Project Brief
Member of a cross-functional design team tasked with reducing customer frustration and enabling a quicker turnaround when working with large data sets.
My Role
Design lead, user interview facilitation
Problem Statement
Customers with policies that contain large risk schedules need to get them into their Policy management systems, and will often need to update them as well. The current product experience makes these imports and updates extremely time-consuming, and sometimes impact business opportunities that require a quick turnaround.
Analysis of Current Pain Points
To kick off the project the team created a Lean UX Canvas. This established a shared understanding of the different problems that needed to be solved, how we would measure success, and set priorities for what needed to be done first.
We also produced an as-is workflow map to provide a baseline of understanding for the current product experience and create empathy for our users and their daily struggles.
The insights gained through these activities were then re-framed as opportunities. The team collectively asked how might we:
- Improve the speed of initial data entry
- Improve system responsiveness (including visibility of what the customer needs to focus on)
- Enable quick bulk changes to data sets
Improving Initial Data Entry
Our primary personas Arnold the Assistant Underwriter and Caterina the Customer Service Rep were the ones doing the initial data collection and entry into the Policy system. The data clean-up and formatting before they could import the data was very labor intensive and time consuming, and unfortunately they would run into business validation errors in the Policy system that would require further clean-up and lead to re-importing multiple times.
It was clear we needed empower our personas to resolve these errors efficiently, and reduce the need for re-importing their data multiple times. This would enable them move quotes through the underwriting process as quickly as possible and not lose out on vital business opportunities.
I produced some initial workflow wireframes that represented an ideal workflow and I knew could technically be built using the Duck Creek Design System. My goal was to follow Nielsen's Usability Heuristic #1 and provide clear "road signs" for the user to find errors and resolve them quickly.
We also introduced a Right Toolbar within the page to keep track of business validation errors. Previously, users would only see the error message on the last step of the workflow and then have to go to a different page to resolve it, adding to frustration as they try to remember the contents of the error message.
Enabling Bulk Data Updates
The second issue users were running into was related to making updates to their very large data sets. These data sets represented annual policies that would renew and have updates every year, but unfortunately due to the lack of excel-style table features in the Policy system they would often just re-import the data from a spreadsheet year over year.
Our goal was to introduce some more powerful table features so that users could filter by specific criteria and apply bulk data updates based on those criteria.
Our initial designs explored introducing the following table features that have come to be standard for most web users today:
- The ability to customize table columns (add, remove, reorder)
- Excel-style column filtering that was dynamic and searchable
- The ability to apply a bulk update based on specific filter criteria (e.g. Update the collision deductible for all Trucks under the policy)
User Feedback
We conducted two focus groups with representative persona users, which provided valuable insight and validated the approach we were taking.
After incorporating feedback from the focus group, I produced some high-fidelity design prototypes that simulated our core workflow improvements.
Improving Data Entry
The below screen capture provides a walk-through of my design prototype, introducing the Work Item Assistant (Right Toolbar) component that followed the user as they went through the quote workflow, keeping the business validation errors at hand and providing quick links to jump straight to resolving the error.
Column Customization
Another feature we wanted to introduce was the ability to customize the columns being displayed in tables. We knew that our users would need to make updates to these policies over time, but those updates will vary based on the company and user. By introducing more Excel style pivot table features like column customization, we ensured that users could get the right layout for their context of use:
Column Filtering
Another crucial feature for making data updates to large data sets was the ability to filter based on specific columns. This feature didn't exist in our table component yet, so this was our opportunity to design and validate the feature with users:
Bulk Updates
Finally, we wanted to provide a powerful feature that would really speed our users up; the ability to apply a data update to an entire data set, or, only those records that met the current filter criteria. After a few revisions this is what we came up with:
What we heard in user testing
Resolving Rating Errors
"Not only did you give the option to navigate in the tabs at the top but the options on the right to go to the error. The more you can flag the problem, the better!"
"The rating errors on the side give me a to-do list, which I appreciate"
"I think fixing the rating errors was pretty user friendly and smooth"
Column filtering / Bulk Data Updates
"Very interesting, there are some very good things that I would have never though possible, but that would be very valuable"
"I like that filter tool! I'm a Sr. Underwriter and I look at data differently than the data entry person, so this will be handy to help me find what I am looking for easily!"
"Doing bulk edits like this seems to make it really simple to get all changes done when you have a huge schedule to work on. It seems really fast!"
Conclusion
After overwhelmingly positive feedback from users in our usability tests, it was clear to the internal product stakeholders that these features would have a significant impact on our users daily work. The final designs were translated into interaction specifications and taken up over several sprints by the product platform team. Today, all of the above features are available to use in the entire Duck Creek product suite.
