/

mfaAuthentication

Multi-Factor
Authentication

Multi-factor authentication (MFA) is a security process in which a user provides two or more independent forms of identification, called factors, to prove one’s identity.

UI design

UX research

Service design

Internal Tooling

About

Problem & Scope

The Capital One internal MFA team had been struggling with users being unable to self service registration of their new multi-factor devices in their current portal. Our support channel was having to spend countless hours walking new users through registration due to confusion.

The task was to redesign an outdated, non-compliant application with corporate design standards.

My Role

I was a UX designer and UX design lead.
I worked with a design team, tech lead, and product owner at Capital One Bank.


Research

Discovery

I looked into how industry practices were conducted throughout other websites to fulfill the same outcomes I was hoping to achieve (i.e credit card companies often display “bright” banners signaling a special offer/promotion that attract users to select preferred options).

Problems Uncovered

I did not conduct formalized empathy research as it was done prior for me but essentially the MFA team discovered that users were unaware of what MFA options currently existed, had issues visually discerning between iconography, and unaware of the difference between Passwordless vs. Non Password options.

Ideation

We sat with our product team and diagrammed potential solutions we could see our customer happy paths leading to. Here – we brainstormed on some different UI sequencing that would make sense for users to visualize different options and threw together a bunch of different iconography solutions.

Drawing the happy path diagram for our MFA solution

Prototyping

While prototyping some different outcomes, to entice users to select Passwordless options, I tried to divide options into separate rows and add a collapsible that only displays Passwordless requirements.

Collapsibles in this context felt like it added too much human effort to hide and show information – we quickly scrapped the idea.

During our prototyping, we also made sure to check for ADA compliance and brush up font/color and iconography to our Capital One Gravity Standards

Changes in Requirements

As most things in development never go to plan, during the middle of testing and prototyping we were told by leadership that current state could no longer recommend users to opt for Passwordless for the time being. We decided to then offer an MVP state and a future state design.

Testing


We worked with our research team and set up 10 interviews with power and non power users. With our first high fidelity mock up completed, we wanted to mainly focus on seeing if users would opt for Passwordless options (or figure out how to educate them) and if users could follow the tasks for registration completion. Our results were a bit all over the board:

  1. Users felt that iconography changes did not make a large difference on whether they understood the differences between Yubi key and Security key (it was still confusing for nearly all interviewees)

  2. All users tested felt that the new UI was more intuitive and visually pleasing

  3. All but 1 user opted to choose the Non – Passwordless option.

    With our fairly positive test results, we did one final UI polish and launched our design.

High-Fidelity Designs


Registration and Select Device

After testing a few different registration screens, we kept our original UI formatting because users had become so accustomed to it that some of our new test screens were seen initially confusing and out of user mental models. Our Iconography and overall UX clean up solved many user complaints and confusions.

Register a New Device

This Register a New Device screen was selected over 80% of the time above other options. It gives more clarity into the differentiation between Passwordless vs. Non Passwordless options and additional context between each device. It gives a light "nudge" into choosing the preferred method of Passwordless for our first MVP launch.

Results


After deployment of our new designs:

MFA registrations hit their goal of 85% of users (60,000) onto Passwordless, with sometimes 2,000 or 3,000 people were registering a day sometimes

Data collected showed that there was a 15% increase (from 50% to 75%) of users that registered for Passwordless devices

The support team saw a 10% decrease in slack questions being asked after our launch

Learnings

During a MVP validation session of 20 users where Kai was first tested in front of users (during production), it was apparent that there was no design process taken when designing the product previously by the team. Customers complained about a visually confusing product that wasn’t intuitive and didn’t have the features they even wanted.

Having to retrace our steps to completely, I reworked a lot of product features to make sure we were building a customer-centric bot. I learned a lot about the detriments of implementing bad design.