


I designed a quote comparison tool that lets insurance brokers upload carrier quotes and see the differences that matter at a glance, replacing hours of spreadsheets and highlighters.
Lead Product Designer
Jan 2024 to Feb 2024
Middle-market insurance brokers review quotes from several carriers before recommending coverage to a client. Doing that by hand could take more than five hours for a single set of quotes, and every manual comparison was a chance to miss a difference that mattered to the client.
I led design from research through launch, working with product managers, subject matter experts, and front-end and back-end engineers.
I interviewed insurance brokers and observed how they managed their clients' quotes. Brokers built spreadsheets by hand or printed quotes and highlighted differences with a pen. The work was slow, error-prone, and produced a stack of paper for every client.
The insight was clear: brokers did not need more data, they needed to see which differences mattered without hunting for them.
“Streamline the insurance quote review process by making coverage differences easy to compare, so brokers spend less time reviewing and make better informed recommendations.”
Before Quote Compare, there was no place in the platform to review quotes side by side.
Quotes arrived as separate files, one per carrier, each with its own format.
Comparison happened outside the product, in spreadsheets or on printed pages.
Differences were found by reading, line by line, with nothing to call attention to what changed.
Before
After: My Redesign

I sketched early concepts and studied how comparison tools in other industries present side-by-side data. Working with product managers and subject matter experts, we prioritized ideas based on build time, the data brokers needed to compare, and system performance.
Upload and label quotes in one flow. Brokers upload carrier quotes and tag each with its policy type, so the tool knows what to compare.
Select the quotes to compare. Brokers choose which quotes sit side by side for a given policy type.
A comparison built around key data points. The view focuses on the coverage details brokers actually weigh, instead of every line in the quote.
A downloadable comparison. Brokers can export the comparison to share with their client, replacing the printed and highlighted pages.
Placing quotes side by side was not enough. Brokers still had to read every value and do the math themselves to know which quote was better on a given line.
I pushed for each data point to show its percentage increase or decrease against the other quotes, paired with a matching color. The difference, and its direction, reads at a glance, so brokers can scan a comparison and go straight to the lines that need a conversation with their client.



To make the new feature feel native to the rest of the product, I built it from our design system, including type styles, icons, buttons, fields, and calls to action. Where new components were needed, I created them within the system and updated its documentation at the end of the project cycle.


I moved from low- and mid-fidelity layouts to a high-fidelity prototype of about 22 screens covering upload, labeling, selection, and comparison. Usability testing included 15 brokers, both new and experienced. Early feedback was positive but showed the flow needed to feel more seamless, and two more design rounds refined the details.
The hardest problems were making sure the right data was pulled from each quote file and letting brokers download and share a comparison. I worked through both with front-end and back-end engineers, product managers, and subject matter experts.
After quality assurance testing, Quote Compare launched about eight weeks after the first concepts. It cut broker review time by 60% and reached 70% platform adoption.
5+ hrs
Broker review time per quote set before the tool
60%
Faster broker quote review
8 wks
From first concept to launch
I handed the final designs to front-end engineering and stayed involved through the build to make sure the tool shipped as designed.
Brokers didn't need more data, they needed the difference. Showing the percentage change instead of raw values taught me that in data-heavy tools, the most valuable design work is often removing math the user shouldn't have to do. Building from the existing design system is also what made launching in about eight weeks possible.
