. Nutrition Facts needs a food, drink or supplement whose panel could be read from your photo, and Claim On Skin is for cosmetics only. Many types disqualify themselves before you make a single decision.

Picmato recommends listing image templates tailored to your product, with a distinct job for each image. Pick one Hero, one Lifestyle and one Detail shot, then spend what is left on the objection that loses you the sale. Leave out any type your product data cannot support, and review anything generated before it goes live.