← All work

Automatic Required Minimum Distribution

Simplifying retirement taxes

A self-service tool for clients aged 72+ who want to set and forget their required minimum distributions online.

The Auto-RMD tool shown on a laptop
Role
UX Lead
Platform
Responsive Web
Focus
Workflows · Responsive Design
Timeline
2019-2020

Overview

Once an investor turns 72, the IRS requires them to withdraw a Required Minimum Distribution (RMD) from their retirement accounts every year, or face a steep tax penalty. For T. Rowe Price, this meant a recurring flood of phone calls and manual transactions from a client base that would rather not think about it at all.

My Role

UX lead responsible for defining the workflows, detailing the use cases, producing all design deliverables, running user testing, and ensuring design quality through development for an audience that is not always comfortable with technology.

Uncovering the Business Logic

The project began with stakeholder interviews. I met with product managers and sales representatives to understand the total flow of the experience, including its edge cases and error handling scenarios.

Mapping the complete workflow up front gave the team a shared understanding of every decision point and the inputs required from clients, laying the foundation for the design work that followed.

Auto-RMD workflow diagram
Auto-RMD workflow

Designing the Entry Page

The entry page sets the tone for the whole experience: it has to explain what the tool does and get people comfortable enough to begin. My first direction paired that explanation with a warm lifestyle photograph of a retiree.

In moderated testing, the photo backfired. Instead of reading the value proposition, participants fixated on the person. They commented on her age and how she was different from them, or spun stories about what her life must be like. The image pulled attention away from the task and, for some, created distance rather than connection.

Photo-led entry page in user testing
Testing the photo-led entry page. Participants kept describing the person, not the tool.

In response to the user testing results, I moved to an illustrated entry page. While the illustration was not universally praised, testing proved it was far less of a distraction. It set the right mood and got people into the experience more easily.

Illustrated entry page
The illustrated entry page. Less distraction, easier to step into the flow.

Flow Breakdown

With the framing set, I designed the full flow as five primary steps. Users are shown information and inputs progressively, so the experience is chunked into manageable decisions rather than one long form.

Step 1 on desktop
Step 1 on mobile
Step 1 of the Auto RMD experience across desktop and mobile

Error Handling

The happy path was only part of the work. Underneath it sat a long tail of edge cases that all had to be designed and written. A single decision, how the client wants their distribution delivered, branched into 13 potential error scenarios across mailing a check, direct deposit to a bank, and reinvesting into a non-retirement account, each with its own message, cause, and recovery action.

Error states by delivery method
All 13 error states for the delivery-method step, mapped by mail, bank, and reinvest

Another example can be found when clients split their RMD across funds by percentage. It has to respond in real time to an amount that exceeds an account balance, or allocations that do not add up to 100 percent, guiding the client back on track without ever making them feel stuck.

Allocation validation and information states
Inline validation on the allocation step: insufficient balance and a running "must equal 100%" check

Additional Information States

Beyond errors, many screens needed information and validation states to guide clients before anything went wrong. I worked closely with a content strategist to define each one: what triggers it, what it says, and how to phrase it plainly for an audience that is not always comfortable with technology or financial jargon.

Tell me more information states
Tell me more information states
Information states written for each step of the flow

Calendar Complexity

Scheduling looked simple on the surface: pick the months and the day you want your distribution. Underneath, the date logic was some of the trickiest work in the project. A client can choose any combination of months, but individual months carry their own blackout days.

December blacks out the 25th through the 31st. January blacks out the 1st through the 5th. February only has 28 days, so the 29th through 31st never apply. Because a single day has to work for every month a client selects, the calendar has to resolve the combined rules and only offer the days that are valid across all of them.

Calendar step in context
The scheduling step: choose months, then a valid day
Calendar with different month selections
Calendar with different month selections
Left: Jan, Feb, and Mar selected. Right: all twelve months selected, leaving only the days valid in every month.

Results

T. Rowe Price's Auto-RMD tool lets clients aged 72+ set and forget their Required Minimum Distribution independently, on their own time, with the peace of mind that once it is set up they do not need to come back and do it again every year.

By designing for the edge cases as rigorously as the happy path, and by validating the flow with real clients, the tool turned an annual source of calls and anxiety into something people could handle on their own time.

01
Every edge case, detailed
Delivered comprehensive designs that granularly documented each edge case and error scenario, not just the happy path.
02
Validated with clients
User testing on usertesting.com validated the design's ease of use to a high level of confidence with potential clients.
03
Reduced calls
The release of the Auto-RMD tool resulted in fewer calls to the call center.
04
Created an internal tool
As an unintended bonus: the customer service team started using the tool to help facilitate customer calls, improving the internal experience as well.
Keep exploring
See more case studies
All work →