top of page
bxwith-pc.png

Aetna Benefit Express

Business to Business Application Design Case Study

The Application

Benefit Express is a business-to-business application used by health insurance brokers to generate quick rate quotes from Aetna. Originally it was a Coventry Health Care application, but was rebranded upon Aetna's acquisition of that company.  To provide accurate results, brokers must input extensive customer data into the system, after which they can review eligible products and plans. The platform also allows secure communication with underwriters and Aetna representatives.

While the tool is essential to the broker workflow, the current application is limited to desktop use and suffers from significant usability challenges. Brokers report that data entry is cumbersome, navigation is unintuitive, and the overall experience slows down their ability to serve clients efficiently. These pain points present a clear opportunity to modernize Benefit Express and bring it in line with current UX and design standards.

B to B vs B to C Design

In B2B, I’m designing for people doing repeatable work.  As a result,  I optimize for speed, clarity, and error prevention in complex workflows. That means role-based permissions, robust filters, bulk actions, and clear states, with success measured by time-on-task, accuracy, and reduced support load. In C2C, the core design challenge is trust between two individuals—so I focus on discovery, confidence signals like ratings and verification, safe communication, and frictionless checkout or handoff. Success there is conversion, retention, and healthy marketplace activity.

  • Primary user goal

    • B2B: Productivity & accuracy (finish tasks fast, avoid errors, meet policy/compliance).

    • B2C: Confidence & ease (find the right match, feel safe, complete the transaction).

  • Users & context​

    • B2B: Known users, repeated use, often trained, working in constrained environments.

    • B2C: Mixed experience, emotional decisions, often first-time or occasional use.

  • Complexity

    • B2B: High complexity is normal—roles, permissions, workflows, edge cases, integrations.

    • B2C: Complexity must be hidden—progressive disclosure, guided flows, simple mental models.

  • UI patterns​

    • B2B: Dense screens, tables, filters, bulk actions, audit trails, keyboard efficiency.

    • B2C: Browse/search, profiles/listings, messaging, reviews, clear CTAs, mobile-first flows.

  • ​Success metrics

    • B2B: Time-on-task, error rate, throughput, adoption across teams, support ticket reduction.

    • B2C: Conversion, trust signals engagement, completion rate, retention, marketplace liquidity.

The Application

Benefit Express is a business-to-business application used by health insurance brokers to generate quick rate quotes from Aetna. Originally it was a Coventry Health Care application, but was rebranded upon Aetna's acquisition of that company.  To provide accurate results, brokers must input extensive customer data into the system, after which they can review eligible products and plans. The platform also allows secure communication with underwriters and Aetna representatives.

While the tool is essential to the broker workflow, the current application is limited to desktop use and suffers from significant usability challenges. Brokers report that data entry is cumbersome, navigation is unintuitive, and the overall experience slows down their ability to serve clients efficiently. These pain points present a clear opportunity to modernize Benefit Express and bring it in line with current UX and design standards.

Working with Product Managers/Owners

I partner with product owners as a thought partner, not an order taker, so we can design the right thing, faster, and with less rework.

  • Start with outcomes, not features: I align on the problem, success metrics/KPIs, constraints (timeline, compliance, tech), and who the primary users are.

  • Clarify scope & assumptions: I help break big epics into shippable slices, call out risks, and document assumptions so we can test them.

  • Make the work visible early: I share quick sketches/low-fi flows first to confirm direction, then move into high-fi when we’re aligned—so we don’t waste cycles.

  • Use research to limit risk: I bring in existing insights, propose lightweight validation (5-user tests, quick surveys, analytics checks), and translate findings into design and prioritization.

  • Collaborate in the backlog: I help write/reshape user stories, acceptance criteria, and edge cases, especially around complex workflows, tables, and error states.

  • Communicate tradeoffs clearly: When priorities shift, I explain UX impacts (speed vs. clarity, flexibility vs. simplicity) and recommend the best option given the goal.

  • Stay close through build: I do iterative design reviews with engineers/PM, answer questions fast, ensure design system consistency, and validate the final experience before release.

  • Measure & iterate: After launch, I review metrics and feedback with the PO and propose improvements for the next sprint.

​

Working Together on Project

Project Goals

  • Business goals

    • Bring an existing application in-house (It was currently in the hands of a vendor).

    • Replace the existing application code with new Java code.

    • Create a series of standard screen templates.

    • Create a CSS library to ensure ease of updates.

    • Replace custom JavaScript code with a JavaScript library.

  • User goals

    • Identify pain points and redesign screens for optimal usability.

