The main project of Designlab's Interaction Design course is an online grocery store app. A brick-and-mortar grocery story is experiencing shrinking sales and stiff competition from online shopping and delivery services like Instacart, so they want to expand into that space to remain competitive.
UPDATE 06/2020: This post was first published 11/2019. The project has expanded to include prototypes and some interaction design, so I have updated and "republished" this post.
The user persona was provided for me, and several people I know happened to fit it pretty well so I utilized them for research. They were given a spreadsheet list of grocery items and told to categorize them however they see fit. A few insights came out of this (as represented on the sitemap): an essentials category, a diet type category, and the user story for flexibility in categorization (ex. ravioli was categorized under "frozen" as well as "canned goods").
The persona highlighted a lack of time and interest in grocery shopping, so in building the sitemap I focused foremost on personalization (for re-purchasing) and then saving time and money (essentials, shop by ad, shop by recipes). One thing I was careful to include, based on competitive research, was to make as little of the app require sign in/registration as possible. Users can browse and build carts without an account. Only when accessing personalization or checkout features are they prompted to use an account.
Product Requirements and User Flow
I then created a list of product requirements and a user flow. The product requirements sheet was pre-organized into a few groups of user tasks. I completed it based on a previous project involving reverse engineering screens from a finished online shopping UI, the competitive research and analysis, and the user research. I found putting myself in the user's shoes and making design choices felt surprisingly intuitive, because I had done so much pre-work.
The product requirements sheet helped me to identify the core screens to be designed. I drew each screen on grid paper first, then digitized them, then iterated to a third version for the sitemap below. Doing everything on paper first felt tedious but ultimately it made things easier once I was in Figma because I wasn't starting from scratch and I already had ideas for changes to make as I digitized. Ultimately Figma was a better option than grid paper because it allowed me to get to the level of precision (or, fidelity) I desired.
Ideating on how the sketches could meet requirements is where the screens really started to take shape. As I fleshed out each screen, I updated the list of features/content for each screen and copied things to similar screens as necessary. I also did this when doing version 1 on paper: I designed the homepage first because it is often the user's first foray into the product, and the navigation I designed for the homepage ended up becoming a static global navigation. I re-used many elements over and over for ease of understanding and use.
I worked on the high fidelity prototype and interaction design more than six months after the wireframes, so a lot changed based on what I had learned.
You can view and interact with the full prototypes here: https://www.figma.com/proto/YYejV9xKEntXcenmYExC62/GoodMarket-Prototypes?node-id=0%3A1
High Fidelity Prototypes
I went for a more image-focused look based on my personal experience with online grocery shopping. I also streamlined some of the information as to not overwhelm the user with too much. A major component is the use of pop-overs for "quick actions" when possible. Very few screens are actually full-size pages, and that was a deliberate choice for the most foundational pages. A quick view of the product, a quick add to the cart, and a quick log-in did not necessitate navigating forward to entirely new pages.
I was extremely careful to design all interactions that move the user forward in the flow as pages sliding in left from the right. Inversely, all interactions that move the user backward in the flow are designed as pages sliding in right from the left. Think about flipping forward a page in a book, and then flipping backwards a page.
I missed a few important pieces in these screens, namely, the credit card information fields. I also struggled to keep things aligned and with re-use. After this project, I learned about the components feature, which would have been a huge time saver with this project. It's my goal now, after this project, to learn more about how to automate alignment/distribution.
UPDATE 06/2020: I did learn more about components and alignment tools. :) Now I seem to find myself a bit limited by prototyping interactions in Figma, namely, states of components. This requires a lot of extra artboards to illustrate basic functionality. Supposedly Figma is adding this feature in soon, but Adobe XD has it already...