Tug Brice
Researcher for Hire
I am a UX researcher and conceptual designer. I've got an MS in Games for Learning from NYU and a BS in Psychology.
I founded Ice 9 Design in 2011 with several friends to work on projects and train up the skills we would need to get jobs. Since then, we have published Notes and String, set up a nationwide non-profit to help the homeless, developed some board games, and created an AI.
I'm looking for work as a UX researcher or designer.
Click the button below to set up an interview.
$O$
National scale effort to end homelessness.
Technology incubator repurposed to do good work.
Accomplishments
Helped more than a dozen people back into permanent housing.
Fed more than 100 people.
Met with policymakers from local to national level.
Prototyped an AI framework used in a later project.
Created a gamified, automated survey engine.
Problems
How do we reach our user base?
How do we effectively help people in an organization with no resources?
How do we create change both on an immediate, personal level, and a long-term, overall level?
How do we automate this process effectively?
Research Questions
How do we reach our user base?
Text Messages.
Homeless people have no fixed address and very little access to the internet.
Homelessness happens primarily in urban areas.
Urban areas have excellent mobile phone coverage.
Most people have mobile phones or know someone that does, and those that don’t can get one cheap or free in the US through the Obamaphone program.
How do we effectively help people in an organization with no resources?
Leverage research and information.
$O$ was a small group with very little money.
There are lots of programs and organizations that already offer help that have more money than we do.
It can be difficult to find information about these programs and figure out if you qualify.
We could create a survey that matched people to programs they already qualified for very inexpensively.
How do we create change both on an immediate, personal level, and a long-term, overall level?
Gather and create narratives.
A significant part of the problem with getting help for homeless people is a perception problem.
Changing people’s perceptions of homeless people makes them more likely to help them.
Stories are a very effective way to change perceptions.
Politicians love a good story.
How do we automate this process effectively?
Use AI.
Open-source tools existed that would allow us to create and modify chatbots that could talk to each other.
Coordinating these chatbots into a team creates an AI framework that could automate the process of asking questions and recording the results.
Leveraging existing academic research can guide the creation of algorithms to do this effectively.
Solutions
Reaching our audience:
Surveyed the relative availability of cellphones and text messaging among a homeless population.
Researched the Obamaphone program.
Confirmed that the vast majority of users either had a cellphone of their own or could access someone else’s.
Created envelopes designed to fold around money or gift cards to be handed out in person with the text # on it. This would build credibility as users could get immediate assistance then, and could text us for further assistance later.
Leverage research and information:
Tasked volunteers with gathering information on as many local, state, regional, and nationwide programs and charities as they could.
Collated the criteria necessary for entry and transformed it into a “master survey”.
The master survey ranked questions by how valuable they were (how many programs they were relevant to), and was designed to be fed to the automated survey engine.
Users would take the survey and then be given a list of services they qualified for based on the questions they answered.
Gathering and creating narratives:
Designed a semi-structured interview designed to gather qualitative information about a person’s experience with homelessness. How they became homeless, what their life was like before and after they became homeless, how people could help them, and what they wished people knew about homelessness.
Designed a more structured qualitative interview designed to be given to users who contacted us by SMS by volunteers before they were put through the automated survey.
Combined this information with the survey results to create a quantitive data package.
Picked several good interviews and refined them into narratives suitable for presentation to community leaders and policymakers.
Creating the AI:
There were multiple problems encountered in designing the automated survey engine algorithm, including engagement and diminishing returns.
Used existing academic research to help define the problem space and create an algorithm to manage these issues.
Created metrics to measure the performance of the automated survey engine.
Used metrics and academic research to improve the performance of the algorithm.
Takeaways
Always be ready to pivot. We shifted course 5 or 6 times during the development process but being flexible and research-oriented helped us hit the mark we wanted to hit.
You can do a lot with a little. By harnessing the power of information and narrative, we were able to spend the vast majority of our meager resources at Taco Bell, feeding people who desperately needed a good meal.
Don’t be afraid to go back to the books. The vast majority of the problems we had during the development process were solved by a single research paper. Since grad school, I’ve made it a habit of running searches on Google Scholar just to see what’s out there and that habit got us around a huge brick wall.
Good stories are incredibly effective. I took a class on narrative theory in grad school, so I had some idea of how useful stories were, but watching how well the narratives we crafted for $O$ worked was still an education.
Accomplishments
Almost 10,000 unique users.
Research-driven market identification and design.
Problems
What do we do with it?
How do we exploit it?
Research Questions
What do we do with it?
Use simplicity to drive mass-market appeal.
Notes and String was originally created because the developer used literal notecards and string in design sessions and wanted something less messy to use.
Using skeuomorphic design meant that form was function. Minimal instructions were needed because it was literally just notecards and string.
Other productivity apps were mainly aimed at designers and business markets and hadn’t found their way to the general public.
The simplicity of our design meant it could be used by anyone, making it accessible to everyone.
How do we exploit it?
Identify and understand our audience.
Because this started as a personal project, no thought had been put into a target audience.
While anyone could use it due to its simplicity, we didn’t know who would use it. Before we could successfully market it, we needed to see if it was marketable.
The flexibility of the app also meant it could be used for many different things. Understanding how people used it would be critical.
Solutions
Use simplicity to drive mass-market appeal:
Simplicity was the key design principle.
Every feature was designed to take no more than two sentences to explain its full use, and the goal was to have no instructions at all.
We focused on using familiar elements as often as possible, mimicking real-world items everywhere we could to lower the cognitive effort necessary to use the app.
This focus on simplicity and accessibility meant that anyone could use it, no matter what their level of tech skills or familiarity with design
Identify and understand our audience:
We did ad-hoc market identification by simply putting the app in front of anyone who would look at it. We had no preconceptions, so we spread the net as wide as possible
This approach showed us that the app appealed most to female college students and single moms.
Follow-up usability and qualitative research revealed that the college students used it to organize their materials for class and school life and the moms used it as a virtual refrigerator door.
Understanding these use cases provided a very clear direction for further developing the product, which we supplemented with surveys and user interviews about what features to add over the course of development.
Takeaways
When in doubt, keep it simple. Unless there’s a reason to make things complicated, simple is almost always better. Everything should be made as simple as it can be, but not simpler.
There’s no substitute for understanding why. Maybe it’s my psychology background, I’m always interested in the user’s why. When it comes to metrics, I focus on purely behavioral measures, but knowing WHAT a user does is no substitute for understanding WHY they do it.
You never know what you can do until you try. We had no idea whether Notes and String would be a success. We had some vague intuition that other people might like it, but it was a personal project hacked together by someone who wasn’t really a coder and never intended to see the light of day. But we decided to put some research into it and see what we could make of it, and it’s one of our biggest hits.