SEP 2015 - NOV 2017 | Senior Product Designer
Learning Momentum is very challenging for new government accountants and analyst. This has overloaded senior associates who spend 30% of their day fixing errors for junior team members. Help is often not contextual and finding a document requires mastery of advanced search within the system.
Results. Leading a team of product managers and developers, we redesigned a transactional web app, identified and prioritized a roadmap of improvements, optimized common workflows, organized page layouts and content groupings, documented many new features for future sprints, created a design system, a custom icon library and visual design language that aligned with the brand goals while setting up a framework for continued product innovation. Then we plugged it all up to be a part of the build process gaining the attention of the executive team and become a blueprint for a company wide design pattern library innitiative.
The Momentum redesign project was a code modernization effort to update an aging, bloated web application into a modular MVC framework that could serve other development initiatives company wide. This was also an opportunity to address challenges in adoption, retention and overall usability.
Executive management's position was that all they needed to do was give it a fresh paint job and that would make the underlying problems disappear. We would find learning the product, finding documents in a large database, and addressing error handling would expose a host of architectural, workflow and inconsistencies. We would also discover that these usability and workflow issues put an undue burden on senior associates. We also found that most organizations using the software would wait 2 or 3 updates before they would even consider the upgrade because of additional training and system performance issues of the past.
This was also the first time the product team had even consider the user experience, relying on developers to craft workflows, define taxonomies and come up with a visual design while developing a complex transactional application.
Not all projects start from a happy place of discovery, ideation and iterative testing. As Designers we always have to use our experience, powers of persuasion and storytelling to help negotiate for a user centered design process so we can do more than just put some lipstick on the pig and actually gather some intelligence to help us form our design decisions. This was the perfect challenge for me to not only integrate repeatable processes to understand, explore and materialize in an Agile environment but to also position UX as a way to help them identify, define and prioritize improvements to the system.
Every contract is different. Some organizations embrace a user focused development process while others develop first and ask questions later. Along with discovery activities I would need to champion a user centered approach while I untangling years of adding features and workarounds while trying to get any insights on existing users. Agile was also new to the team so it was the perfect opportunity to layer UCD and Lean UX practices into the new development process.
Starting with some older research done in combination with a State and Local version of the software I needed communicate the need for more relevant information and create a strategy to get enough data to start to make decisions supported based on user feedback. I would need to present the strategy for discovery activities while demonstrating the benefits to the approach to get buy-in from the executive, product and development teams.
Then we would enter the ideation phase to come up with a number of potential solutions to make the system easier to learn, more collaborative, efficient and scalable while maintaining the existing functionality.
Belongs in Results
Leading a team of product managers and developers, we redesigned a transactional web app, identified and prioritized a roadmap of improvements, optimized common workflows, organized page layouts and content groupings, documented many new features for future sprints, created a design system, a custom icon library and a visual design language that aligned with the brand goals while setting up a framework for continued product innovation. We plugged it all up to be a part of the build process, garnering the attention of the executive team and establishing a blueprint for a company-wide design pattern library initiative.
Stakeholder Interviews. Like most new projects it's important to talk to the stakeholders to find out what the goals are, what they've tried and how we would measure success when we achieved it. Most everyone knew there were a number of opportunities for improvement and they were also acutely aware of the adoption and retention issues facing the product. The truth was that the product was so bloated it caused performance issues and had created a big backlog of bugs in features their users relied on.
UCD Planning. To understand the problems Momentum users faced, the research that had been done, the business goals and to assess the current application to identify problem areas, I reviewed an enormous amount of information, talked to stakeholders and got lost in the vastness of the application trying to figure it out. What I found shocked me.
Product Evaluation. I'm not an accountant and with a 3 to 6 month learning curve there was no way I would even know where to start. I did click around a bit and got immediately nauseas and also thrilled because the opportunities that would lie before me. In order to evaluate the product I would need to rely on customer service logs and the product teams to help me understand workflows and common usage patterns. I was also blessed with reading through old installation documentation to understand the complexities of the build process and all the things that could go wrong for customers upgrading.
Where I got my lucky break was discovering a number of SMEs who came from agencies where they used the product day to day. As you might guess they were some of my first interviews and would give me a great picture of the challenges facing our users.
User Interviews. Interviewing external users was a bit challenging and contextual inquiries, field studies and observations would not happen inside agencies dealing with top secret documents. We had to rely again on internal resources and quarterly user group meetings to capture the problems, pain-points and frustrations. As time passed we would also find an opportunity to talk to more people during an annual event for both customers and potential buyers.
A bonus was the opportunity to use interns as a representative sample of a new Momentum user perspective. The interns actually gave us some great insights and helped us explore a number of ideas to make the product easier to use.
Empathy Mapping. Building empathy for users wasn't hard. They were faced with dense screens of data using a 9pt font in a series of random layouts as they would move through a task. Using empathy mapping allowed us to share those frustrations with product and development teams and help inform our ideation activities and become a tool we would use to validate our ideas to be certain we were achieving the desired results.
Personas. Leveraging the older contextual interviews and observations from the field studies and the insights gained from internal resources the team was able to start defining some composite personas. There were a number of meetings to gain a deeper understanding of the motivating factors, demographics and roles Again these would become part of the record as we would explored some solutions.
Definition. In order to communicate the improvements and negotiate improvements to workflows the product team would hold daily backlog grooming and user story creation to address improvements and new features. At this point we had enough qualitative data to identify 20% of the problems that would the biggest impact to the usability of the system. We would also collaborate with the lead developers to access the difficulties in addressing some of the bigger problems like searching and error handling.
