Redesigning SAP's Help Content Publisher

RESEARCH  ·  PRODUCT REDESIGN
INTRODUCTION
The Help Portal is SAP's product documentation platform which is accessed by millions of SAP customers worldwide. The companion Publisher platform used by Authors to publish content to the Help Portal was starting to gain more users, and couldn't scale to handle new and complex use-cases. I undertook a complete redesign to provide an intuitive product experience for Authors.
ROLE
Lead UX Designer
WHAT I DID
Research, analysis, Product design, user-testing and handover.
TEAM
1 PM, 1 Design Intern, 4 Full stack Developers and 2 Testers
research
To discover the key problem areas, I ran a series of Contextual Inquiry sessions with the key users of the system - Content Authors. During these sessions I observed their main workflows and the interviews resulted in creating a clear picture of the key problems they were facing. ‍
PROBLEM
The first version of the Content Publisher had no user centred design principles applied to it. More than 60% of all tickets created in JIRA were related to usability issues. With more than 300 Products being supported and 200+ Authors relying on the system to publish content to millions of SAP customers, overcoming the usability problems through a redesign became crucial.
Breakdown of the Problem
🤷 Complex, unintuitive workflows
Publishing and authorizing  documents were the two key user workflows, but they were far from intuitive.
🥴 Heavy cognitive overload
Finding content was hard in an already overloaded interface leaving users feeling anxious and confused.
🖥 Dated user interface
The system used outdated UI patterns derived from a very dev-focused thinking with critical usability issues.
overview of old experience
process
To discover the key problem areas, I ran a series of Contextual Inquiry sessions with the Content Authors. During these sessions I observed their main workflows and the interviews resulted in creating a clear picture of the key problems they were facing.
process
Identifying main user jobs
I ran a workshop with designers, developers and the PM to present takeaways from the research and identify user jobs-to-be-done. This exercise helped every member develop empathy for users.
process
Mapping jobs with the user journey
I compiled a list of all the user jobs that came out of the exercise and mapped them back to the user journey. Using this, we identified the priority workflows and as a bonus - redundant, complex ones.
Design goals
Simple
Intuitive
Informative
process
Collaborating with Development Team on the product redesign vision
As with all of the work I do, I ensure to stay connected with the Dev team from the very beginning. At this point, I ran a workshop involving Dev Architects, PM, Front-end and Back-end developers to communicate the outcome of my research and to plan the larger product vision together. Since we also were limited to using in-house development frameworks (SAP UI5)  and the related Design System (SAP Fiori), I felt it was important to have technical feasibility conversations ahead of time.
Problem 1
Making key workflows intuitive
Publishing and authorizing were the two key tasks that authors performed, but the workflows were unintuitive. Authorizing content was also difficult due to a high possibility of errors and there was no way to undo them.
process
Exploring Document Lifecycle Management
Every document goes through multiple phases before it goes live to customers. We explored reframing this collective process of managing a document into a concept of 'Document Lifecycle' management.
outcome
A failed concept, but a winning design solution
Through user testing of these explorations, I found out that the concept of "Lifecycle management" did not land well with users but one of the design patterns we tested - a wizard model, resonated so well, that it ended up being the winning design solution.
solution
Publish Wizard and a new Authorization model
Users found the step-by-step wizard approach to publishing and authorizing tasks highly intuitive and sped up workflows instantly. We also introduced a new Authorization model, one where permissions were grouped based on roles that users played, instead of the old version based on an outdated data model.
Design goal: intuitive
Step-by-step Publish wizard
The Publish Wizard followed a step-by-step approach to publishing documents. Users could be done with the task in just 3-clicks. This also made it easy for new Authors to learn workflows quickly.
Design goal: intuitive
Review and Undo errors
The old workflows had no way to undo errors easily. I ensured that the wizard approach had a way for users to review their work and undo errors before they Published content live to customers.
Design goal: intuitive
Permissions that align with real-world use cases
A new Authorization model was introduced in this redesign, one where permissions aligned with real-world use cases and were grouped based on roles that users played - Author, Reviewer, Content Lead, Administrator.
Problem 2
Reducing Cognitive Overload
This was singlehandedly, the most often reported usability issue. Documents were presented in a flat, spreadsheet-like manner but without the capabilities to Find, Sort or Filter through them. The resulting effect was overwhelm. Users found the interface hard to consume and work with.
Design goal: simple
Show only what users need to see
The old version, displayed ALL documents present in the system, which was a major security risk. Instead, the new version showed users only the documents they had access to, only the functions their authorization level could perform. A simple solution to reducing overwhelm and preventing errors.
design goal: simple
Introducing visual elements
Heuristics revealed that Text was the only way that information was conveyed to users which created overwhelm. I introduced the simple measure of using icons and visual indicators to inform users about various elements like document status, permission levels, document visibility etc.
Problem 3
Updating the interface to be more intuitive
A key element to usability issues was the fact that the old interface was completely built without any user-centred design principles, only aimed to solve an immediate need in document management. Over time it became a giant hodge-podge of
complex features added as and when a need arose.
design goal: informative
Increasing trust by providing information as needed
Doing interviews, I observed users having a complete lack of trust in the system because there weren't enough messages or feedback to the user. Introducing better UI Error messages, progress indicators and alerts dramatically increased user trust.
Design goal: intuitive
Filter, Sort and Share views
A simple, yet underrated capability that the old interface lacked was the ability to filter and sort documents. Users also wished for a customize and share filtered view capability. These features were added to the new interface much to the delight of authors.
Validating and Shipping the Content Publisher
To ensure that usability bugs were caught ahead of time, I set up a Beta Validation program which allowed interested users to test new features early, and provide feedback. I recorded both qualitative and quantitative feedback that directly contributed to design improvements and to the product roadmap for long-term improvements.
impact
Shipped Content Publisher dashboard and 5 other completely redesigned companion dashboards to more than 50,000+ Authors.
Usability related tickets reduced from 60% to 5%.
Post-launch survey indicated an increased trust in system. Authors felt like their needs were met and felt engaged by having their feedback heard through my continuous conversations for research and beta validation program.
Things I would have improved if I had more time..
Prioritize visual design
Improved the MVP rollout process
During workshops where we decided the Product Vision, Senior Dev Architects decided to use in-house development frameworks (SAP UI5)  and the related Design System (SAP Fiori) for various technical and system architecture reasons. I felt limited by the framework chosen when designing the product. Some features could have been designed better, from an experience and visual standpoint, if we had used a different framework or if we'd gone the custom route.
With 6 dashboards requiring a complete redesign, rollout was no small feat. If I had more time, I would have laid out a more iterative rollout plan and could have improved upon my product management skills.
Let's Connect!
LinkedinXemail