Wholesale Insurance Brokerages are a highly competitive space, and the more niche insurance products (like the kind B&W offers) don't come up very often. When insurance agents look to source a solution with Burns & Wilcox they would often end up in an unfamiliar, very manual process. It can take 1-2 days up to 60-90 days depending on the scenario.

While there are still scenarios where complex insurance products require a human touch, Burns & Wilcox hoped that our team at MetaLab could come up with an elegant digital solution that could contextualize those situations.

Burns & Wilcox had already signed a contract with a development partner specializing in insurance software, but over the course of their partnership realized that their platform's user experience left a lot to be desired. Our internal assessment and subsequent unmoderated user tests of their existing platform confirmed this. B&W's dev partner would bring up other challenges, but more on that later.

We quickly realized that it was going to be a lot more nuanced project than a visual re-skin.

MetaLab conducted an on-site kickoff in their Detroit HQ and I participated remotely from Bangalore. Our main goals through this initial phase were:

1. Understanding an insurance agent's workflow
2. Analyzing their existing insurance platform
3. Getting stakeholder insights & concerns
4. Estimating the project timeline & deliverables

While we got a general understanding of B&W's goals for this project through our stakeholder interviews, we needed more quantitative and qualitative feedback on their existing platform and do a comparative analysis of the best practices in a niche such as this.

Our research team reached out to me to make a simplified prototype that captures the essence of their existing flows that would be used for user testing with our recruited test participants – insurance agents.

We took what we learned from these test results and interviews and laid out areas where we'd focus our attention on. We used what we internally call prior-frames (essentially open-ended exploratory wireframes. MetaLab's version of sketches if you will) to explore potential touch points and patterns which help set the foundation on which we would later start working on our wireframes. This also gave us an insight into the gaps in our understanding of the processes that insurance agents follow, and ask the right questions.

Through the kick-off activities, our initial user testing, research & priorframing – we had a pretty good grasp at what a day in the life of an insurance agent looks like and what they needed out of a platform that would replace their time tested workflows. The average age of an insurance agent in the US is 59 after all (McKinsey, 2016).

First and foremost, we tasked ourselves with figuring out ways to simplify the labyrinth of a platform that B&W was working with. The current flow was confusing agents. They didn’t really know what to expect going into it and therefore weren't really sure of the type of quote available on the platform. I explored different flow options and decided on one similar to the existing one while addressing the main issues: flow, messaging & efficiency.

I was responsible for 2 of the 3 main products that B&W planned to launch with – Personal Article Floater & Cyber Liability. Basically insurance for your unconventional physical assets that a regular homeowners insurance wouldn't cover and for protecting your businesses from cyber security threats like hacking and data leaks.

An agent would typically talk with their clients to figure out what kind of insurance they're likely to find useful. They would then recommend one that fits their needs and if the client gives them the green light, that's the point when their actual work began. They would then gather information from the client to understand if they're eligible for it (basically making sure they engage in insurable business), then note down the clients personal details followed by details about the  risk (i.e. assets/digital records etc. they're looking to get insured). This is the part of the process that's extremely manual and somewhat prone to human error on account of the sheer number of data points. After all this is done, the agent would get in touch with a broker like B&W and submit this information to an underwriter to get a quote. If the client found the quote acceptable, they'd proceed with issuing the policy. But, if at this point, the client found the quote too high, or had some other changes in their risk, the whole process would need to be repeated. It was rife with time (and hence, revenue) consuming inefficiencies and potential for error.

Keeping in mind the circumstances under which an insurance agent would most likely be using this platform (i.e. while on the phone with a potential lead), and after much exploration – we arrived at a visually sparse step-by-step flow, with a focus on clear call to actions, contextual explainers/warnings wherever applicable and ease of navigation between sections for rapid data changes.

After a few rounds of client feedback on the wireframes, and numerous collaborative iterations of some of the common parts of the flows across these insurance products we had some approved end-to-end flows.

At this point, we were 5 sprints into the project. Due to the nature of this project, in parallel to the wireframing, we were also exploring high fidelity directions that could help set us up with the design components that we'd be working with for the remaining 6 sprints. This was a little unusual for me, since it felt like a stark deviation from the processes I'd been used to pre-MetaLab. In an agency environment that caters to businesses & timelines of all kinds, you kinda learn to go with the flow.

Since our wireframes were still in the making when we started exploring these design directions, we chose to apply our hifi thinking on some of the more generic aspects of the app. This enabled us to remain unblocked while still establishing some guidelines that could be translated fairly easily to the wires when the time came. While I initially had inhibitions about how these 'integrated' sprints would turn out, they went quite smoothly. Not to say that they didn't have their own set of challenges, but we dealt with them as and when they came up.

For the remainder of the project, we worked exclusively in high fidelity. But in parallel to the continued iteration on the insurance quote flows, we continued the iteration on the hifi direction.

Another oft under-thought aspect of web apps that we paid attention to was the dashboard. Here, the goals were to provide an easily parsable page for agents to get a birds-eye view of their workload and quick jumping points into the various insurance quote flows. I explored & iterated on some layout options, and variations of the components used on the dashboard along with optimizing the work items.

3 weeks from the end of the project, we ran our second round of moderated usability testing. We enlisted 6 participants – 3 internal and 3 external; specifically testing the Onboarding, Insurance Quote, and Issue Policy flows.

The testers rated overall usefulness (4.5/5) and ease of use (4.75/5).

As is the case with a lot of projects at MetaLab, our B&W deliverables would be used by a non-MetaLab development team. So we take extra care to continually maintain a documented component library with small & medium sized projects like these. The larger projects, depending on the nature of the project, can necessitate time investment into more comprehensive design systems. All of this is not just a hand-off plus, but a requirement when you're remotely working with a team of designers spread across timezones. I collaborated on the UI Kit with my teammates, and also looked into the responsive (primarily iPad focussed) layout updates.

Here's what Burns & Wilcox's agent platform looked like when all was said and done:

As of January 2020, the V1 of the platform was live with 150 agents in Burns & Wilcox's national partnership network. They received overwhelmingly positive feedback on the platform, specifically the ease of onboarding & the overall ease of use. They were slated to go public later in the year to coincide with their fall marketing push.

Collaborating with 3rd part development teams can sometimes be hit & miss. B&W was working with a development partner that had their own platform for building these insurance product flow and working within the constraints of the types of components this platform supported was challenging to say the least. Still, I would say that we tried to use these limitations to simplify down to the essentials and I’m pretty happy with the results.
‍

Designing for niche use cases like the speciality insurance industry meant that we had to learn how their current suite of insurance products work but also how their future offerings could work. This forced us to think in a very modular way, breaking down these flows into logical groups that B&W could use as a framework for future work.
‍

We knew early on that we’d be working in parallel on high fidelity as well as wireframes for a portion of this project on account of the tight schedule. We learned the importance of developing re-usable patterns and a comprehensive UI Kit early in the project, which left fewer net new curveballs later on.

Adapting to non-conventional workflows was probably the biggest take-away for me on this project. Each project since has had its own set of workflows and while a lot of it is standardized and methodical there are unique challenges associated with each project and B&W was when I got my first introduction to it.
‍

This project being my first time collaborating with an external development team, I learned a ton about things like documenting my design work, presenting it and establishing a shared understanding of the product experience.
‍

Knowing whom we are designing for is a foundational step in any design process. When it came to B&W, it helped us challenge our assumptions about accessibility and usability and our work benefitted from it greatly.