Scentfine.

Context

Scentfine brings fragrance discovery and resale into one consumer marketplace. Its product scope connects finding a fragrance, understanding its value, and completing a purchase across web and mobile.

Customer problem

Selling fragrance online is high-friction: sellers have to create trustworthy listings, describe condition accurately, and price bottles without clear market data, while buyers need enough evidence to feel confident purchasing from a stranger. The product serves fragrance buyers, sellers, and collectors. Buyers need useful discovery and confidence in a purchase; sellers need a clear listing and fulfillment flow; collectors need ways to understand their collections.

What I owned

Product Manager, June 2025–August 2026, Houston. Owned end-to-end product across iOS, Android, and web: seller onboarding, listings, discovery, checkout, Stripe Connect payments, shipping, community, and trust & safety.

Strategy & prioritization

Redesigned seller onboarding to separate listing creation from payment setup. The change increased listing completion by 34%. A detailed account of the research and prioritization process is still to come.

Constraints & options considered

Capture the decisions behind recommendation relevance versus explainability, listing speed versus moderation, and price guidance versus seller control. Include the alternatives considered and the actual reasons for choosing a direction.

Key decisions & why

The three decisions that shaped the core buying and selling experience:

1. 01

   Listing · Start from the catalog, not a blank form

   Sellers pick their fragrance from a 40,000+ catalog, so name, brand, and notes come filled in and the listing is accurate from the start. Selling becomes a few choices instead of a form to write.

2. 02

   Trust · Make condition verifiable

   Required photos, fill level, condition, and packaging are captured as structured fields, so buyers can judge a secondhand bottle before they pay.

3. 03

   Pricing · Guide the price, don’t set it

   A suggested price weighted by market data, fill level, condition, packaging, and availability gives sellers a reference point, while they stay free to list at any price.

Execution

Launched personalized AI recommendations across a catalog of 40,000+ fragrances. Built AI message moderation and a Price Guide that aggregates and weights pricing data from five sources across 40,000+ fragrances. The marketplace also connects Stripe Connect payments, shipping, community, and trust & safety.

Outcome & learning

Increased listing completion by 34% by separating listing creation from payment setup. Automated 72% of messages previously requiring manual review. Delivered personalized recommendations and market-value guidance across 40,000+ fragrances, with pricing aggregated from five sources.

Challenges & lessons

Add a specific example from fragrance identity, condition assessment, pricing coverage, or moderation. Explain the failure mode, the response, and what changed in the product—not just the technical fix.

What I’d change

A proposed next evaluation would trace the journey from discovery to purchase, identify where buyer confidence breaks down, and assess recommendation usefulness before expanding automation. This is a proposal, not a completed experiment.