Samantha Spark

Creating a Landscape of Ideas & Information

The Basics:

Client organization: Class Project, University of Michigan

My role: Researcher, Content Strategist, Product and Interaction Designer, Graphic Designer.

What I did: Researched and analyzed student needs, distilled components of user flow, conceptualized touchpoints, wireframed, prototyped, iterated, and branded a software.

Tools: Sketch, Protopie

“I wish I had known that sooner”

Individuals between the ages of 18 and 24 are at a pivotal time in life. This is a time when people have energy and independence, and the stakes of learning are becoming higher as individuals think about how they can sustain themselves financially. Finding that first job, and especially one that you enjoy and can offer you financial stability is a big step.

This is also a time period that offers immense academic and professional resources, but often people in this age group don’t find the information they need until after they need it.

I noticed this lack of awareness and direction when I taught high school, and validated it with recent college graduates. In interviews, people I talked to were able to identify and articulate this problem more clearly as they discussed it in retrospect:

• “It feels like I wasted the first couple years of college floating around aimlessly.”
• “I tried going to advising and the advisors were not helpful. It wasn’t their fault, they just didn’t know my situation. How can they figure out in 20 minutes what I should do with the rest of my life?”
• “I wish there was a way to dip your toes into multiple professional fields — you’re at such a young age and it’s hard to know what’s ahead.”
• “I tried economics and business. I was afraid of math and computer science. Now I’m kicking my younger self — Why didn’t I do that?”
• “I was worried about spending too much money. I didn’t understand that I would get more value out of a double major and another year of school. This would definitely have increased my earning potential.”
• “In general, I think the college system is flawed because you’re forced to make a decision about your future… when you’re uninformed.”

When college students get information too late, they often end up on a path they didn’t want to be on, or not on a path at all. Sometimes individuals end up struggling financially or in a career they don’t like, simply because they didn’t know the right information to get where they needed to go.

On top of that, the academic pressure of college can be overwhelming, and there can be a stigma in asking for help or acknowledging that you don’t have direction.

College graduates I talked to said that they didn’t know the questions to ask to find grounding in their search for the right path.

Currently, individuals gather information to make career-related decisions via the internet and through their connections.

Guidance counselors, mentors, teachers, professors, parents, peers, and friends all can offer valuable resources and ideas to guide an individual to the right career. I call the information we get from human connections social capital.

I learned about this problem through user interviews and qualitative data analysis. I found user goals and needs from these interviews and used those as the basis for every big and small decision I made.

Other design research methods I used:

• Card sorts
• Direct Interviews
• Email Questionnaires
• Rapid Prototyping
• Moderated Usability Testing
• Observation

Interviews were especially valuable because I was able to dig deeper when my user offered valuable insights. I learned about how individuals gathered information and made decisions about their future.

My interview outline:

Here are the goals and needs I discovered my users had.

Equipped with this new understanding, I brainstormed ways to help individuals with these needs:

I focused on user goals 4, 5, and 6. My users wanted to be able to talk and they wanted to see information about what they were talking about.

I decided to begin designing a natural language processing platform that listens to users and then gathers relevant information off of the internet to present it on a screen. The keywords of the conversation are represented as the nodes of the mental model representing the logic of the discussion.

The current name of this project is Samantha Spark. The name Samantha comes from the main character in the movie Her, and she develops the ability to understand nuanced, human-like thinking and conversational skills. The conversational design of this project is based on NLP that does not exist at this level yet. I call the mental models that go on in these screens a ‘Spark’.

In order to meet this need, my job was to combine the social capital of a productive human conversation with a tool that effectively gathers and illustrates online information relevant to the logic points of the conversation.

This tool listens to an informal mentoring conversation, collects the keywords as the nodes of a mental model representing the logic of the discussion, and scrapes the internet using these keywords, to gather relevant information around what the user would like to learn about.

I came up with features to meet each goal and need of a user, except for user goal number 3. At this point, I couldn’t figure out how to guide the user to the right questions.

Below is my first paper prototype, with features aligned to the goals and needs of my users.

When a user clicks on a node of the conversation that they would like to focus on, the clicked node expands and more circles populate around it with relevant information.

In testing the prototype, I developed the order and decision points of user behaviors and movements into a flow.

I mapped patterns that I observed in my affinity map.

Feedback from my first iteration that informed my second iteration:

• My users wanted more of a welcome/greeting/acknowledgement from the system; they wanted the system to gather information about them.
• My users told me that after they found the piece of information that they wanted, they wanted to bookmark or save it.
• My users wanted to talk to people who had been in their same situation.
• I cut out features users didn’t use

The new goals and needs and their correlating features in my second iteration are in red below.

In my next paper prototype I added introductory questions to develop rapport and gather information about the user:

I created a Save feature and buckets:

and I added circles for connections of specific individuals related to the topic of the node of conversation.

The Story of Encouragement

Listening to users walk through these prototyping conversations, I heard a common thread of fear. In different phrasing, users said they were afraid of various things- failing, spending too much money, asking questions, asking the wrong questions.

I wanted to create a feature to dispel that fear.

I decided to have the tool scrape the internet for stories of encouragement specific to the topics of the discussion, to add as nodes of the model. These stories can help a student connect with and feel inspired by another individual facing a similar challenge.

“I didn’t know what questions to ask”

Finally finding the right feature

At this point in my process I came up with the missing feature to satisfy my user need #3. I simply had the tool create nodes with “A Question you might ask” link.

Even if students don’t come away with the right information, having a better understanding of what questions to ask and variables to address can give action items and momentum. This circle populates outside and near the node of logic the students are discussing.

At this point, I rearranged my goals, needs and features within three bigger overarching themes:

• Focus
• Encouragement
• Truthful information

As I moved through the mid to high level prototypes and designs, in addition to answering to my user’s goals and needs, I asked myself these questions:

• Does this feature improve the user’s focus on their goals? and how.
• Does this feature encourage the user towards their goals? and how.
• Does this feature offer the user truthful information? and how.

Digital Prototyping

Here are some of the features I created and tested in my first digital prototype.

Establishing rapport with the user

How users can select what they would like to talk about. This creates the scope of the conversation and aggregation of information. On the next screen, circles populate and extrapolate as users speak.

The circle users click on, expands and becomes the focus of the screen and the Spark populate more nodes around the selected focus.

A higher fidelity version of this landscape might look like this:

When a student expresses five majors he or she might be interested in, the screen might populate with information looking like this screen below:

The Results

Addressing the future of employment is a herculean task. It can be so overwhelming, that young people sometimes choose to avoid it.

But as I showed Samantha to college students, instead of feeling confused and powerless, they were surprised by an empowering, appealing picture of their future. Instead of wanting to hide from the tasks ahead of them, they instead were drawn into examining the tools at their disposal to optimize their time. They were moved to action.

My hope is that as NLP develops, this tool will too, and students will no longer need to say, “I wish I had known that sooner”. Instead, students equipped with Samantha will be saying things like, “I’m so glad I found that information in time”, “This is the perfect profession”, and “I love being a millionnaire”.