Christine YimDigital Visual Designer

New York

Berkley Technology Services

Usability testing for the quote flow in Berkley’s new agent portal.

  • UX case study
  • 2023—2024
  • Berkley Technology Services
Role
UX research: prototyping, note-taking and synthesis
Duration
2 months
Tools
Figma, Microsoft Suite, Miro

Berkley Beyond is a new online portal that brings policy servicing and payments into one place for brokers and policyholders. A critical part of it is the quote flow, where agents gather details from a policyholder and return an estimated premium. The flow for two lines of business had been designed by an outside agency but never tested.

As developers prepared to build it, the lead UX designer and I ran usability tests and analysed the results over two months, to decide what the next design iteration should fix first.

Deliverables

  • Clickable prototype
  • Usability tests
  • Note-taking template
  • Affinity map
  • Prioritised findings

01

Problem

The quote flow had been designed, but never tested.

The screens for General Liability and Commercial Property quotes existed, but nobody had watched a person use them. The aim was to collect first impressions, find the pain points, and rank what to improve before release.

Constraints

  1. Prototyping limits

    Figma could not reproduce every interaction the built product would have.

  2. Recruiting

    Agents were hard to reach, so subject-matter experts stood in for them.

  3. Learning curve

    I had to learn how both lines of business are quoted before I could test them.

02

Prototype

A prototype real enough to test, and no more.

  1. Learn the workflow

    I started by learning how agents work when they quote, and reviewed every screen and its data inputs with the product managers and the lead UX designer.

  2. Build with real data

    I built a clickable prototype and filled it with realistic dummy data from our database software.

  3. Audit along the way

    I noted inconsistencies between components as I went, for later revision.

  4. Prototype the components

    People were thrown by what a clickable prototype could not do, so I built component-level prototypes for inputs and single-select dropdowns. It reduced confusion and kept the file smaller.

  5. What I learned

    Match fidelity to what the test needs. I shared the approach with the UX team.

03

Testing

Four experts, one script, and every pause on record.

  1. Who we tested with

    Agents were hard to recruit, so we tested with four subject-matter experts who work closely with agents and know their workflows and frustrations.

  2. How sessions ran

    The lead UX designer and I wrote a script that followed how an agent would naturally move through a quote. Sessions ran over recorded Microsoft Teams calls; the lead designer ran them and I took notes.

  3. How I took notes

    I used an Excel template with timestamps and a coded dropdown for each note: process, interruption, quote, issue, positive finding, recommendation, error. The elapsed-time column showed where people paused or struggled.

4
subject-matter experts tested the flow
60—90
minutes per session
13
actionable findings carried into the next mockups

04

Findings

The biggest clusters marked the worst screens.

I gathered the key findings, quotes and notes from every session in Miro and tagged each note with the expert it came from. Then I grouped related findings into an affinity map, which the product managers and UX team worked through together.

Takeaways

  1. Cluster size showed priority

    Groups with the most notes showed which screens confused people most.

  2. Questions for product managers

    Mapping exposed inconsistencies between the two quote flows that needed a decision.

  3. A clear order of work

    We could point design effort at the areas with the greatest effect on usability.

05

Outcome

More data than expected, and a method the team kept.

Impact

  1. Enough to move forward

    The sessions produced more data than we expected, enough to move into the design phase.

  2. Positive feedback

    Product managers and subject-matter experts praised the depth of the research.

  3. A method shared

    I presented component-level prototyping to the UX team and recommended adopting it.

Next time

  1. Use AI for synthesis

    Explore AI tools to make the data synthesis faster.

  2. Recruit working agents

    Get direct insight from the people who will use the quote flow.

  3. Write it up

    Turn the Miro board into a structured document that is easier to read.