Ideation. This was also another challenging area of the engagement. I believe everyone knew the product was hard to use and they were faced with loosing market share but no one had an idea how to fix it. In order to make this part of the journey we needed to pick the top 3 problems and focus on them individually while trying to create a more holistic experience.
So we locked ourselves in a conference room with a whiteboard and started to explore the learning curve with a requirement to not change core functionality or taxonomies, which we would discover evolved randomly as user requests came in.
Our sessions would ultimately lead to a more contextual help system and an onboarding system to help new users navigate the complexity of the software to address the learning curve. It would also prove to relieve some of the burden on more experienced analyst.
We also start to reimagine a tasked based workflow to help folks keep up with a growing list of projects and quickly identify the priorities in their workloads. And while I believe that we could have simplified the interface considerably we got a lot of feedback from users they liked the dense displays with lots of options. Our solution would try to account for that by providing 3 levels of granularity to a given task.
To address the challenges with findability the team would collaborate again to improve search and more importantly the way users could filter the results in a more intuitive manner to find documents indexed by a randomly generated number at the time of creation.
In retrospect I believe the ideation process would have been better by running individual design sprints but tight deadlines and the technical requirements resulted in a number of features that we just couldn't change.
Another initiative was a code modernization effort where we would explore better workflows, more consistent taxonomies and improved layouts in a modular system designed to align development teams with a user first approach.
Wirefames. The next step in the process, which was actually in parallel with some of the discovery, was creating wireframes to start to visualize some of the solutions we explored during ideation. I don't remember how it was that we started working on advanced search but that was the first challenge we would address. One of the challenges we faced was the executive teams expectation of high fidelity wireframes so a lot of the earlier low fidelity work would quickly get overwritten as we would push our way through the sprints in order to give development actionable documentation to build the templates.
These wireframes were created in illustrator and pushed out to InVision so we had a clickable prototype early on.
The feedback we would get would ultimately lead us to solution that was hidden from basic search, had the ability to save the query and return a set of facetted results using the filtering capabilities of the SQL database.
Prototypes. The goal was to create a high-fidelity interactive prototype to achieve executive buy-in, validate our decisions, communicate design and workflow changes to development teams, and promote the improvements to a user base resistant to change. It's always fun to see a project like this take form. Starting out from sketches on whiteboards and copy paper, the Axure prototype quickly took on Visual Design and complex workflows to inform a set of blueprints accessible through Jira and Confluence. The prototypes would quickly become the tool we would use to validate our concepts with both internal and external stakeholders.
Twenty or more prototypes were created ranging from low- to high-fidelity that would serve three development teams over the release lifecycle. We used those prototypes to conceptualize workflows, interactions, layouts, typography, iconography, color and visual design.
Ultimately, the prototype would walk us through complex interactions supported by workflow documents, define sprint-related features, and establish a format for comprehensive user stories and Confluence articles linking you to the feature.
Our core prototype would become a template for two additional initiatives underway in parallel with the redesign project. It would also become a library source for any new projects providing interactions, branded UI and layout templates.
Visual Design. The primary objective was a redesign so there was a strong visual element to the project. From a customized set of icons, logo mark, branding guidelines, a 508 compliant color palette, and the fonts, we created a unique visual language that would help communicate the business vision and a consistent experience for the next generation of the app.
Design System. Early on, working the code modernization team we came up with a way to integrate a design system into the build process for both the legacy product and the new framework. The comprehensive system would provide developers with a cookbook of design patterns, accessible HTML, CSS/Sass and react components for the modularization effort. The design system was built to easily scale allow developers to change colors, fonts and iconography in less than 10 minutes.
Leadership. Throughout the project I was in a leadership role providing the vision and guidance for the project. I was also responsible for managing UX resources and mentoring the product team faced with a new development process.
Testing. It was difficult to do any heavy testing due to the complexity of the workflows and the customization features. Although we were able to get users to click through a given task in our "Speed Testing" event at one of the Momentum Days, we would never be able to formally test a prototype externally. The challenge of the learning curve and the extensibility feature of the real app, allowed the individual agencies to customize the way the software worked for their specific use case resulted in many being reluctant to adopt an upgrade without a lengthy staging phase. This meant they were likely working on two or three versions prior to the last release, and left us relying on in-house resources for feedback and contextual inquiries to identify problems with older versions.
The tests conducted were think out loud observations. After the testing the team would gather the feedback and trying to fix what we could to stay on course.
The primary team delivered a major release on schedule that comprised an extensive redesign including a workflow-based dashboard, significant improvements to onboarding and user collaboration across documents. We were able to negotiate small improvements to existing workflows, advanced search and faceted results that would have massive reach. Artifacts of the exercise were included in a UX roadmap that would provide a prioritized list of improvements for the next three major releases, untangling workflows, improving layouts and addressing user feedback as part of an ongoing continuous innovation practice we established.
Additionally, we designed and built a unique design system using React and 508 compliant CSS that could be used in the build process to inform look and feel for non-React elements providing the code base for development efforts using the new framework. Securing adoption on a corporate level, the system was a scalable, opening it up to easy 'look and feel' changes that fit into the branding guidelines of other CGI projects.
And finally, we implemented a process for continued user feedback that will be used to inform future sprints. The coolest part of the process was sharing our efforts with large groups of users and future customers during three-day semi-annual Momentum events. Where we introduced 'Speed Testing' to capitalize on the audience size and availability in order to test easy workflows and obtain valuable feedback from a customer base that isn't all that thrilled about change.
