My first Product Design project

I wanted to understand the underlying structure of a good ecommerce app before mocking up any screens. Object-oriented UX was the key, here—realistically, I had to know what kind of code entities I’d be working with, regardless of focus or features.

I mapped out everything I could think of related to:

• Products
• Industries
• Product function and ensembles (e.g. shoes vs. sandals vs. boots)
• Transactions, taxes, and credit
• Delivery and logistics

I also mapped out a hierarchy of metrics for the app, connecting every ground-level metric to a high-level growth number. It was important to me that we should have an analytics system to build our product intention and measurements.

The new insight: this exercise convinced me that price tracking would work best styled as an ecommerce app—not to compete with the current players, but to:

1. Tap into already-familiar patterns (the “1,000 websites theory”).
2. To lay the foundation for excellent product data in the future.

EJ led development of the high-level app flow. He taught me the art of creating a recursive flow—a series of loops based on interactions and jobs to be done. His inspiration actually came from his love of game design, which we baked into the these concepts.

I added to that foundation by developing:

• The checkout flow
• More product page entry points (to support the user flow)
• List management and meta data for products

During this exercise, we realized that this app flow could become twice as potent when paired with subtle AI agents working in the background to make increasingly precise and relevant product recommendations that improved with every pass the user made through the “loop.”

We planned to design for a mature product state (as a practical road map) with the understanding that we’d scale it back for the minimum viable product.

We put aside plans for AI recommendations and subscription tiers to focus on creating the core price tracking experience.

This. iswhat we settled on for MVP:

1. Tracking product prices at 3 major retailers
2. Letting users track products manually.
3. Comparing current prices with historic highs and lows.
4. Notifying users when a tracked item’s price decreases by 20% or more.

I booked a working session with one of my teammates to walk through their feedback and the problems they saw with the existing design. It was a great learning experience to get critical about my own work.

We started with a teardown, and it wasn’t as bad as I feared. My design had a few specific issues—and other platforms had already solved some of them.

1. We merged the price and “should you buy?” components into one.
2. Too many product specifications too early: the SHOP app solved this already with drop-down menus. I used this pattern to combine two sections into one.
3. We changed the “more products” slider into a grid to give the user more options at a glance, minimizing interactions.
4. I added a comparison function to “Similar products” cards so that users could immediately compare what they saw to the current product they were viewing in a bottom sheet.

Just about everything I wanted in the design still remained. I (re)learned that most complexity can be drip-fed through progressive disclosure.

I’d intentionally left the home screen and search flows  until after I’d refined the product screen, the core experience.

I used Conversational Design principles to craft these functions:

1. The search bar at the top of the screen
2. Tracked items on sale
3. Your product collections (with suggestions)
4. Lates product finds (pro-active suggestions)
5. Current and upcoming store sales

The persistent bottom navigation included:

• Home
• Saved items with tabs for: Everything, Collections, and Sales
• Orders
• Profile and Settings

The search bar placement encourages immediate interaction, but I intended long-term engagement to revolve around price notifications and product suggestions.

The smart recommendations would eventually eclipse most manual searches by anticipating user interests based on in-depth information architecture and meta data. That completed the core recursive loop that we designed at a systems level way back at the beginning of the process.

Want to work together? Let's chat.

Feel free to send an email if you'd like to chat.

I love talking shop over a cup of coffee, even if you just want a second opinion on something.

Coffee's on me.