Find your wine, not a feature list.
WineVibes is a mobile app for people who actually enjoy wine. Not sommeliers, not learners chasing tasting-note vocabulary — the person who wants a bottle that fits the mood they're already in.
The primary user goal isn't learn about wine. It's get to the right bottle in under a minute, with confidence. Every design decision had to serve that goal, or get out of the way.
I gave myself one non-negotiable rule before I sketched a screen: the entire app had to live in three pages. That constraint wasn't aesthetic — it was protective.
Wine apps have a tendency to bloat. Every feature — social feed, virtual cellar, food pairings, vintage database, barcode scanner — sounds useful in isolation and adds friction in aggregate. Users open the app to make a decision, not to browse a catalog. Every additional feature is a small cognitive tax on the moment they came to solve.
Three pages meant I couldn't hide behind features. It also meant the user could learn the entire app in about ten seconds — which matters when you're designing for someone who came to feel empowered, not schooled.
Each cut was a bet on the primary user's goal. No social feed. No virtual cellar. No sommelier-grade tasting notes nobody finishes. No barcode scanner. No "people you might also follow." I held every proposed feature against one question: does this shorten the distance between "I don't know what to drink" and "I'm getting this one" — or does it sit between the user and the result?
The three pages became: match your vibe, review your history, adjust your preferences. That was the entire product. The rest of the work was refinement — button size, form field order, the language of the vibe prompts — because when you strip a product this bare, the small typography and micro-copy decisions do the heavy lifting.
The work was in deciding what wasn't there.
The visual identity had to feel like the wine aisle of a small, considered bottle shop — not the fluorescent supermarket version. Warm cream backdrop. Deep burgundy for the primary CTA. A soft rose accent for the match reveal. The wordmark sits in a rounded warm serif because the app itself is casual, not clinical. Nothing about the language says expert or system. Everything says friend who knows their bottles.
Fifteen seconds from I don't know what to drink to I'm getting this. Users don't notice the design when they use it — which is exactly what a goal-directed tool should look like. The user's goal is the star; the interface is the stagehand.
The three-page rule earned its keep. The next version would push the vibe input further — voice, free-text, time-of-day inference — but the constraint stays. The moment you add a fourth page, the app becomes about the app, not the wine.