Delta Dental Enrollment Dashboard

Project overview

The problem

Analysts had to find critical errors across disconnected systems

Enrollment Analysts inspected more than 150 files each day using five to eight legacy tools that did not communicate reliably with one another.

Errors had to be identified manually or through overnight reports that could arrive late or fail to arrive altogether. Analysts also had no centralized way to confirm whether files and emails had been successfully processed.

When an issue affecting the Coca-Cola account went undetected for nearly a year, Delta Dental incurred more than $2 million in losses.

How might we help Enrollment Analysts identify and resolve issues before they become costly mistakes?

Research

Fragmented tools made problems difficult to find and even harder to resolve

I partnered with UX Research to facilitate a three-day workshop with Enrollment Analysts and subject-matter experts. Through interviews and journey mapping, we examined analysts’ daily responsibilities, gaps in the existing workflow, and opportunities to reduce manual work.

Research revealed four recurring problems:

• Analysts used as many as eight tools to investigate a single file.

• Errors were often discovered through delayed overnight reports.

• Failed emails could go unnoticed.

• Analysts lacked a shared view of file and issue status.

Strategy and scope

We focused the first release on visibility, prioritization, and traceability

I facilitated an impact-versus-feasibility exercise to balance analyst needs, business risk, and technical effort. We prioritized four core capabilities to deliver incrementally over nine months:

1. Dashboard overview

2. Centralized error management

3. File-receipt tracking

4. Email-status tracking

We used four principles to guide the work:

• User control: Let analysts focus on the accounts and information relevant to them.

• Actionable information: Prioritize information that requires attention.

• Current system status: Show the most recent data available.

• Transparency: Help analysts understand what happened and what to do next.

Decision 1: Centralize the daily workload

Give analysts one place to understand what required attention

The first capability established a centralized overview of file activity, errors, and email status. Rather than checking several systems independently, analysts could begin from one dashboard and investigate issues from there.

The design prioritized:

• Customizable views by analyst and date

• High-visibility failed and pending states

• Direct access to detailed records

• Only the information needed to make a decision

• Progressive disclosure for large datasets

Place these visuals together:

1. Annotated dashboard

2. Large-data explorations

3. Final dashboard walkthrough

Working with subject-matter experts, I identified which information analysts needed immediately and which details could remain hidden until requested. Tables supported comparison, while expandable sections reduced the initial information density.

Decision 2: Surface errors and missing files earlier

Help analysts identify problems before they escalate

Previously, analysts depended on manual checks and overnight reports to discover processing errors. We combined error management and file-receipt visibility so analysts could identify failed, missing, or delayed files sooner.

The experience allowed analysts to:

• Review affected files and error details

• Distinguish pending and resolved issues

• Identify missing file receipts

• Prioritize high-risk problems

• Track the status of an investigation

Decision 3: Make system communications traceable

Show analysts whether important emails actually reached their destination

Analysts previously copied themselves on outgoing emails because the system did not clearly confirm whether messages had been delivered.

We created a centralized email-status view that showed:

• What was sent

• When it was sent

• Whether delivery succeeded

• Which messages failed

• Which messages could be resent

Testing changed the product

We removed a feature that felt like surveillance

We explored comparing individual analyst performance with team performance. Although the concept introduced greater visibility, analysts felt that it emphasized employee monitoring rather than helping them complete their work.

Based on that feedback, we removed the comparison and kept the dashboard focused on workload and issue resolution.

We reorganized navigation around analysts’ workflows

We tested the dashboard with 14 Enrollment Analysts and two managers. Participants understood the overall product direction, but the second round revealed that navigating between file receipts, email status, and related information was more difficult than expected.

We revised the information architecture by:

• Giving file receipts and email status dedicated destinations

• Grouping related operational information

• Making high-priority tools easier to access

• Creating a clearer path between overview and detail pages

We replaced a real-time promise with honest system status

Engineering constraints revealed that some error data could not update in real time because it first needed to pass through system processing.

Rather than implying that all information was live, we redesigned the experience to show pending and resolved counts. This gave analysts useful transparency while staying within the technical limitations.

Validation and impact

The dashboard made investigations faster and reduced operational risk

We tested the dashboard with 14 Enrollment Analysts and two managers to determine whether they could navigate the experience, find important information, and understand what required attention.

Reflection

What I would improve next

• Investigate data limitations earlier in the design process

• Define what “real time” means for every system dependency

• Plan for delays, missing data, and exceptions from the beginning

• Continue improving navigation and issue prioritization

This project taught me that designing an operational dashboard is not simply about putting information in one place. The experience also has to help people understand what requires attention, what has already happened, and what action they should take next. Working closely with analysts and engineers helped us turn a fragmented workflow into a more transparent and manageable system.