Redesigning a location intelligence platform for better usability
First version 2017. Second version redesigned in 2022
Embedded Files
My Role:
Design system: creator and curator
Research: getting expert reviews, conducting interviews of the stakeholders, actual users of the MVP and understanding the pain points, translating the problems into tangible, actionable stories.
Solution: coordinated with BA team to understand new features, generate solution to tackle pain points and incorporate new requirements
Management: Team management, Timelines and effort estimation, feature prioritisation
Branding: Created the enhanced logo for the platform
PROBLEM STATEMENT
PROBLEM STATEMENT
Our customers preferred task management via Excel sheets and WhatsApp, which were cumbersome and not user-friendly, especially for field agents with limited tech experience. This was due to the convoluted structure of the products, where actions were not obvious. The existing system lacked real-time insights, making it difficult for managers to assign tasks efficiently based on agent location and status.
The task was to create a robust and user-friendly foundation upon which, the product offerings can be rebuilt.
ACHIEVED IMPACT
ACHIEVED IMPACT
Improved satisfaction scores for the users from the workflow perspective.
Reduce the learning curve, potentially saving a week of effort in a month-long training.
Average 50% reduction in the number of clicks required to reach the desired information.
Reduced cognitive load by filtering the information shown, based on archetypes.
Improvements in the data input features, backed by 88% of the interviewed users
Improved aesthetics with standardization of front-end components and a design system
EXECUTION
EXECUTION
Research
Research
Included heuristic evaluation by multiple experts on a 30-point scale
Getting a leadership buy-in to execute the redesign and on the estimates.
Setting up the REMOTE research (during COVID lockdowns) to get insights from non-tech-savvy users. Read more about the research journey here
Persona
Persona
One of the earliest feedbacks was that the platform was complex.
The user distribution was 10% highly educated office bearers who needed quick views
While the remaining 90% of the users were field agents with a language barrier and needed simplicity.
With persona definition, we decided to address this by
Providing information that is needed at that point
Providing information that is relevant to the user
Overall 9 distinct personae were identified and they were further classified into SOMI system. The system comprises 4 verticals
Selling: e-Commerce, inventory
Operation: Quick response, day to day activities, dispatch
Management: people, long term adjustments, better reporting
Insights: Business intelligence, strategists, Data analysts
This was an effort to reduce the cognitive load by dividing specific functions and placing them slightly away from the day to day activities
Task flows / web
Task flows / web
For each persona mentioned a logical starting point was identified based on the interviews and then a task web was built, where the journey of the persona was mapped through the modules and submodules they have access to.
Design System:
Design System:
A design system for components was created to introduce consistency as well as enhanced accessibility. For this, various design systems like carbon by IBM, Material Design by Google, Ant Design system, and Atlassian design system were studied. A comprehensive design system with the specialized components needed by the application was created in Figma and published for the team
For the design system, the team came up with the design philosophy for the entire redesign exercise.
Accessibility is a set of requirements every designer must follow. Efforts were made to build accessibility at the component level itself. The entire system was designed to be at least WCAG2.0 AA compliant. Guidelines were also set in the description to what component to use and how.
Web UI design
Web UI design
Based on the fine-tuned design system and the concepts developed on paper, Miro was used to create the initial building blocks of the UI. These were then developed in figma with the new system in place. The designs tested for the look and feel and change in the user flows with a focus group within the interviewed users.
Task management screens were redefined.
Included insight (doubling as quick filters) that need human intervention
Each table header becomes a filter
Task/customer/asset details on an overlay not to obscure the insights
Task timeline on the right, for extra cost
Quick insights about auto allocation, job size, and predicted SLAs
Dispatch could see the status of each job along with the upcoming jobs
Simple drag and drop to assign a job
A comprehensive view of all field assets, including
Battery percentage
Signal strength
Breaks and Attendance
The cockpit view was made more usable
Quick filters on the top
Job size and SLAs
Job-status
The number of unallocated tasks was reduced to virtually zero*
Dispatch could see the SLA as well as ticket size of a job in the same map, resulting in focused push towards productivity
Managers used to caution filter most almost 100% of the times.
Report Dashboards saved time for reporting, otherwise done via excel sheets.
Testing : Web
Testing : Web
The Figma prototype was tested with the same set of users to understand if they could;
easily navigate through the portal and understand the entire structure. The time taken by the users to perform repeated tasks was measured for both old and new designs. the target was to improve the time by at least 30% for the new users and 10% for the veteran users (since veteran users use a lot of recall while using the application)
understand and recognize the programmatic enhancements made in the form of auto-selection of certain components based on persona and,
measure the usefulness of the merged dashboards and lists, through simple open-ended questions.
CLOSING THOUGHTS
CLOSING THOUGHTS
The system is under development. However, the initial feedback from the users we conducted the interviews with seemed positive. The final 🔒documentation can shed some light on the overall process.
More details can be found at dista.ai
Google Sites
Report abuse