Karl Vochatzer (Design Lead) | Jack Alford (UX) | Jun An (UI) | Stefanie Owens (Research) | Milena Pribic (Dev)
The redesign of the public-facing IBM Power Systems website; in one section alone we went from 107 pages to 7, increasing the overall System Usability score by 52%.
Project Objective: IBM Power Systems
Our team was tasked with figuring out how to improve the usability and overall user experience of the public-facing Power Systems website (www.ibm.com/power). I conducted in-depth design research regarding the user needs of our most critical users:
Power Systems customers (AIX, IBM i and Linux system administrators)
IBM Business Partners
The site is not only a technical resource for Power system administrators to refer to, but also a vital sales reference for Business Partners and IBM Sales to refer customers of all levels of technical understanding to learn about Power Systems. Through the methods above and working closely with Sponsor Users, my team drove the new design strategy of the Power Systems website around user needs and their pain points with the existing website.
Following IBM Design methodologies, the "as-is" describes the current state of the product or site; as a starting point, our team did an overall heuristic evaluation of the 1200-page Power site to establish the most glaring issues.
We then conducted an array of discovery interviews with Power customers, salespeople and business partners in order to better grasp what each constituent was looking to accomplish with the website, as well as what their biggest issues were with the current website. One-on-one interviews were the chosen methodology for this generative research phase in order to dig in deep to the qualitative reasoning of our users.
The team used the generative data to structure our plan of attack for our proof-of-concept of the new website design. Collaborating with others in marketing and IBM Digital, we held an IBM Design Thinking workshop to internalize the issues we had heard and gain alignment on what needed to be prioritized.
The exercise above was one we conducted to help the extended team internalize the user research that had been done thus far. The quotes were color-coded based on user type (business partner, customer, or IBM sales) so that the team could see who was having the most issues where, as well as where issues overlapped across user type. I had the extended team read through quotes from discovery interviews and post them to a phase of the website usage process. I guided the team through this affinity diagramming process in order to gain alignment and articulate the pain points and focus areas, as seen below:
Overall, we employed the following research activities throughout the course of our project sprints in order to observe, reflect, and make what would best fit user needs. I selected each methodology as deemed necessary by the needs of the project at the time.
Market research interviews
Guerilla research at SXSW 2016
Customer focus group
Monthly client council calls
Monthly 1x1 interviews with Sponsor Users
System Usability Scale (SUS)
Time on task measurement
Internal user testing with IBM subject matter experts
Each of the research activities above would be a lot to dig into. As an example, I've included an excerpt from a moderated Time on Task test I did in order to establish a baseline of usability on the Power Systems website. I chose this test in order to have some quantitative measurements of time and success/failure in addition to the qualitative data produced as the research participant thought aloud. We performed this test on the existing Power Systems website (to measure the As-Is) and then later on our redesigned website (to measure the To-Be), in order to gauge a change produced by design. After the participant completed their Time on Task, I also collected System Usability Scale (SuS) data in order to have another quantitative benchmark before and after the design process.
Synthesized from all of the generative research activities above, the following principles acted as guiding measures not only for our team's prototype, but also for the future Power Systems website as it is currently being implemented by the marketing group within IBM. We reflected on these four ideas when making any strategic decisions throughout our wireframing and prototyping processes.
3. Create fewer discrete pages to enable findability, while also finding ways to break up text blocks and be more visually compelling but translation-friendly
1. Provide clear and quick navigation to technical resources (with context of what info you’d find where)
4. Cut down on redundant navigational elements to give users a clear, consistent path to relevant content
2. Separate marketing and technical content as much as possible while still giving the user a cohesive, consistent experience
In the end, our interdisciplinary design team collaborated to create a high-fidelity prototype of the AIX portion of the IBM Power Systems website, as well as a hardware comparison tool interface. Our two prototypes and their final UX/UI design are not published here for confidentiality purposes as they are not yet public.
Through these prototypes, we solved the user problems identified in the research above for each of our personas as they tried to discover, share, and compare technical information on the Power Systems website. In a final round of evaluative testing on our wireframes and prototypes, we received feedback from IBM customers that added validation to the new direction we created, as well as measured our final System Usability Scale score:
The team's work was handed off to another team within IBM to carry forward at an IBM Systems-wide level, using the design principles and patterns that our team created in redesigns across the websites for Power Systems, Z Systems and IBM Storage products. The finished design work is not shown here for confidentiality purposes, however you can view the most up-to-date Power web presence at ibm.com/power.