top of page

Building a Multi-Pathed Insurance Quoting Software

Updated: Nov 2, 2023


Black and white photo of two women looking at a computer screen in an open floorplan office.

The Problem

Clearlink’s insurance sales team used a legacy software system that had been patched and propped up for almost a decade. Call times were long, software failures and glitches were high, and callers were frequently frustrated by the effects of the old system. Sales reps went through extensive training, then learned workarounds from veterans on the floor, and attrition rates were high, especially early in the training periods.


The Goal

Create a new order entry system that reduced call times, made training easier, and dramatically lowered system errors. A secondary goal was that this product might be white labeled and sold to mom-and-pop insurance firms.


Constraints

For this project, the only constraint was know-how. The development team had a new agile coach, but were learning agile principles as they went. This was my first time working on the insurance backend, and the learning curve was steep.


Research

We first looked at other market competitive softwares and tried to parse out how they worked. I spent the majority of my time on the sales floor, watching over reps’ shoulders on multiple calls a day, cataloguing all the details I needed more information on. I also spent a lot of time with our developers, trying to understand all the API calls, possibilities, and constraints -- we were pulling quotes from a third party system, and it required a very specific number of details in order to pull accurate quotes. As pieces began to come together, I started a rough wireframe of the new system.


Validate & Iterate

At first, I spent a lot of time running back and forth to the sales floor with usability tests on small design features or developed prototypes. But the more I kept observing their workflows and interviewing them about the current software, I had a realization: an insurance sales rep’s success depended on their ability to follow the caller’s conversation.


If we asked reps to take down information in a specific order (like the old system), they’d either end up frustrating their callers or creating workarounds (like one rep who literally wrote all the details down on a dry erase board, asked the caller to wait while she entered it in the correct order, and then read out the quotes).


The Solution

I ended up building a custom home screen with tabs that broke insurance out into its respective information buckets (policy holder, home, auto, leisure, health, other…). Reps could click to the tab the caller lead with and were given clear visual cues for any field in the flow that needed to be satisfied before they could request quotes. The flow automatically adjusted requirements as soon as state information was collected and updated any flags for the rep to follow up on before quoting.


By introducing multiple valid pathways through the flow and creating a highly visual error system, we allowed sales reps to follow the customer's lead and build natural rapport rather than forcing the conversation through a rigid funnel.


For this project, we took great care to work with the design team on the error handling and notifications. Symbols plus labels and carefully tested contrast to account for colorblindness made the system accessible to every sales rep.


The Results

A 72 year old lady joined our insurance team because she had been through Medicare sign ups and wanted to help other folks her age get the care they needed more easily. She was in the first training class we had on the new system, and she said, “I feel like I don’t even need you to show me how this works! Just get me on the phones!”


Call times were dramatically reduced, reps were happy with the ease of the new training programs, and callers sounded pleasantly surprised to get quotes as quickly as they did. We saw less than a handful of major glitches in the first month, and other sales teams were quickly jealous of the “pretty” new order entry system insurance had.



Photo from Arlington Research on Unsplash


Recent Posts

See All
bottom of page