Platform intent: to become the platform of choice for industrial IOT development.
  • With over 1000 developers building the platform, we had enough resources.
  • But users were struggling! And you know it's hard when engineers admit to struggling.
  • It was because the organization assumed its engineers would build for engineers. They didn't.
So in addition to focusing on how I understood the particular painpoints, this is also about how I influenced the larger organization to focus on adoption.
I synthesized a single persona: to help us collaborate on what it means for it to be drop-dead-simple to build an app, so we can become the app store of the IOT.
GE is a "six sigma" type of organization, where engineers call the shots, so I had to communicate our findings in a comically quantitative way. This is where I started to break up the experience into five key sections.
I synthesized the solutions into a succinct, high-level list of design guidelines. These were used by designers to help them think through a comprehensive solution, and to get buy in with engineers and managers about where to invest effort to increase adoption
This is the next level of detail - there are recommendations for each design principle and each phase of the experience.
I developed a storyboard that gets into more detail to take the conversation to the next level with product managers, and lead the way for junior designers.
Each of those sections had a sketch of what the experience should be like. In this case, the homepage, users said the original version felt too "marketing-y", so this sketch suggests the following features:
  • Value propositions: one for each major persona
  • Community and tutorials - devs often look for strong forums before anything else
  • Overview of platform architecture: to get her oriented
  • catalog items - don't need to be images for this user
Here's how the site looked when I was done with it. Our efforts drove tens of thousands of new developers to sign up for Predix, and made it significantly easier for them to build applications.
Later, I wanted to find out how successful our efforts had been. So I kicked off a new effort with the team of 4 researchers I was leading. Read more about this in my Medium article.