Security center information architecture refresh

Project Details

The Security Center exists to protect the financial security of our members and provide peace of mind regarding USAA products and services through educational, helpful and well organized content. In doing so, we also maintain and, where applicable, exceed regulatory requirements, and aim to provide members a way to self-service their problems using their channel of choice.

The Problem

Security Center pages were being transitioned to a new platform and a new IA was created without consulting design. The proposed IA buried links to security resources, making it more difficult for members to find them.

The objective (WIP)

In order to support this effort, we evaluated the proposed information architecture (IA) through testing prior to the transition of Security Center pages from Tridion to AEM.

MY ROLE (WIP)

I worked as the primary researcher, facilitating working sessions with various partners to refine the Information Architecture of USAA’s Security Center.

Notes: provide current metrics for people visiting — is it because they don’t need it or because they can’t find the right info? Once in production, we will measure metrics and adjust.

We did a tree test with the following scenarios. We also wanted to evaluate whether users expected to find them under support or security, so we copied the page titles under both support and security. The options were randomized and we included 3 non-security related tasks to throw them off.

We knew we wanted to do a tree test to have a benchmark and then create a new IA that was user-centric to test it against.

Activity Summary

DURATION
4 months

MY ROLE
UX Researcher

TEAM
SEO Specialist
Security Experience Owners
(Jeremy Fischer Role?)

METHODOLOGY
Content Audit
Competitive Analysis
Card Sorting
Tree Testing


The challenge

Our security partners had created an IA without engaging with design. We were concerned because it removed the current link for security resources from the main page and decided to put it the Footer under Support.

They already had a plan to consolidate 44 pages into 22 and shared a spreadsheet with us with links to the existing pages. The pages combined similar pages and prioritized content that met FFIC (??) requirements. They were about to put the proposed IA into production without any prior testing.

I clicked on the pages provided and discovered a few more issues that were not addressed:

  • There were still redundancies between many of the 22 pages

  • The titles for the pages included a lot of jargon

Problem statements

  • Security Center pages are neither easily findable by members nor prominently displayed within usaa.com navigation

  • Next step actions are not always clear - either to reach the intended solution or leave the Security Center experience

Findings

  • Participants struggled to find pages that used jargon or were too vague. Proximity to links or pages with similar titles also resulted in wrong selections.

  • Tasks with the lowest success rates also had the highest number of participants who selected Contact Us, which can result in increased strain on our call centers.

  • SC’s current information hierarchy is consistent with participants’ ranking of most important content. We should keep these resources prominent and in visible locations for easy discoverability.



Project timeline

 

Benchmarking


What i did

I decided to click through all the security center pages, starting on the main one, and clicked on all the other linked pages to map out what is in production and which ones were included in the spreadsheet. After discussing with my manager, we agreed that a more detailed audit was necessary. I took screenshots of the pages and put them in Mural, and then numbered them to map against the spreadsheet our partners provided.

  • Took screenshots of the pages and put them in Mural for easier analysis

  • Analyzed the content that was being combined



Audit findings



After reviewing with our partners, we decided to move forward with testing. We conducted a series of tests.

USER TESTING


We conducted card sorting and tree studies since they are complimentary methods:

  • Card sorting - Participants were given cards with security topics to group in a way that made sense to them. This helps us create IA models.

  • Tree Study - Participants were given security scenarios then asked to find the correct page using a website navigation (IA). This helps us evaluate IA models.

 

Tree test - partner proposed ia

We designed an unmoderated tree test to determine if users are able to find the correct Security Center resources in the current proposed IA.

RESEARCH FINDINGS

  • Majority of participants chose answers under security vs. support

  • Findability influences likelihood of contacting customer support

  • Need to avoid ambiguity of titles and use of jargon

  • Need to either distinguish between Mobile device security, biometrics, and general App info or combine to eliminate confusion

 

 

Content Audit + CompARATIVE analysis

 

 

Closed Card Sorting

 

 

Open Card Sorting

 

 

Create new IA - Partner workshop

 

 

Tree test of new IA

 
 

Results

Overall results:

  • Content from 22 Security Center pages were consolidated to create a total of 10 pages, which were distributed between three main categories: Advice (located in the global nav), Security, and Support (both in footer nav)

  • Version 2 performed better overall, showing statistically significant improvement in 5 out of 11 tasks. V2 performed worse than V1 in only one task.