Pick one aspect ratio and never break it
Shopify themes crop product images to fit the grid, and they do it without asking. If half your catalogue is square and half is portrait, the theme will either letterbox some products or crop into them, and both look like a mistake. Decide on a single ratio before you generate anything — square is the safest default because it survives most themes, most collection layouts and every social surface you will later repurpose it for. Portrait ratios photograph apparel better but need a theme that expects them. The decision is cheap to make once and expensive to revisit across two hundred products.
Variants need images or the picker does nothing
Shopify can swap the product image when a shopper selects a variant, and stores that skip this leave a real conversion improvement unclaimed. A buyer choosing between four colourways wants to see the fourth colourway, not read its name. Generating one image per variant is tedious with a camera and trivial when the colour is a setting rather than a reshoot. The same applies to size-adjacent variants where the product genuinely looks different — a folded versus hung garment, a small versus large print area. Where the variant does not change appearance, do not manufacture a difference; a duplicate image is worse than none.
Image weight is a storefront speed problem
Product imagery is usually the largest thing on a Shopify page, and page speed affects both conversion and how Google assesses the store. Two habits keep this under control. Export at the resolution the theme actually needs rather than the largest you can produce — around 2048 pixels square is enough for zoom on most themes, and a 6000-pixel file is bandwidth nobody sees. And prefer modern formats where your theme supports them; Shopify serves WebP automatically in most cases, but starting from an unnecessarily heavy source still costs you. The gain compounds across a collection page showing twenty-four products at once.
Where a mockup stops being enough
Generated imagery covers product shots, variant coverage and social crops well. It covers two things badly. The first is the hero banner, where a store's identity is set and something bespoke usually earns its cost. The second is any product whose appeal is material — texture, weight, the way a fabric falls in real light. A shopper buying a heavyweight garment specifically because it is heavyweight wants evidence, and a render is weak evidence. Use mockups to make the catalogue coherent and fast, and spend real photography budget on the two or three products actually carrying the store.
