Ask three indie developers how they make store screenshots and you will get three religions: the automation camp (Fastlane), the design camp (Figma), and the generator camp (web tools). They are not actually competing: they solve different layers of the same pipeline. Here is an honest map of who does what, and the combination that works.
Layer 1: capturing raw screenshots
Fastlane snapshot (iOS) and screengrab (Android) automate the capture step: they drive your app through UI tests and produce raw screenshots for every device size and every locale, reproducibly. If your app has 8 languages and localized content, this is the only sane way to get 8 sets of authentic, in-language captures. What Fastlane does not do: marketing. Its output is naked screenshots, and naked screenshots convert poorly and read as amateur.
If you have two languages and five screens, manual capture in the simulator is honestly fine.
Layer 2: composing the marketing slides
This is where captions, backgrounds, and device framing happen, and where the two other camps live:
- Figma (or Sketch, or Canva) gives you unlimited design freedom. The cost is discipline: every store size and every language is an artboard you manage by hand, nothing warns you when a dimension is off, and text changes across 8 languages mean 8 manual edits. Great for a one-language app with a designer in the loop; increasingly painful as the language x device matrix grows.
- A dedicated web generator trades freedom for workflow. LookFrame is built around the matrix problem: one shared design preset, captions per language, screenshots per device and language with inheritance, export of every store size in one click. Its honest limits, as always: no video previews, browser storage instead of a real folder outside Chrome and Edge, and less freedom than Figma for wild one-off compositions.
The combination that actually works
For a localized app, the pragmatic pipeline is: Fastlane captures, a generator composes. Fastlane produces authentic in-language raw screenshots per device; you import them into the composer, write captions once per language, and export the full Apple + Google set. Figma stays in the loop for brand exploration, one-off launch visuals, and the feature graphic if you want something bespoke.
For a single-language app: skip Fastlane, capture manually, compose in whichever layer-2 tool you enjoy. The tool comparison goes deeper on the options.
The short version
Fastlane is a capture robot, Figma is a freeform studio, a web generator is a workflow machine. Pick per layer, not per religion: automate capture when languages multiply, and compose in the tool whose constraints match your matrix.
