EEZY Mobile App
EEZY is an A.I. assistant that aims to truly understand its user and offer the most relevant recommendations around food, events, and activities. My role was to create a prototype of the onboarding experience for testing purposes.
For the A.I. in EEZY to provide the most relevant recommendations, the app required a lengthy onboarding questionnaire of 14 questions, some with more than a dozen answers to choose from. The concern was that potential users would not finish the process due to being overwhelmed.
To solve for the problem of overwhelming users, I broke up the 14 questions into more bite-sized chunks and approached each of these chunks with the goal of making them fun to engage with. This would not only help the user take in information easier but gave the product more opportunities to show its “human side” and help the user feel that EEZY was really getting to know them.
A breakdown of the onboarding process helped reveal potential experience issues.
Once wireframes were completed, they were made interactive in Justinimind. Initial testing was done through the client’s team to assess which style to move forward with.
From Discovery, the strategy for onboarding became akin to gamification - making the lengthy questionnaire fun to move through.
For testing with intended users, a placeholder design style was applied to the wireframes. The visual style aimed to enhance the bright cheeriness of the A.I. character.
Rounds of wireframes were created to define the specific charm that was needed for the prototype totaling in three different styles.
A common challenge with A.I. products is building trust between the user and the machine. In the case of EEZY, the A.I. would use the information it gleaned from a user’s Facebook account to start off the process. A question that came up was how can the A.I. use this step to propel the user forward into answering questions and so that the user does not expect that the sign-up process has been completed? Two approaches were created and mapped out to figure out where to go.
From the user journey maps, giving the user the idea of a “reward” at the end of the process was tempting and seemed to ensure the most success. However, when it was paired with the rest of the experience that the app would give, beyond the onboarding, the A.I. being more up front about it needing information seemed the best choice. The A.I. needed to start off a relationship with the user on a footing of honesty and trust.
Mid-fidelity wireframes were created for the internal team in three different styles for the questionnaire. The first focused on the most direct, but rather plain approach to provide a baseline to compare the other two with. The second approach brough in imagery to liven up the answer choices. The third approach was the most extreme approach and had different interaction types for the different types of questions. The third approach was the one chosen to move forward because it provided the most enjoyable experience that broke up a potentially monotonous time.
Quick turnarounds of prototypes were created in Justinmind. This project held complex interactions within each page and Justinmind was able to handle creating that logic.
A muted turquoise tone and bright yellow A.I. character greet the users when they enter the app. Large images were chosen to entice tapping interactions and bring more natural elements to the interface. However, this is placeholder visual design and not the final design of the app.
Final Product & Insights
The final prototype was taken out to testing in multiple locations around the US and UK but unfortunately this tale ends without knowing the results of those tests. So for now I look upon this project as an exercise in feeding a large amount of information to users. I am proud of the discoveries made in the process of creating this prototypeand getting to ask questions about creating relatable A.I.