    • Improve screen validation.

    • Redesign screens for more accessible color and text.

    • Speed up screen load time.

    • Increase broker trust and confidence in the application

    • Include a mobile solution.

Project KPI's

  • Data entry difficulty rate: (Expressed as % rating, “Easy/very difficult”) .

  • Time to input customer data / time-on-task: baseline: 25–30 minutes per client.

  • Navigation clarity score:  Is it intuitive? Expressed as a percentage.

  • System reliability / performance incident frequency (slowdowns or errors)

  • Side-by-side comparison demand (feature need signal) 

  • Messaging workflow integration satisfaction (communication effectiveness) 

  • Overall satisfaction score / CSAT-style measure 

  • Would they recommend?

  • Screen validation quality: form error rate, validation success rate

  • Accessibility/readability compliance :contrast, text legibility; potentially WCAG pass rate.

  • Screen load time: page performance

  • Broker trust & confidence: perceived accuracy & trust rating)

Understanding User Roles

To emphasize the complexity of this application, there were six distinct roles with accompanying work flows and unique screens.  For this case study, I will be focusing on the brokers.  Brokers are the external and internal sales people in the health insurance industry. 

Screenshot 2024-12-12 151539.png

Basic Broker Workflow

Screenshot 2024-12-12 151522.png

The existing application

  • BenefitExpress was created in 2014 with the current standards of the time.

  • It was previously developed by an external vendor that utilized custom code that lacked efficiency and slow performance.  In addition, no efforts were made to even attempt to make the application accessible or user friendly.  The application was developed by developers and product owners with no input from user experience design.

BEX_edited.jpg

Start with a survey

We began with a quick Google documents survey to get a quick understanding of the existing site Below you will see the results:

Survey Results:

  • Ease of use: 7/10 brokers rated data entry as difficult or very difficult.

  • Efficiency: Average time to input customer data was reported as 25–30 minutes per client.

  • Navigation: 8/10 brokers described the interface as confusing or not intuitive.

  • Errors & performance: 6/10 experienced system slowdowns or errors weekly.

  • Features: 5/10 wanted a way to compare product options side-by-side; 4/10 requested mobile access.

  • Communication: 7/10 found messaging with underwriters helpful, but poorly integrated into workflow.

  • Satisfaction: Only 2/10 reported being satisfied overall; none said they would recommend the tool to peers without improvements.

Key takeaway:

Brokers value the platform’s purpose but struggle with its usability, efficiency, and lack of modern features—highlighting a strong need for redesign.

Word Cloud:

Screenshot 2024-12-12 145644.png

Design Challenges

  • Layout & Workflow Complexity

    • Overwhelming density of information due to wide data tables and form-heavy screens.

    • Lack of clear visual hierarchy made it hard for users to know where to focus or start.

    • Sequential flow wasn't intuitive, requiring users to jump between tabs or sections without a guided path.

    • Minimal affordances for progressive disclosure, meaning all data was shown at once instead of being chunked or layered.

  • Screen Real Estate & Responsiveness

    • Wide tables exceeded standard screen width, causing horizontal scrolling and reducing usability on smaller screens or laptops.

    • No adaptive layout to accommodate different roles or user contexts (e.g., brokers vs underwriters may have needed different views).

    • Limited use of collapsible sections or accordion patterns to manage screen length in long forms.

  • Text & Typography

    • Small font size made dense data harder to read quickly, especially in long tables.

    • Inconsistent type hierarchy across labels, values, and headers led to scanning fatigue.

    • Crowded rows with minimal padding between text elements made the interface feel cramped.

  • Color & Visual Design

    • Monochromatic palette was consistent but lacked enough contrast to guide the eye or indicate interaction states (hover, selected, editable).

    • Color-coding (e.g., row highlight colors) wasn’t explained, potentially confusing users.

    • Error or alert states were not clearly emphasized through color or iconography.

  • Use of White Space

    • Very limited white space, especially between rows, sections, and columns, created visual clutter.

    • No breathing room between major components like action buttons, forms, and tables, increasing cognitive load.

  • Interaction & Accessibility

    • Dense interface was not keyboard-friendly or screen-reader accessible.

    • Form fields lacked helpful tooltips, inline validation, or assistive text, slowing down completion and increasing errors.

    • No clear call-to-action hierarchy—buttons like "Back" and "Print Screen" were visually buried or lacked emphasis.

  • Data Visualization Challenges

    • Numerical data was visually undifferentiated, making it harder to interpret values at a glance (e.g., rate, PMPM, and total premiums all look the same).

