When I first joined Broker Insights I was their first dedicated user experience designer. Up until then the software engineering and testing teams had been responsible for that part of the process. I know that they won’t mind me saying, but it did mean that some of the UX had been tackled in a more pragmatic way. As broker insights were planning their next major release for their flagship platform, I was tasked with taking a look at where the most improvement could be achieved for the least development time, and reviewing their information architecture from the user’s perspective.
I started by conducting a review of their current system and working with the SMEs to understand what the system was currently doing and why that was useful to the users. The insurance sector is a large and complicated space, that has ways of working that go back centuries. It is a place that is very wary of change. Given that they very nature of our product was meant to change and modernise the sector, finding the right way to approach users was important.
Once I had a better understanding of the insurance sector as a whole, I wanted to understand the different stakeholders within it. Especially those who were the main users of the system. Again a lot of the information on these users came from the SMEs who had held these roles in the market previously, but I wanted to add more context to include users who currently held the roles and were using the BI system to help them. In conjunction with the marketing department I created a short survey to help build a better picture of our users; what motivates them, how they use the system currently, and where they find they are experiencing issues.
From the survey we discovered that our system was mostly being used by the manager-level user, to help them with placement and managing their teams. They had the most frequent logins and longest session times, whereas lower level users logged in less frequently as it was not as up-to-date as their internal systems due to the way our system updated the policy data. While the accuracy and completeness of the data was a long-term goal of the business, due to inter-business challenges it was not an overnight fix. Our focus was therefore to help the user by making the system easier to use and easier for them to access useful insight.
Using the reviews and research that we conducted, we decided to focus first of all on some of the quick wins that could make an impact for users.
Consistency
The first of these was around consistency. In the current version of the platform there was a lack of consistent placement of elements. For instance, the filters were on different parts of the page depending on which page you were on. This meant that it slowed down user actions while they searched for these elements. This was particularly evident with users who had been more recently onboarded.
Another issue that was often brought up was the lack of screen space for the data, especially for the majority of users who used quite small desktop/laptop devices.
These two issues together meant that we could make two gains with one change to the interface. We moved the filters to an expandable drawer on the side of the screen, so that the filters could be minimised when they were not in use, freeing up more screen space for the data. This however gave us a new challenge, as some users used the filters frequently, they took an extra couple of seconds to access them in the drawer. We increased the hit zone for the toggle that opened and closed the drawer which reduced that time to access the filters. This is just one example of the changes to ensure greater consistency for the users.
Big data, long waits
The next challenge was one of the main concerns with any large data-driven platform: load times. Due to the large datasets, each call to the backend could take some time to load. Part of the work around this was done by the software engineers implementing caching, but an area I identified in the UX was with the way the system handled filtering. The system would call for updated data when the user started to change any of the filters. This meant that it was making multiple calls for every filter change. I suggested that we take a different approach where the user could make the changes but the call would not be sent to the backend until the user applied the changes. This meant that one one call would be needed, reducing the load times and the amount of work the backend would need to get through. This did add an extra step for the user, but on balance it had a net positive effect of the whole experience.
Expandability
While the first two challenges I have mentioned were driven mainly by the end user’s needs, the next was driven by a business need as well. When I arrived at Broker Insights, the platform used a horizontal navigation at the top of the pages. This worked when the platform was in it’s infancy, but as the platform offering expanded there was less and less real estate for the new features.
The solution that I proposed, after researching other successful approaches to similar problems, was a vertical sidebar navigation. This allowed for the expanding offering, and for the ability to group the features into sensible sections, helping reduce the cognitive load as they tried to find what they needed. The sidebar navigation also provided a more natural and familiar way to accommodate new features by allowing vertical scrolling within the navigation section.
While this is not an exhaustive list of the challenges I tackled during my 3 years at Broker Insights, I think it gives a good picture of some of the foundational changes that were made.
I learnt many important lessons while working at Broker Insights. Here are a few of the most important learnings:
The irreplaceable value of direct user engagement
There is no substitute for communicating directly with the people who are using the system. This was something I learnt continually as the access to the user base was very limited. Making the most of every conversation, listening attentively, and asking questions provided context to some of the issues we were seeing in our more quantitive analysis. It also helped us gauge the mood of the user, not just the friction points themselves.
Alignment as the foundation for cross-functional collaboration
No man is an island, or so it is said. As the sole designer for the majority of my time at BI, building good relationships with the people you work with is key. Not only for your mental well-being, but to give you access to a wealth of diverse knowledge and insight. Also, when working with colleagues who have differing priorities (e.g., commercial, technical, or operational); understanding those priorities and how they affect our shared goals really helps smooth out communication between departments.
The importance of persistent, constructive advocacy
As mentioned above, I was the sole designer for the majority of the time at BI, and sometimes it could feel like you were fighting against the waves when you were trying to highlight an issue, or suggest alternatives. It is easy to become disillusioned and just accept not being listened to. This is especially true when you have a non-confrontational nature. Keeping the flag flying for those things that you know are important is essential. Keeping notes, building evidence, and persistence on the important things will ultimately benefit the business.
Balance experience with outcomes
In one of my examples I mentioned that an extra step for the user (generally not ideal) resulted in a net benefit in terms of load times. This is just one instance where a slightly negative impact on one stakeholder’s experience improved the overall experience for them or others. Finding the right balance is not easy, but when you find that the ideal solution for one is not the ideal solution for all; choosing to make it a less ideal experience for one, but ultimately a better fit for all, is a challenge worth pursuing.