From collection sheet to decision system
A collection spreadsheet is good at answering “What do I own?” It is much weaker at answering “What should I wear today?” The gap between those questions became the product brief for Scent OS.
The useful unit was not another perfume card. It was a decision: a fragrance that fits temperature, rain, humidity, location, time and social context, with enough explanation that the recommendation feels credible.
Why context matters
Fragrance behaves differently across heat, humidity, indoor air-conditioning and rainy weather. A system does not need to pretend that taste can be reduced to a perfect formula, but it can remove obviously unsuitable choices and make the remaining options easier to compare.
Scent OS therefore separates a weather-based daily pick from discovery modes. The first is practical; the second preserves play and surprise.
Building a useful taxonomy
The collection is searchable by obvious fields such as name and brand, but more interesting navigation comes from perfumers, scent worlds, concentrations and release periods. These dimensions reveal patterns that are difficult to notice when bottles are only displayed as a grid.
Making layering explainable
Random layering recommendations may look creative but are difficult to trust. The formula model uses roles—base, lift and veil—so each component has a job. The user can understand why a combination may work and can replace one role without discarding the whole formula.
The broader product lesson
Personal tools become valuable when they shorten a repeated decision while deepening understanding over time. The best result was not simply “choose a bottle faster”; it was turning the collection into a living knowledge system.