Source
rawraw/my-app-design-process.md
url

TL;DR: Chris Raroque’s design process is not mystical taste. It is a repeatable loop: mood board strong examples, prototype options, implement the best option on a real device with real data, then cut and refine until the UX gets simpler.

Biggest lessons

  • Mood board even simple features. Chris starts by collecting examples from Mobbin and the wider internet because even obvious features can reveal better patterns when viewed side by side.
  • Prototype alternatives before committing. Search-result cards, grouped results, and traditional lists were compared before the team chose the boring but higher-UX list.
  • Respect the data model. Grouped search results looked cleaner, but they broke relevance ordering from Typesense. UI structure cannot ignore result semantics.
  • Test with real data on a real device. Figma shows UI; the actual phone reveals whether keyboard focus, filters, result density, and scan patterns work.
  • Cut anything that does not earn its place. Icons, filters, and indicators were removed or replaced when Chris noticed he ignored them in real usage.
  • Small details finish the experience. Highlight color, divider gray, search-box height, haptics, and auto-opening the keyboard all affected the final feel.

Why it matters

  • This is the process backbone for app-product-craft.
  • It also connects to app-market-research because Mobbin-style reference research applies to product UX, not just ads and paywalls.
  • It gives vibe-coding a practical design constraint: code fast, but judge the result on a real device with real data.