Design tools for App Store screenshots: Figma, Sketch, Affinity, Photoshop

This is the other half of the tools roundup. That post covers generators and template builders — things that make the set for you. This one is for designing it yourself.

If you have design skill or a designer, this is the route with the highest ceiling. Nothing purpose-built will beat a good designer with a real tool.

Figma

Best at: collaboration, components, variables, and the fact that everyone already has it. Community files include App Store frame templates with correct sizes, so setup is close to free.

For store listings specifically: components with a screenshot slot plus variables for captions is the pattern that works. Build it once and a release becomes swapping instances rather than editing frames.

Weakness: no concept of store output rules. Nothing stops a frame being nudged off an exact pixel size, and nothing warns you before upload.

Sketch

Best at: being Mac-native, fast with large documents, and having a mature symbol system. Its export presets are genuinely good — multiple sizes and scales per artboard, defined once.

For store listings: the export preset system maps well onto display classes, and there is a long tail of iOS-specific plugins.

Weakness: smaller community, fewer ready-made store templates, and macOS only.

Affinity Designer / Photo

Best at: one-time purchase, no subscription, and genuinely capable raster plus vector in one document — useful when a panel mixes a real screenshot with vector artwork.

For store listings: export personas handle multiple sizes cleanly. If you resent subscriptions, this is the serious answer.

Weakness: weaker collaboration, no component/variable system at Figma's level, so multiplication across languages is more manual.

Photoshop

Best at: raster work, compositing, and anything involving real photography, shadows or texture. If your art direction involves a physical scene rather than flat shapes, this is still the tool.

For store listings: smart objects for screenshot slots work well — replace the source and every panel updates.

Weakness: heavy for what is mostly a layout job, and layout is the bulk of a listing.

What all four share

No exact-size enforcement. App Store Connect compares integer dimensions. None of these four will stop you exporting a near-miss, and the error arrives hours later without naming the file. The fix is quick once you know it.

No fan-out. Panels × display classes × languages is manual in all four. Components and variables reduce it; nothing removes it.

No localized layout logic. German runs long, Japanese short, Arabic right-to-left. Each is a human decision per frame.

No opinion about composition. Which is the point — and also why so many hand-designed listings land on the same device-on-a-gradient layout. A blank canvas offers no resistance to the obvious answer.

The discipline that makes any of them work

1. One artboard per exact export size, named for its dimensions. Never scale on export.
2. Screenshots as replaceable instances — components, symbols or smart objects — never pasted pixels.
3. Captions from variables or shared text styles, so size stays consistent across the set. Inconsistent caption sizes are the clearest amateur tell there is.
4. Fixed-width caption containers with predictable wrapping, so a longer translation does not collide with the device.
5. Verify before upload: magick identify -format "%w x %h %f\n" *.png | sort | uniq -c — one row per display class means you are clean.

When to stop hand-designing

Not when the design gets hard. When the maintenance stops you improving.

The signal is specific: you know your first screenshot is weak, and the reason you have not fixed it is that fixing it means opening sixty artboards. At that point the constraint is production, and production is what a pipeline is for. Until then, a design tool and some discipline beats anything in this category.