How many product images does a listing actually need?

Most briefs arrive with a number already attached: “we need about twenty images.” Twenty is not a set, it is a budget. The useful question is which questions a buyer asks before paying, and how few images answer all of them.

The questions, in the order they get asked

A buyer scrolling a listing runs through the same sequence every time:

1. What is it? — recognisable in a thumbnail, at 200 pixels wide.
2. How big is it? — against something familiar, in a real room.
3. Where does it go in my life? — the context that makes it plausible.
4. What makes it worth this price? — the detail the copy is claiming.
5. What do I actually get? — contents, dimensions, what is in the box.

Five questions, and no image answers two of them well. That is where the count comes from.

The spine: seven to nine images

• Hero, plain background (1). Product filling the frame, neutral light, no props. This is the thumbnail, the ad crop and the retailer feed image. It has one job and it should not be asked to do a second.
• Scale shot (1). The product next to a human, a hand or a room element that reads instantly. Skipping this is the single most common cause of returns on furniture, pet products and anything sold by dimension.
• In-context lifestyle (2–3). The product in the environment it lives in. Vary the room across these — a kitchen, a hallway, a living room — rather than shooting the same corner three times.
• Feature close-ups (2). The sensor, the seam, the hinge, the coating. Whatever the price is defending, shot close enough that the claim is visible rather than written.
• Dimensions or contents card (1). Flat, legible, measurements in both metric and imperial if you sell in the UK and the US.

That is seven at the bottom end, nine when the product has two features worth defending.

Where sets go wrong

Repetition dressed as variety. Six lifestyle images of the same sofa at slightly different angles read as one image to a buyer. If two images could be swapped without changing what the buyer learns, one of them is not paying rent.

A hero doing two jobs. A hero on a styled surface with props looks better in isolation and performs worse everywhere it is actually used — cropped square in a feed, shrunk into a comparison grid, placed on a retailer’s white page.

No scale, ever. It is the cheapest image in the set and the one most often cut for budget.

What actually changes by channel

Channels change the framing and the count at the margins, not the spine. Marketplaces impose a white-background main image and reward a dimensions card. Your own product page can carry one extra lifestyle image and a wider crop. Paid social wants the hero re-framed square and vertical, which is a re-frame of an existing image, not a new one.

This is why we price multi-format packs separately from images: a 4:5 crop of an approved hero is a production step, not a new creative decision.

How we get there

We start from the set, not the shot list. The brief names the five questions for that specific product, each image gets assigned exactly one of them, and anything left over gets cut before a single frame is generated. Then the direction is locked — palette, materials, light, camera height — so that image seven still looks like it came from the same afternoon as image one.

Seven images that each answer something beat twenty that all answer “what is it”.