2017-03-29 Prototyping 1006.jpg

What Works Cities


Case Study:
What Works Cities


Research, Prototyping, Service Design, Visual Design, storyboard illustration



Public Policy Lab



Bloomberg Philanthropies,
Results for America


What Works Cities is an initiative of Bloomberg Philanthropies, started specifically to help cities improve services for their residents by using data – doing things like making school rankings public, sharing open data, and aggregating public safety numbers.

The Challenge

Bloomberg Philanthropies wanted to know if WWC is actually having an impact. WWC had occasionally come across success stories in their partner cities, but it wasn't happening as often as they hoped. Plus, they suspected they weren't mining the entire field; were there stories floating around that they just didn't have access to? So, WWC came to the Public Policy Lab, looking for a way to create systematized mechanisms around success-story collection. 



What We Learned

We sat down with Results for America, the on-the-ground partners for the WWC Initiative, and learned that their current system only allowed for high-touch story collection; one individual was responsible for making progress calls to every participating city. With 80 participating cities, this "manual" method made it difficult to monitor the whole WWC landscape.

Photo Mar 30, 11 14 14 AM.jpg

What We Did

We created three proof-of-concepts, each one meant to be deployed at a different stage in a city's engagement with WWC. My favorite of the bunch was something we called TickleBot – a quick, automated survey that would poke participants for success tips during the course of their regular WWC communications.

In order to stand out from government and city communications, which are usually formal and impersonal, I created a character we named "Herman." A little jellyfish avatar, he appeared throughout the survey as a personification of the warm, inviting tone we hoped to convey.

I created the TickleBot survey prototype in InVision and we took it, along with the other two proof-of-concepts, to the What Works Cities conference to test. Some of the questions we hoped to answer about TickleBot were:

  • What would incentivize someone to share their successes with WWC?
    • Most people told us they wouldn't need an incentive as long as the survey took no longer than 5-10 minutes to complete
  • How do users respond to the visual design and tone of TickleBot?
    • Reactions were positive; people liked Herman, and appreciated the clean design
  • Would participants actually have information to share with WWC?
    • Most said they would have enough material to give us an update around 4 times a year
  • How frequently would they be willing to get tickled (poked)?
    • Same as above

Next steps

After presenting our findings and design concepts to WWC, they began discussions about which concepts they will carry forward and who they will approach to design/implement the concepts. With the limited man hours they currently have, the main challenge will be finding the right hours or people to implement the concepts. Meanwhile, we've been able to apply some of our learnings from this project to our other projects. Love that cross-pollination.