Context:
Our team decided on this project in order to resolve uncertainty when it comes to choosing events to attend and how we can make sure that users find the right event for them. Having an app based on people's experiences allows users to see first hand whether one would be worth their time and money before making a decision.
Description of team members/roles
Website Editor, Interviewer. Helped initial conversion of lo-fi to hi-fi prototyping.
Interviewed testing participants, contributed to ideation, assisted in prototyping the hi-fidelity solutions
Website editor, Video Editor, helped draft hi-fi prototypes and collaborated with rest of team to identify key features and narrow down problem statement.
Website Editor, helped draft lo-fi prototypes and obtained sketches from interviews
Website Editor, brought together ideas and helped make them cohesive on the website
Problem Statement
Problem background :
The majority of our participants mentioned having difficulty deciding which live events are available in their area. All participants had trouble on deciding on which are worth going to (and are worth investing their time and money).
Additionally in 2023, a report by Eventbrite found that many consumers feel face decision fatigue and feel exhausted by the volume of available live events, struggling to decide what is worth their time and money. (https://www.eventbrite.com/blog/press/newsroom/eventbrite-trends-report-2023-predictions)
People affected by the problem :
This not only affects potential attendees of events, but also event organizers, artists, performers, venues, and ticketing platforms who would benefit from knowing an accurate headcount of an event. Friends or family members making shared plans and deciding which performances to go to may also have difficulty because they are unsure where to look for testimonials for these experiences.
Impact of the problem :
People may miss out on meaningful experiences, experience decision fatigue, or feel regret for having attended something they feel wasn’t worthwhile. Event stakeholders face inconsistent metrics to gauge attendance, lost potential revenue, and reduced engagement.
How Might We?
Final outcome
This is the final prototype we decided to flesh out!
It combines elements both from our most likely to work and delightful prototypes. The event radar compiles events near the user and lists the average ratings given by other users on the platform. Our solution makes users confidently decide on what event to attend based on other users' experiences.
Original research question : How do people decide what live events to attend?
Needfinding Research Summary :
We had a total of 5 participants, all of whom are undergraduates aged 18-22 in various colleges (UCB, DVC, etc.). The design research method we used was 1-on-1 interviewing to understand their process of finding live events and how they develop their preferences. The three key insights we gained from their answers were that:
Interviewees typically attend concerts if their friends and family are going,
Interviewees often look online for reviews on shows or events they wish to attend,
Interviewees always enter a 3rd party app to check the location of the event and their schedule.
These three insights shaped our "How might we...?" question by bringing to light the many changes that could make people's lives easier when trying to discover an event or show, and the features that we wanted to include in our app (i.e. being able to connect with your friends and see what they attended, or integrating a ranking/review system inside the app).
Ideating Summary :
We started our brainstorming through putting down the most common insights from our user interviews and creating mental personas. We then placed all of these ideas on a FigJam, voting with the emojis with the most likely/delightful/dark horse ideas, creating a heatmap. Finally, we then tallied the votes and circled which ideas we believed would be most effective for our HMW question.
Key
GREEN - Most Likely
BLUE - Most Delightful
RED - Dark Horse
Summary of lo-fi prototype process
Taking in the ideas we brainstormed, we voted and picked our top 3 as seen, the "most likely", "most delightful", & "dark horse". The first prototype asked the question of which nearby performances are highly rated and what the environment is like there - answered through the many reviews that are gathered in the app. The second prototype asked the question of how to find out which of your friends are attending a show and if your schedule is empty at those times - answered with the friends tab and schedule tab. The last prototype asked the question of how to quickly find and choose current performances, what they are, where they are, and if they are relatable to your taste - answered by showing you events in a tinder-style swipe right/left environment.
The key takeaways from our lo-fi prototypes were that some of these features may be easier to implement and more useful to the user, and others may seem fun but may not be utilized as much. Since there were many features we wanted our app to have, we also realized that we may need to reorganize the layout to fit the components, and prioritize certain features over others.
Most Likely to Work
Most Delightful
Dark Horse
Summary of hi-fi prototype process
Moving on to the hi-fi prototype versions of our three solutions, we focussed on creating clean UI's that brought to life the lo-fi "most likely", "most delightful", & "dark horse" options.
The “Most Likely” prototype explores how organizational layout and geographic tools impacts the ease of finding an event. Through the use of multiple tabs and filters, it prioritizes efficiency and structured navigation. The “Most Delightful” prototype examines how the presence of friends influences a user’s decision to attend an event, emphasizing social incentives and engagement with its Snackpass-esque interface. Finally, the “Dark Horse” prototype investigates how far intuition can guide a user within a swipe-based interface, offering a more exploratory and unconventional approach to event discovery based on popular dating apps.
From the hi-fi design prototyping process, we were able to takeaway that users are drawn to familiar looking interfaces that feel simple. When a space appears too cluttered in high detail, users are overwhelmed.
Most Likely
Most Delightful
Dark Horse
Summary of usability testing process
Our participant group featured the same demographic of people from our last interview which was undergraduate university students aged 18-21. The design research method we used is short in-depth interviews to understand our users' needs. We also conducted usability tests through our hi fidelity prototypes to understand their thought processes. Key insights included (1) how our event radar prototype was too clunky, (2) how the dark horse prototype was unclear in its action to swipe left or right, and (3) how the aesthetics influenced which prototype our users preferred. The recommendations we received were to emphasize the user experience of our prototypes by utilizing our user's previous experiences with apps like Instagram or Reddit, chancing elements from each prototype that were most intuitive to understand to create our final design. Another recommendation we received was to give simple instructions or add a help button to guide the users when they need assistance. The last recommendation was to keep the aesthetics simple and minimalistic to maintain user focus when navigating the application.
Reflections
We learned that what is straightforward to us might not be for our users - our idea of common sense was not universal. Everything should be displayed in a simple manner so that users aren't confused by how our app functions or how they can navigate it.
Layout is important as it can make or break our app. The design should be aesthetically pleasing but not too clunky otherwise it's too complicated for the user to utilize and will cause decision paralysis.
Focusing on how the user can rely on previous experiences showed us how similarity & resemblance is important to users. Participants reported that some of our prototypes resembled the tasks they do daily such as scrolling on Instagram or writing a review.
Lessons Learned
One thing that could be done differently is to conduct more interviews, especially between the lo-fi and hi-fi design processes. We took more input from potential users to hear more thoughts about our ideas and what they'd like to see implemented. This could give us a better idea on what adjustments may positively impact the UI/UX of our product.
We also want to have a more thorough and comprehensive review of our design process through an ethical lens before we started designing. Understanding the possible impacts and changes to social relationships with our product earlier would allow us to build a prototype that is more human centered.