    • Lack of visual cues like icons, bars, or conditional formatting to quickly surface important or outlier values.

  • System Feedback & Navigation

    • Little system feedback after user actions (e.g., form submissions, error states, or data updates).

    • Top navigation bar was text-heavy and potentially overwhelming, with many similarly styled links.

    • No persistent progress indicators in multi-step processes.

Contextual Analysis in BenefitExpress Research

  • In-context observation: Spent time in brokers’ offices watching them navigate the existing BenefitExpress application to complete real tasks (quoting, enrollment, carrier reconciliation).

  • Observed workflows: Documented step-by-step behavior, including workarounds, pauses, tool switching, and verbal cues that revealed where the interface caused friction in real use.

  • Experience mapping: Translated observations into a screen-by-screen workflow map showing how brokers progressed through tasks, where they hesitated, and where errors or workarounds occurred.

  • Key pain points identified: Found specific issues such as unclear labeling in quoting modules, repetitive data entry across screens, and confusion over benefit configurators that slowed completion and increased cognitive load.

  • Qualitative insights report: Built a contextual analysis report with:

  • Annotated session excerpts (e.g., “Broker stops at Step 3 because the field label doesn’t match the term used by their HR client”)

  • Heat-mapped task flow diagrams showing frequent backtracking

  • Quotes and behaviors tied to specific UI elements

  • Actionable UX prioritized recommendations:  

    • Clarify terminology to match broker language

    • Reduce redundant input steps

    • Improve affordances and feedback in navigation to reduce errors

  • Outcome narrative: Demonstrated how contextual insights explained behind-the-screen decisions (why users took detours, where they guessed, and what slowed them down), making the case for focused design improvements that aligned with actual business workflows.

Wire frames

  • After analyzing the user feedback we thought carefully about information architecture and screen design. 

  • For all screens we created mid-fi wire frames that we then presented to the business for approval.

BexWire.png

Iterating on design and layout

I iterated on the Purchaser screen by starting with mid-fi wireframes to validate the information architecture and task flow, then progressively refining the layout based on stakeholder and broker feedback before handing designs to development.

final.png
 Condensed

What you’re seeing

  • One long, form with little visual hierarchy.

  • Tight spacing, small inputs, lots of inline required indicators.

  • Mixed alignment (some fields stacked, some scattered), with a single prominent SAVE at the bottom-right.

  • Relies on users already knowing what each field means (SIC, MLR, eligibles, out-of-area coverage).

  • ​Strong contrast and good use of company pranding.

Why it was designed this way

  • Optimized for speed for expert users (brokers/agents who do this all day).

  • Those familiar with paper forms responded positively.

  • Desire to limit scrolling for users who might miss fields and cause delays.

  • Error prevention is mostly via required field markers and likely server-side validation after Save.

  • ​Executives and sales felt positive about the use of color for outside brokers.

Tradeoff

  • Fast for power users who memorize the flow, but could be visually fatiguing for newer users.

Card-Based Grouped.png
Collapsible (Card Version)

What changed

  • The page becomes a stack of expandable cards (Group Info, Location Details, Employees & Eligibility, Contact & Key Dates).

  • Each section has its own boundary, icon, and collapse affordance.

  • Keeps the screen visually “lighter” and more modular.

​

Why the designer did this

  • Progressive disclosure: show only what the user needs right now, hide the rest.

  • Supports “review + edit” behavior: users can jump to the section they need without visually wading through everything.

  • Better for smaller screens / responsive layouts: accordion patterns translate well to tablet/mobile.

  • Sets up scalability: if underwriting rules add sections later, the page can grow without turning into an endless wall of fields.

  • Often improves perceived usability because it feels more organized and less intimidating.

​

Tradeoff

  • More interaction cost: expand/collapse adds clicks and can slow expert users if they must open multiple sections.

  • Increased development time and some technical challenges for the time.

  • Risk: users may miss required fields hidden in collapsed panels unless there are strong indicators (e.g., section-level error badges, “3 required fields missing” states).

LongForm.png
One Column Expanded Version

What changed

  • Stronger information hierarchy: page title + short description + a guidance callout up top.

  • Fields are grouped into clear sections (Group Identification, Operational HQ, Eligibility, Timeline & Communication).

  • Consistent spacing, modern inputs, better label readability.

  • Left navigation and a calmer, more modern visual system (likely design-system driven).

​

Why the designer did this

  • Reduce cognitive load by chunking fields into meaningful groups (recognition > recall).

  • Improve scanability and error rates with consistent alignment, spacing, and predictable patterns.

  • Make it more learnable for less frequent users (or new brokers).

