An insurer since 1906
The Standard provides disability, life, dental and vision insurance, absence management and retirement plans to employers and individuals.
Information architecture → design → shipped8 min readOne site,
three distinct audiences
No budget for user research, stakeholders asking “what do the users think?” And a mobile-first redesign that’s live today because I went and found out.
- Problem
- The Standard’s public site served employees, employers and brokers, and none of them could find their way. Search was failing, login was confusing, and accessibility sat at 82%.
- My role
- Lead product designer & researcher. I owned the competitive analysis, journey maps, information architecture, tree tests, wireframes and all 15 usability sessions. I worked with a visual designer.
- What I did
- Built an audience-first IA, validated it with 3 tree tests (29 people), designed 25+ mobile-first wireframe pages, and moderated 15 usability tests, all in 12 weeks.
- Result
- It shipped. The audience-first structure is live on standard.com today, along with the label, contact and claims fixes my testing uncovered.
- Timeline
- 12 weeks
- Audiences
- 3
- Tree tests
- 3 · 29 people
- Usability
- 15 sessions
- Wireframes
- 25+ pages
- Status
- Live
Tools
- Figma
- Miro
- Treejack
- User Interviews
- Zoom
- Octopus
Methods
- Competitive analysis
- Journey mapping
- Content audit
- Information architecture
- Tree testing
- Wireframing
- Usability testing
Context
The company and the stakes
Who The Standard serves
Three audiences, one front door
Employees look for their benefits, employers manage plans for their teams, and brokers sell The Standard’s products to clients. They all start at standard.com.
A rebuild, and a chance to fix it
The site had to move to a new platform and meet accessibility standards anyway. That made it the moment to fix how people found their way.
At a glance
Assumptions → evidence
From hunches to a live site
- Audit13competitor sites ranked
- Hypothesize3assumption-based journey maps
- Structure1audience-first IA
- Validate29tree-test participants
- Test15moderated usability sessions
- ShipLiveon standard.com
Oct–Dec 2022 · two-week sprints
01 · Discover
The real problem
The ask was simple. The problem wasn’t.
The ask
- Meet WCAG 2.1 AA
- Upgrade Drupal 7 → 9
- Refresh the design
What discovery revealed
One website was trying to serve three distinct audiences (employees, employers and brokers), and none of them could find their way.
23%of searchers had to refine their query to find anything
13%left the site straight from search results
82%WCAG 2.1 compliance, against a 95% target
0conversion tracking, so no one knew what the site achieved
The constraint
No budget for user research in discovery. I’d have to start from stakeholder knowledge, analytics and competitors, and prove every assumption later.
Learning from 13 competitors
I ranked 13 insurers on the seven tasks that mattered most. The Standard sat near the bottom for navigation and login and the best-in-class patterns became my starting points.
Best in classGuardian
Pattern I borrowedNavigation segmented by audience, with clear drop-down subcategories that don’t overwhelm.
Best in classPrincipal
Pattern I borrowedPick your role or topic, then get the right phone numbers, links and forms, all scannable.
Best in classJohn Hancock
Pattern I borrowedA simple results page that shows the result count and filters to narrow it down.
Best in classGuardian
Pattern I borrowedOne step per page, with what you’ll need listed up front.
Best in classEmpower
Pattern I borrowedChoose your audience after clicking Log In, from a link that’s clearly site-wide.
Three journeys, written as hypotheses
Without users to interview, I mapped three journeys from stakeholder knowledge and labeled them for what they were, assumptions to test. Each one bottomed out in the same place, with people not knowing where to go.
Gordon · broker
Lowest point · ConfusedHe knows the information is on the website, but he can’t find it.
What fixed itA dedicated broker section, renamed “Brokers & Advisors” after testing.
Employee · enrollee
Lowest point · PerplexedTwo log-in buttons on the homepage, and no idea which one is hers.
What fixed itOne clear Log In in the header, plus log-in pathways in each audience section.
Willa · HR lead
Lowest point · OverwhelmedShe can’t help every employee, and doesn’t know which number to send them to.
What fixed itA Contact page tailored by role and topic, linked from every page.
02 · Structure
Information architecture
Organize by who you are, not what we sell
The old site mixed audiences everywhere. There were two near-identical retirement pages, two near-identical workplace-benefits pages, and an “I want to…” section that duplicated both. I reorganized everything around three audience doorways, with a repeatable pattern inside each.
EmployeesIndividuals & Families
EmployersBusinesses & Organizations
BrokersFinancial Professionals ← flagged later
EveryoneGet to Know The Standard
The turning point
When I presented the first IA, stakeholders got uneasy. They had one question.
“What do the users think?”
My call
Don’t argue. Test. I proposed three tree tests, one per audience. They were fast, cheap, and the most direct way to answer it with evidence.
I also added check-ins so stakeholders saw progress, not just a final answer.
