Case Study
Side Hustle Matcher: Practical Recommendation Engine
Most side-hustle advice is generic and commercially weak. I built this product to solve that by turning real work signals into ranked options users can actually deliver and sell. I owned the scoring engine, explanation layer, and premium results UX from end to end.
Landing experience with a clear practical promise and a three-step explanation of the matching logic.
Problem and goal
Most side-hustle quizzes feel generic and personality-led. They produce weak suggestions that do not match what users can actually deliver in a paid setting.
The goal here was to build a more credible product engine: infer practical signals from real work patterns, then rank opportunities using explicit constraints and commercial fit.
The output needed to feel actionable, not academic. Every recommendation includes a clear fit rationale, a first offer, and concrete steps for the week ahead.
Challenge and solution
One challenge was avoiding generic role-title recommendations that feel random. I solved it by combining task signals, tool confidence, outputs, and constraints inside a transparent weighted scorer, which produced recommendations users could trust and act on immediately.
Core product decisions
Signal-first normalisation
Users are mapped from role family, weekly tasks, tool confidence, and delivered outputs. This produces stronger skill inference than title-only matching.
Transparent scoring engine
Recommendations are scored with clear weightings for skill fit, preference fit, constraints, commercial viability, confidence, and friction penalties.
Commercially useful output
Each top result includes a first offer example, starter pricing direction, outreach prompt, and three immediate action steps.
Trust through contrast
The results page also shows alternatives and one poor-fit option with reasons, which improves user trust and avoids generic hype.
What this project demonstrates
- Designing recommendation logic that is editable, traceable, and commercially useful.
- Translating research into a practical schema that supports rapid product iteration.
- Building a conversion-friendly results flow that explains ranking decisions clearly.
- Owning delivery from product framing through implementation and launch-ready polish.
Next improvements
- Add richer comparison tools between top matches.
- Support user accounts so recommendations can be saved and tracked over time.
- Introduce screenshot-backed social proof and user outcome stories.
- Add regional compliance overlays for location-specific regulation guidance.
Product walkthrough screenshots
End-to-end flow from quiz input through ranked recommendations, explanation layers, comparison tables, and launch planning.
