Decide what to watch, faster.
WatchCircle is a taste-driven movie discovery product designed to help users find something worth watching in under 5 minutes — without endless scrolling.
Welcome Screen
Home Screen
Today, deciding what to watch is harder than ever.
- Too many platforms
- Endless scrolling
- Generic recommendations
- Decision fatigue
People often spend more time choosing than actually watching.
To compensate, they jump between:
- Reddit threads
- Google searches
- Friends’ suggestions
Even then, the decision is rarely easy.
WatchCircle is built around one goal:
Help users pick something worth watching — quickly and confidently.
Instead of browsing catalogs, users go through a simple guided flow and receive curated recommendations tailored to their intent.
Finding what to watch isn’t just about content — it’s about taste.
Today, users rely on external platforms to discover what people like them are watching. WatchCircle brings that layer directly into the product through Taste Circles.
Taste Circles are communities built around shared preferences — like:
- Thriller lovers
- Feel-good watchers
- Mind-bending enthusiasts
Instead of juggling between multiple platforms, users can:
- Join circles that match their taste
- Discover what’s trending within those circles
- Get recommendations shaped by similar preferences
This shifts WatchCircle from a simple recommendation tool to a taste-driven discovery experience.
WatchCircle uses a guided 5-step input system to understand user intent:
Users select what they feel like watching:
- Light & fun
- Exciting & gripping
- Deep & thought-provoking
- Emotional or romantic
- Unique or mind-bending
👉 Defines the core direction of recommendations
Users select how much time they have:
- Quick Watch (< 1h 40m)
- Standard (1h 40m – 2h 20m)
- Long Watch (> 2h 20m)
👉 This is the only strict constraint
Users choose where they want to watch:
- Netflix
- Prime Video
- Disney+ Hotstar
- JioCinema
- Any platform
👉 Used to filter availability
Users select preferred language:
- Hindi
- English
- Korean
- Others
👉 Treated as a soft preference, not a hard filter
Users choose how results should be prioritized:
- Popular right now
- Underrated gems
- Matches your vibe
👉 Influences ranking logic, not filtering
Users receive:
- 🎯 One Best Match
- 🎬 A few strong alternatives
- ✍️ Clear, human-like reasoning
Problem: Users struggle to decide what to watch due to overwhelming choices and weak recommendation systems.
Insight: The issue is not lack of content — it’s lack of decision clarity. Most systems either:
- over-filter (too restrictive), or
- under-personalize (too generic)
Solution: Design a system that:
- uses one hard constraint (time)
- treats all other inputs as ranking signals
- prioritizes decision clarity over browsing
-
Build candidate pool based on platform availability
-
Apply hard filter:
- Runtime (time constraint)
-
Apply scoring:
- Mood match (high weight)
- Picking strategy (contextual weight)
- Language preference (soft boost)
- Popularity & quality
-
Rank results
-
Return top 5 picks
- Strict filter-based logic led to empty or irrelevant results
- Treating mood as a hard filter reduced flexibility
- Language filtering caused zero-result scenarios
- Over-constraining inputs reduced recommendation quality
These failures led to shifting toward a ranking-based system with minimal hard constraints.
- Product thinking over feature building
- Strong focus on decision-making, not just discovery
- Ability to identify and fix broken systems
- Iterative design and refinement
- Balancing UX simplicity with system intelligence
- Time to decision (primary metric)
- Recommendation click-through rate
- % of users completing “Pick in 5”
- Repeat usage rate
- Lovable (no-code/low-code builder)
- Supabase (authentication & backend)
- TMDB (movie data source)
MVP is live and actively being tested.
Currently improving:
- Recommendation accuracy
- Ranking logic tuning
- Edge cases (language & platform gaps)
- UI polish and feedback loops
Note: Currently optimized for mobile experience (UI designed mobile-first; desktop experience is not yet fully optimized).
- Expanding Taste Circles into a richer community layer
- Social recommendations (friends + similar users)
- Personalization over time
- Better handling of regional content
https://golden-stage-welcome.lovable.app
Built by Nikita Malhotra Founder → Product Management | Designing products that simplify decisions
Would love to hear your thoughts — especially on:
- Whether it helps you decide faster
- How relevant the picks feel
- What feels confusing or missing