Three tree tests, 29 people
Who10 employers, 10 employees, 9 brokers, customers of The Standard and competitors
HowUp to 10 tasks each, randomized so people couldn’t learn the tree
Bias checkBrand-neutral labels. “Get to Know The Standard” became “Get to Know Us”
Not pass/failSome tasks were vague on purpose, to watch how people explore
WorkedEmployees went to Individuals & Families, employers to Businesses & Organizations. Log in, Contact and About tested well across all three.
FlaggedBrokers didn’t see themselves as “Financial Professionals.”
Flagged“Support” read as tech support. 8 employers picked it when asked about the company’s technology.
Flagged“Resource Center” was unclear, and Careers was buried at the bottom of the tree.
Instead of guessing at fixes, I carried every open question into usability testing. Stakeholders regained confidence once they saw the IA hold up with real users.
03 · Design
Sketches → wireframes
Every screen answers a finding
I started on paper, testing how long labels could get before a navigation row broke, and how one homepage could welcome three audiences without confusing any of them. Then I built 25+ wireframe pages in Figma for the visual designer to bring to life.
Home with three doorways
Each audience gets a doorway right under the hero, so people self-select in one click.
↳ Answers “Who is this site for?”Mega menu
Audience tabs with their own sub-menus, plus a persistent “Get in Touch” in every one.
↳ Answers “I can’t find how to contact you”Contact, tailored
Choose your role and topic, and only the contact options that apply to you appear.
↳ Answers “Which number do I call?”Search that helps
Recent searches, sorting, filters and tags to narrow results fast.
↳ 23% refined searches · 13% exitedFile a claim
A guided flow that asks your coverage and role, with forms one click away.
↳ Self-service for the top taskFind a form
Keyword search, a state filter and form-type filters across hundreds of forms.
↳ Self-service for the top taskMobile menu
Top tasks like dental login, filing a claim, reporting an absence or finding a form are one tap away.
↳ Employees mostly visit to log in or actCareers, consolidated
Fewer pages, with more on one landing page, moved where people looked for it.
↳ Tree test showed Careers was buried04 · Test
Dec 12–16, 2022
15 sessions. One week.
I moderated all 15 one-hour usability tests over Zoom, 4 employers, 4 employees and 7 brokers each with its own discussion guide, working through a clickable low-fidelity prototype.
“Tell me you have the most providers, the best rates, the details of how you’re going to help me do business with you.”
“This website is a one-stop shop for employers.”
“It’s easy to use. Easy to find things without any trouble.”
Research overruled my own design
The homepage quick links were mine, and employees loved them. But employers and brokers read them as a sign the page wasn’t for them. I cut them.
ChangedMega menu redesigned so every level looks clickable
ChangedPortal pages cross-link, so anyone on the wrong one finds their way
ChangedBroker & employer copy rewritten from marketing to data-backed claims
ChangedProduct pages hold everything, including info, log in, claims and contact
05 · Shipped
Research → live
Financial Professionals
Brokers & Advisors
Every finding followed the same path. Spotted in the tree test, confirmed in usability, shipped to the live site. Pick one to trace it.
- 01 · SpottedBrokers didn’t look in “Financial Professionals.”
- 02 · ConfirmedOnce relabeled “Broker,” every broker recognized their section.
- 03 · Live todayThe main navigation reads “Brokers & Advisors.”
- 01 · SpottedCareers sat at the bottom of the tree; people looked in “Get to Know.”
- 02 · ConfirmedMoved into Get to Know The Standard, every participant found it.
- 03 · Live todayCareers lives under Get to Know The Standard, with a “Join Our Team” block on the homepage.
- 01 · Spotted“Resource Center” was unclear to employers.
- 02 · ConfirmedEmployers went straight to it as “Research & Insights.”
- 03 · Live todayA “Research & Insights” section and homepage feature.
- 01 · Spotted“Support” read as tech support.
- 02 · ConfirmedPeople expected a persistent Contact link in the header.
- 03 · Live today“Contact Us” sits in the header, next to Log In.
- 01 · SpottedHomepage quick links served employees only.
- 02 · ConfirmedEmployers and brokers weren’t sure the page was for them.
- 03 · Live todayThe homepage leads with three audience doorways instead.
- 01 · SpottedEmployees expected to log in or file from product pages.
- 02 · ConfirmedReport an Absence belonged inside File a Claim.
- 03 · Live today“File a Claim or Request a Leave” is one combined task.
Don’t take my word for it. The structure is live.
Visit standard.com ↗06 · Reflect
Looking back
If I ran it today
Even five quick interviews in discovery would have turned my journey-map hypotheses into evidence weeks sooner and spared stakeholders the uneasy moment.
With no conversion tracking, I couldn’t prove the lift. Today I’d push for baselines (search exits, task success) before launch, not after.
I synthesized 15 sessions by hand in a week. AI-assisted tagging would get me to themes in hours, leaving more time for a second round of testing.