Amazon Fresh In-Store Mode
How can in-store grocery shopping be even better?
I joined the Amazon Fresh in-store team in spring of 2020 to help answer this question. The team's charge was to create a tool to help reinvent physical-store shopping and to build a platform where advertisers can reach these shoppers. I was the sole product designer working on this project as I collaborated with product, engineering, research, and other cross-functional teams.
Problems and explorations
Inital research indicated that people enjoy grocery shopping for the most part. However, qualitative data indicated there were definitely pain points, especially related to saving money and finding deals. We learned the primary reason people opened mobile apps while at the grocery store was to check for deals, sales, and coupons, so, while we expolored numberous features, functions, and presentations, we decided to make deals our initial main focus.
Early, low-fidelity explorations. I wireframed early ideas to help the team visualize features and determine our areas of focus.
Mid-fidelity screens. Because Fresh is a newer brand, I moved relatively quickly to higher fidelity UI explorations to help see what a product manifestation of the Fresh brand might look like. I ended up not pursuing the color wash backgrounds shown here because, although they were part of the Fresh brand system, they felt too heavy for an app meant to supplement physical store shopping. I kept the larger font sizes and tap targets used in this exploration with the goal of optimizing for glanceable, single-handed use while shopping.
High-fidelity mocks. These mocks represent the direction we ultimately went with: reduced colors to still match the Fresh brand while letting the content take center stage, large fonts, buttons, and tap targets, and primary and secondary navigation in a menu near the bottom of the screen to support easy, one-handed browsing.
While identifying features and iterating on designs, I was conducting unmoderated usability studies to verify my design direction. I designed and conducted six different studies related to deals, navigation, content, and the name of the product.
One of the key findings was related to onboarding: after testing several different patterns, it became clear that customers needed a detailed explanation of what in-store mode was and when to use it. Although this added friction, it was neccessary friction: without a full explanation, first-time testers all gave different descriptions of in-store mode's purpose; after I added the dedicated onboarding screen, all participants understood the feature and gave the same description of what it was and when to use it.
In-store mode launched as a soft launch in July of 2021 and is still live today (if there is a Fresh store in your area, go to Grocery > Fresh > In-Store in the Amazon app to see it).
The feature was averaging about 500 organic, unique visits per day at the time of launch.
We decided to maintain an unannounced launch due to learning from an in-store usability study I conducted with research prior. In this study, we heard that customers liked the feature but found it too hard to use without search, filters, and full catalog capabilities. These were all features that we'd descoped for our launch, so we decided to wait to formally announce the feature until after these features were added.