MHCI Capstone with Rewyndr

SpringBook Cover resized

Every member of the MHCI program has to take part in a capstone project. At the start of the spring semester, students are grouped into teams of 4-5 and assigned a client company to work with for the next eight months. Each team is tasked with creating a design solution to a problem presented by the client. The spring months are spent conducting extensive user research into the problem space, and the summer is dedicated to an iterative design process, informed by the spring research. Both the spring and the summer semesters conclude with a presentation to the client and a printed report of work. The team is also expected to maintain at least weekly contact with their client, to make sure that expectations are the same for both parties.

I was a member of team Spark (every team chooses its own name) with four other students: Puja Agarwal, Gina Assaf, Jonathan Chan, and Michelle Lew. We worked with Rewyndr, a small tech start-up run by David Palmer and Craig Waller in Pittsburgh. Our task was to design the user experience for their photo-based web application. As part of their business strategy, they focused their product on the alumni relations market as a launch field.

After our January client kickoff meeting, we began our research process. We conducted a literature review of 15 scholarly sources, covering such topics as the perception and cognitive processing of photos, the use of photos online, and the best practices of successful online communities. We also conducted a thorough competitive analysis of 45 companies and organizations that exist in Rewyndr’s problem space. Finally, we created a territory map to visualize the team’s shared understanding of the problem space prior to any work with users. This domain research helped us choose which users to target with our subsequent field research.

We investigated three main groups of users with our field research. We made an effort to work on these groups in parallel, to maximize the amount of data we obtained.

springResearchSpread

First, we interviewed 13 “extreme” or professional photo users, such as historians, photo editors, museum curators, and others. We targeted these people because of their rich domain knowledge in the realm of photos; they work with photos as part of their jobs, which makes them photo experts.

Second, we interviewed and surveyed 79 alumni and alumni relations staff from several colleges and universities. We chose alumni as our second user demographic because of Rewyndr’s focus on the alumni market as a launch field.

Finally, we targeted everyday photo users, average people who only interact with photos casually. We designed seven special research methods to understand how people find photos, how they share them, and what they value about them. Across all of our methods, 121 everyday photo users participated in our research.

Some of these created methods included our Camera Phone Cultural Probe…

cameraPhoneSpread

… our Sensemaking Study…

sensemakingSpread

… and our Organization Study.

organizationSpread

After we completed our research, we began synthesizing our mountain of data. We used affinity diagramming to process the data from our many interviews, and we created sequence models for our Pinterest user contextual inquiries.

affinity diagram spread

The categories we created in our diagrams led us to our main research insights. We used these insights to guide our brainstorming session, in which we generated over 200 design ideas. We took the best ideas and fleshed them out, forming fifteen basic design visions. These simple visions served as conversation starters for additional brainstorming.

We presented our research insights and our visions to our client in May of 2013 in our 148 page research report. We then used the visions in the report to conduct more guided brainstorming sessions, this time with our clients present. We iterated over these ideas several times, refining them and tabling the not-so-good ones. We were finally left with three concepts, from which we created twelve storyboards. We began our summer semester by testing these storyboards with people to get feedback, via speed dating.

visioning spread

Once we had compiled our speed dating feedback, we began the first of our eight iterations of product design (which is more iterations than any other project group in our year). During each iteration, we generated design ideas (based on the earlier iteration’s feedback), refined our ideas into a new design concept, and created concept prototypes, which we then tested with people. The testing methods we used were scenario walk-through, think-aloud, card sorting, and desirability testing.

design methods spread

We dedicated our first three iterations to concept validation, where we tested different product designs to find a concept that resonated with our participants. Then, we spent our next five iterations developing the features of the concept we chose. Once our design concept was finalized, we created personas to represent the demographics of individuals that would use our product.

The prototypes we produced were broken down into three degrees: low-fidelity, medium-fidelity, and high-fidelity. Our lo-fi prototypes were made on paper, index cards, and post-it notes; they had no functionality, and served to validate concepts. Our mid-fi prototypes were made using Balsamiq Mockups and Axure RP; they had partial functionality, and allowed participants to understand interaction techniques and features. Finally, our hi-fi prototypes were created using a combination of JavaScript, HTML5, and CSS; they had near-full functionality, and let participants fully explore our product.

Once we began hi-fi prototyping, we started considering our design language. We brainstormed a list of five “design keywords” that we used to define our product’s aesthetic, and created a mood board on Pinterest based on our keywords. These two methods influenced our font choice, color palette, and iconography.

After completing our last iteration, we gave a final presentation to our client in August of 2013, where we debuted our working prototype and handed off our project assets (including suggestions for moving forward, a design specification, and some extra design ideas) in our 208 page design report. Our presentation included a concept video and an interactive demo of our prototype.

We left our mark on both of our reports in the same way:

back spread

My contribution (Spring): During our domain research, I conducted the literature review almost entirely on my own, and helped formulate our territory map. In our field research, I assisted my teammates in conducting the interviews and user studies, as well as conducting several interviews myself. I also wrote and distributed our alumni survey (using SurveyMonkey), and then tallied and interpreted its results. During our synthesis process, I helped create our affinity diagrams, and uncovered one of our five research insights (each team member had their own insight area to tackle). I also transcribed several of our interviews. For our research report, in addition to writing several sections myself, I proofread every section of the book for both mechanics and word choice; every paragraph is written, to an extent, in my style. For our spring presentation, I made the best of a bad situation by keeping the audience placated while my team solved a technical issue.

My contribution (Summer): During our visioning and speed dating process, I created the concepts for three of our storyboards, including our most successful one. I also helped test them with other members of our program. When we entered our iterative design process, I focused my design ideas in two areas: navigation and unconventionality. I made sure that any ideas we came up with were simple and easy to understand, but I also strove to create designs that were unorthodox and unheard-of. I also stressed that the user test is the best way to solve design disputes: if one is at odds with a co-worker about a particular design choice, do an A/B test, and let the users decide. In addition, I contributed prototypes for every lo- and mid-fi iteration. As the summer semester drew to a close, I spearheaded the production of our concept video, creating the script, the initial storyboards, and the narration. For our final presentation, I wrote and delivered the opening and closing speeches, which added an emotional, human element to the otherwise informational and business-like material in the presentation.

My contribution (Throughout): As I was the Documentation Lead for our project, I was often responsible for taking meeting minutes, as well as capturing the events of our research and user testing (including taking notes, and recording audio and video for later transcription). I was also the opening speaker in every one of our several presentations, including the two to our client (wherein I was also the closing speaker); my segments (thanks to my public speaking ability and knack for summary) brought the audience onto the same page, garnered a laugh or two, and left them with something to think about afterward. Finally, I made a conscious effort to bolster team morale, maintaining a positive attitude and offering reassuring words when people needed to hear them.

Tagged , , , , , , , , , , , , , , , , , , , , , , . Bookmark the permalink.

Comments are closed.