My role:
Worked with:
Overview:
Twitter brings in $400 million a year with its API via enterprise contracts and data licensing. Many of the largest business utilize streaming endpoints to gather data from Twitter, and then they use application code to parse the data. In order, to make sure that they are getting the right data from Twitter, and that their streams are working properly, they need monitoring tools. V1.1 of the Twitter API has been out for years, and the Developer Platform team recently released V2 of the API for enterprise customers and planned to deprecate V1.1 in the coming year. However, enterprise customers told their account representatives that they cannot adopt V2 of the API without core monitoring tools, which are currently only available on the GNIP (legacy) console. This project aims to provide parity for the most common and useful monitoring tools shown on GNIP, directly on the Twitter Developer Portal
Prototype
This GIF represents the design that was queued up to be built by the engineering team in Q4 of 2022. It shows the redesigned Developer Portal dashboard, clicking into a specific App that already has monitoring data loaded, and the redesigned experience for high-data streams that are partitioned. For that use case, the user would see an aggregate status of all their redundant streams and the partitions, and from there, they could drill into a specific graph to see how that specific partition was performing. The secondary graph would load in a drawer that opens from the right of the screen. The empty monitoring interface can also be seen in the screenshots below.
Prototype of the new monitoring experience in the Developer Portal
Process:
This project didn’t have a dedicated PM. There was a Principal engineer who was working as interim PM and had a lot of the core technical and product knowledge, but he was not familiar with common product management practices. So the first thing I did when this project was assigned to me in Q3 of 2022 was to start interviewing the Principal and lead full-stack engineer about how the product worked. I asked a lot of questions to understand the use cases, the shortcomings of the live product, the needs of the businesses, and to understand what success would look like. I requested access to the GNIP console, where I was able to log in and see the experience from the eyes of some of our biggest customers. I took a ton of screenshots of various scenarios and tools to reference how we were handling monitoring today.
I requested time to do interviews with customers to ensure we understood their use cases, but I was informed we did not have time to do that research before our Q1 launch. I told the team that I would design the best solution I could with the time provided, then usability test it, and in the future we could continue to interview our customers to gather data about how they’d like to use these tools and improve them moving forward.
I began with paper and pencil, sketching out some of the intricacies involved with figuring out where the monitoring information should live in our public Developer Portal. I then took the outputs of the requirements gathering, sketches, and the screenshots of the existing monitoring tools and I began to look through our design system components. I learned that our graphing components were not yet live, but was provided access to the beta of the components. The designer working on these components for the design systems team asked for feedback on the components, so as I began designing with them, I provided feedback on the graphing components to the design systems team. I completed the designs in October 2022, with the intention of usability testing them in December or January.
Outcome:
I was just finishing up the design and collaborating with the engineering leads on implementing these designs when 90% of the individuals in my organization, including myself, were laid off. I’m unsure of the future of these designs at this time. We would have measured efficacy of the tools to meet enterprise customer needs, and we also seeked to make these monitoring tools available to developers using the public API, so we would have measured page interactions, time on task, as well as CSAT.
