Product Photos for an Online Store (With AI)

The online buyer cannot touch the product; their "touch" is the picture. The same product, two cards: in one a dark kitchen photo, in the other a clean white background + detail shots. Even at the same price, the second sells. Visual quality is not an aesthetics matter; it is a conversion variable.

To take product photos today a studio is not a precondition: a phone + the light rules + the AI background tools give 90% of the professional result. This article is the practical protocol: the shoot, the processing, the card set; and at the end, the AI tools' proper place.

The foundation: light is everything

The one-sentence law: in good light a phone shoots better than a professional camera in bad light. The practical setup: next to a window, daylight (not direct sun; softened with a curtain), the product at 45 degrees to the window. If the shadows are harsh, a white cardboard reflector from the opposite side (a one-manat solution). If artificial light is needed, two simple softboxes or even white-shade lamps; avoid coloured/mixed light (the white balance problem). The evening shoot + ceiling lamp combination, meanwhile, is the answer to "why do my photos come out yellow."

The background and scene: two modes

• The catalogue mode (a white/neutral background): the marketplace standard; white poster paper curved into an "infinity background" (the floor-wall line invisible). The goal: the product, only the product.
• The context mode (the usage scene): the bag on a shoulder, the cup on a table; the main language of social selling. The goal: how it looks in life.

The card set combines the two; below. The prop rule: let the scene support the product, not compete with it; a second "star" pulling the gaze in the frame splits the sale.

The shooting protocol: 6 frames for the card

The phone settings: the rear camera (not the front), wipe the lens (a one-second, most-forgotten step), 1x focus (the wide angle bends the edges), a support/tripod against shake. Portrait mode with caution on products: the artificial blur can "eat" the edges.

Processing: at the natural boundary

Processing's goal is showing the product not better than it is but brightly as it is; the returns-and-disputes line runs from here. The base processing (possible on the phone too): brightness + contrast, the white balance (let the whites be white), the crop (the product as 80% of the frame). Colour accuracy is critical: in clothing, the "the colour wasn't like the picture" complaint sits in first place; rather than "beautifying" the colour in processing, capture it correctly at the shoot.

The AI tools: where gold, where trap

The gold zone; the background work: AI background removal (one tap, a clean cut) + background replacement (placing the product "in a studio", "on a marble table"); a ready feature in Canva and the specialised tools. It speeds catalogue production many times over. The grey zone; retouching: cleaning dust and small scratches is normal; hiding a defect returns with the return. The red zone; generating the product itself: a product image "shot" by AI does not match the real product; a sales photo is like a document — it is not invented. The rule: let the AI image tools change the scene, not touch the product.

Sources and further reading

Where to verify the source

The platforms' image requirements (size, background) are changeable:

• Umico Business and the card rules of the platforms you sell on

Continuing the topic

The sales visual's neighbouring articles:

• The product description that sells
• The Instagram sales storefront
• The AI image tools
• The marketplace card
• Other articles on this topic

This weekend's one-hour experiment: reshoot your best-selling product with the 6-frame protocol and refresh the card. Two weeks of statistics will be this article's most convincing continuation.