  • Improve accessibility: larger targets, clearer labels, more whitespace, clearer focus states (usually), and a more modern contrast model.

  • Support the workflow: the guidance box clarifies why some fields matter (initial quote vs. final proposal).

​

Tradeoff

  • It’s easier to understand, but it’s still a long scroll. For heavy daily use, scrolling can feel slower than compact layouts.

Group 35.png
Users Chose the "compact" design in testing

  • Everything is visible at once:  no scrolling through sections or expanding accordions to find a field, which reduces “where is it?” hunting.

  • Fastest for expert users: compact, dense layout supports high-volume data entry and muscle memory.

  • Fewer clicks / less interaction cost: no collapse/expand states, fewer UI behaviors to manage while entering data.

  • Better for cross-checking: users can quickly compare related fields (employees vs. eligibles, dates vs. prior carrier) on one screen.

  • Feels more “form-like” and familiar: matches legacy expectations and training materials, so adoption friction is low.

  • More predictable "save" behavior: users trust a single, straightforward flow: fill fields and hit Save, without worrying if something is hidden or incomplete in another panel.

  • Traditional Menu Design: Unlike the other designs users were presented a familiar horizontal menu rather than bread crumbs for navigation.

Final Designs 

  • UX and UI worked ahead of developers to ensure that they were busy.

  • We conducted research into how other companies designed similar applications in 2015.

  • We established that a handful of templates would suit our needs:

    • Form

    • Tabular data

    • Text only

    • Results

  • As we redesigned the existing screens, we then presented them as static designs to our users to get their input.

  • We then made any necessary changes and then handed off the designs to the developers.

  • Below are some example screens:

Untitled-2.png
Group 35.png
Untitled-1.png

Design Solutions

  • Simplified Workflow

    • Reorganized complex flows into guided, step-by-step sequences with progress indicators.

    • Modularized long forms using collapsible sections for better focus and manageability.

  • Optimized Layout

    • Column-level controls (sort, filter, collapse) improved data management without needing to leave the screen.

    • Added sticky headers and smart defaults to streamline data entry and navigation.

  • Improved Readability

    • Increased font size and spacing for easier scanning of dense tables and forms.

    • Applied a clear type hierarchy for better distinction between headers, labels, and values.

  • Refined Visual Design

    • Maintained a neutral palette with strategic use of accent colors for clarity and emphasis.

    • Clarified color coding in tables with legends, tooltips, and hover states.

  • Better Use of Space

    • Increased white space between sections and elements to reduce visual clutter.

    • Used cards and section dividers to chunk content into digestible units.

  • Enhanced Interaction & Accessibility

    • Made forms keyboard-accessible and screen reader-friendly.

    • Provided inline validation and helper text for smoother completion.

  • Smarter Data Presentation

    • Highlighted key values and totals with bold styling and strategic placement.

    • Added visual cues like icons and subtle charts for quicker interpretation.

  • Clear Navigation & Feedback

    • Simplified the top navigation to be task-based and contextual.

    • Integrated real-time feedback and status indicators for user actions.

bxwith-pc.png

 This is so much easier to use—night and day compared to what we had before.”

​

I can finally find what I need without clicking through five different screens.


It’s faster, cleaner, and just works—huge improvement.

BenefitExpress 2.0 user

Conclusion

  • BenefitExpress was a mission-critical quoting tool used primarily by insurance brokers to generate real-time small group insurance quotes. The legacy system was large and complex, with hundreds of screens containing dense forms, tabular data, and multi-step flows. It also served a diverse group of users including underwriters, product specialists, and account representatives—each with unique workflows and usability challenges.

  • I led the UX redesign of BenefitExpress with a focus on simplifying the interface, streamlining user flows, and improving performance. Through user interviews,  journey mapping, and iterative prototyping, we identified key friction points and reimagined the application from the ground up.

  • Key Improvements:

    • A modern, user-friendly interface that dramatically reduced training time and user frustration

    • Streamlined quoting workflows that increased speed and accuracy

    • Improved performance and stability, supporting hundreds of simultaneous users in real time

    • Internal cost savings by replacing the need for outsourced tools with an efficient in-house solution

  • Executive Impact:
    Leadership was especially pleased with the rapid turnaround and reduced development costs. The redesigned tool not only improved the user experience but also delivered business value by keeping the solution in-house and increasing quoting efficiency for hundreds of brokers.

  • Although BenefitExpress has since been sunset, the redesigned version had a lasting impact—enhancing productivity, reducing errors, and setting a higher bar for UX within the organization.

bottom of page