A digital concept developed for the Autodesk Data Platform (ADP). Researched and designed over ten weeks, an MVP of the concept was presented to senior leadership in the summer of 2018.
I was embedded as a PM intern on the Data Platform & Experience (DPE) team, reporting directly to a Senior PM, and working alongside an internal Data Engineer and an external consulting team. Our unit lead a project to re-conceive the suite of products on the ADP platform, I lead research and design of the digital experience of one such product.
Just as AWS has a suite of products that enable users to leverage Amazon’s cloud platform, Autodesk, too, sought to deliver a seamless cloud experience to their users.
Over 6 months in the summer of 2018, the DPSE team worked alongside a consulting group from Fjord to design this suite.
My internship coincided with the final phase of the project, so my challenge was to create a proof-of-concept for one of the products in the suite.
To develop intuition, multiple research efforts were conducted including over 20 interviews, two workshops, a community survey, and countless whiteboard diagraming sessions.
The community survey revealed distinct user groups within ADP, each of which had their own data needs, technological proficiencies, and preference for learning.
Though previous research had been conducted on ADP’s knowledge bases, I needed deeper insights into how users would actually leverage the plartform in their role. Building on insights from a community survey and a three-day workshop, I lead a content audit of existing ADP documentation and interviewed 17 users about their experience using existing ADP tools and services.
While knowledge gaps were discovered throughout the entire ADP process, it was the uncertainty at the beginning stages of engagement – when a business team investigates capabilities of the data platform – that offered the most potential. A team cannot use the platform if they don’t know what it can do.
Further, interviews with data engineers, data analysts, product owners, and team leads suggested that knowledge needs varied significantly across teams. Whereas individual contributors commented on the need for step-specific, detailed instructions, team leaders spoke of the need for high-level use cases – how to offer both in a platform experience?
"My biggest complaint about ADP is that no one knows what data is in there."
– Data Analyst
"Understanding the process and contact points would have cut the time in half."
– Data Engineer
Building on key insights uncovered during research, a How Might We Statement was created to frame the objective of the project:
How might technical process documentation communicate the value of ADP?
At this time in the project, the DPSE team and Fjord had presented the final recommentation on how to divide the ADP product suite: Analytics, Data, Core Services, and Support.
The Data Modeling & ETL product within the Core Services package was chosen as the experience to design because the ETL process touches nearly every role on every team that engages with ADP.
A solution that improved the user experience of running an ETL on ADP would serve as a template to be extended to the rest of ADP products.
Autodesk already had a Data Portal platform, but it was created for an earlier generation of ADP. As such, information was outdated and did not address the capabilities and processes of current ADP offerings.
Card sorting exercises were conducted with a Data Engineer, Data Scientist, and Product Manager to refine an intuitive categorization schema.
Defining the hierarchy of knowledge for the site allowed a distinction to be made between high-level ADP value (which appeals to Data Readers and Team Leads), and low-level product details (which appeal to Data Analysts, Engineers, and Scientists).
The conceptual knowledge hierarchy was operationalized into a comprehensive restructuring of the Data Portal’s information architecture.
Mockups of the Data Portal with the revised information architecture were created for the Learning and Development team to share with contractors who manage the site.
A user journey map synthesized research and visually represented paint points, dependencies, and areas of opportunity that exist throughout the process of using the “Data Modeling & ETL” product.
The exercise suggested that technical users required three distinct pieces of knowledge throughout their engagement with ADP:
An overview of the process required to use the various softwares of ADP
Detailed explanations of permissions, points of contact, and use case-dependent instructions
Notes of what works and what didn’t for teams that have gone before
Early prototypes were tested with users to refine the design of a boilerplate webpage for ADP products.
Users praised the high-level overview and guidance that the site would offer, but commented that many of the details required to actually run an ETL are not documented on any source.
A wiki renovation was being simultaneously conducted by the Learning and Development team to address this concern, but maintenance of the technical knowledge base was acknowledged as a critical assumption to the design’s overall success.
Comments from usability tests were incorporated into a final, high-fidelity mockup that was delivered to the Data Strategy and Experience team.
The project was interdisciplinary from the very beginning, crossing boundaries between user experience design, research, data science and business, and I realized a personal strength in synthesizing and communicating across those team and boundaries.
I discovered a tremendous passion, not only for service design, but for working in the capacity of a consultant.
As a platform that facilitates team-based big data explorations, designing for ADP required more than a list of needs and a few personas. It required an understanding of the working relationship teams have, and of how data helps facilitate that relationship. I discovered a passion for mapping processes, measuring across dimensions. and prioritizing efforts.