Over six weeks, my design team and I were tasked to produce a MVP database migration tool called Dataworks Lift that would allow IBM’s enterprise customers to move their databases from on-premise servers to the cloud. Design was challenged to bring clarity to the workflow, while still allowing for complexities of data migration.
We kicked off with a two day delivery workshop with development and offering management. Design lead the team through an exercise to define the initial golden path. We used voting with orange dots to indicate what was beyond scope for the MVP, to help provide constraints and focus to the technical discussions.
We constructed a road map to align around capabilities of the MVP release , as well as near and long term additions when maturing the product over future releases. Design also tracked questions that we needed to answer as a larger team once we left the workshop that would ensure success of our product.
We tracked the as-is scenario for users attempting large scale data migration, noticing the large amount of switching between programs and interfaces. We then constructed a to be scenario, that was technically agreed upon with development, which would minimize the number of context switches during data migration.
Out of the delivery workshop, the design team established goals that would lead our work:
- Simplify and automate (when possible) the data migration flow for a clearer, more seamless experience.
- Guide the users through the high level steps and give them necessary information to complete their task
- Create a flexible design framework to accommodate for rapid iteration and future capabilities.
- Integrate with both IBM Design Language and Cloud Data Services Design Guide
We came up with a series of sketched, low fidelity wireframes to pitch the design concept and flow to development and offering management.
Working with another UX designer, we continued to refine the sketches and screens, bringing up the fidelity on each round. We then presented to the larger team, including development, offering management and executive leadership. By showing developers screens with real content, we provoked hard technical conversations and pushed development forward to figure out the critical details.
Visual design then took our wireframes, moving from visual explorations to final deliverables. I provided support to explain the strategy and be a user advocate.
Along with offering management and software engineering, the design team and I successfully shipped the MVP of Dataworks Lift in six weeks to meet our deadline. Leadership, including the GM and a VP of Cloud Data Services at IBM were incredibly pleased and used our project as an example of a team successfully implementing agile development.
Ideally, if our team had continued on this project, I would have conducted generative research and usability testing with data engineers. This would help the team better understand how the Dataworks Lift tool was used in real world, data migration projects at IBM's enterprise clients. Conducting think aloud usability testing would also help us identify design elements which needed to be refined, fixed or removed.
Unfortunately, due to a change in business strategy, the Dataworks Lift project was moved to another group and our team was placed onto a new project. While disappointed, I was grateful to be able to get the chance to practice rapid, iterative design. The time constraints removed many of the bureaucratic hassles and required full attention from the development team, who typically juggle multiple projects. I also got the chance to develop my leadership skills, acting as the day to day lead on the project, in regular contact with our executive stakeholders, offering management and development.
- Joe Meersman
- Russell Parrish
- John Howrey
- Tim White
- Christopher Moody
Front End Development
- Jesse Crow
Front End Development
Tasks & Deliverables
- Co-facilitation of a delivery workshop
- User stories & written scenario
- Project management