988Suicide&CrisisLifelineHomepageRedesign
Redesigning a national crisis service to reach people in their darkest moments
Role
Lead Product Designer & UX ResearcherSolo Project · Commissioned by Vibrant Emotional Health
Scope
- Google Analytics Audit
- Unmoderated Usability Testing
- Information Architecture
- High-Fidelity UI Design
- Stakeholder Presentation
Stakeholders
- Vibrant Emotional Health
- Substance Abuse and Mental Health Services Administration (SAMHSA)
Status
In stakeholder review with SAMHSA & Vibrant Emotional Health
Disclaimer
The views, opinions, and content expressed in this case study are proposed solutions and do not necessarily reflect the official views, opinions, or policies of the Substance Abuse and Mental Health Services Administration (SAMHSA), or the U.S. Department of Health and Human Services.
TL;DR
- —
Problem: Over 1M people visited the 988 Lifeline homepage in Q4 2025. 65% left without taking action, with an average session of 18 seconds.
- —
Approach: Combined a Google Analytics audit with a 9-participant usability study to pinpoint where the IA was failing, then redesigned the page around the psychological questions a hesitant help-seeker actually asks.
- —
Design goal: Reduce the distance between someone in crisis and the support they need by surfacing privacy reassurances, trust-building content, and community resources at the moment users need them.
Before
AfterHamburger menu hides navigation on desktop. No trust-building content, privacy information, or community resources above the fold. 65% of visitors left without taking action.
Navigation exposed directly in the header. Privacy reassurances and eligibility content surfaced above the fold. Community-specific resources accessible from the homepage.
Context & The Problem
The 988 Lifeline Homepage is a high-stakes, critical national service. It is the most visited page on 988Lifeline.org and plays a vital role in guiding help seekers to immediate support.
A Google Analytics audit revealed the true scale of the problem: over 1 million total users landed on the homepage in Q4 2025 (Oct 1 – Dec 31, 2025). The page had a 65% bounce rate, a significant gap for a service designed to connect people in crisis with immediate support.
1M+
Total users, Q4 2025 (Oct 1 – Dec 31, 2025)
65.34%
Bounce rate — left without taking action
18 sec
Average engagement time
660,000+
help-seekers left the homepage without finding support — in a single quarter
On a crisis line, that number is not a metric. It is the magnitude of the problem.
The funnel below draws from a separate GA4 path analysis filtered to users who entered directly at 988Lifeline.org — isolating help-seekers who came straight to the homepage, separate from traffic arriving via landing pages, external links, or email campaigns. Despite the difference in scope and time period, both datasets show a consistent bounce rate in the mid-60s. The pattern holds whether you measure all traffic or only direct help-seekers, which rules out the possibility that bounce rate was inflated by irrelevant traffic sources.
User Flow — Funnel Drop-off
Help seekers who arrived directly at 988Lifeline.org
31,409
Bounced without action
63.9%
9,584
Explored other pages
19.5%
8,172
Reached pre-chat survey
16.6%
Why this dataset: Sitewide GA4 data tracks conversions, bounce rate, and engagement but cannot isolate how help-seekers specifically moved from the homepage to the pre-chat survey. This funnel uses a GA4 path analysis filtered to users who entered directly at 988Lifeline.org, making it possible to precisely track the three outcomes above.
Bounce (GA4): Session lasted less than 10 seconds, triggers no conversion events, and does not lead to a second pageview or screenview.
A Signal Hidden in the Data
While the homepage was losing 65% of visitors in 18 seconds, two pages deeper in the site told a completely different story. Users who found them stayed and engaged deeply.
"What to Expect" page
49s
Avg. engagement
38%
Bounce rate
"Help Someone Else" page
49s
Avg. engagement
35%
Bounce rate
The content wasn’t the problem. The architecture was. These pages were performing; they just weren’t reachable. This became the central thesis of the redesign.
Research & Methodology
To understand why 2/3 of users were leaving without interacting, I designed and conducted an unmoderated usability test via Qualtrics with 9 diverse participants. My goals were to evaluate how easily users could find immediate help, identify gaps in trust regarding privacy, and assess the discoverability of community-specific resources.
As a solo project without a moderated testing budget, I used unmoderated sessions to maximize reach within constraints. The tradeoff: less ability to probe in the moment. I mitigated this by writing scenario-based tasks with open-ended follow-up prompts, and by triangulating qualitative findings against the GA data to separate signal from noise.
To ensure findings held across the communities the 988 Lifeline is designed to serve, I recruited participants across a spread of age, gender, ethnicity, and geography.
Participant Breakdown
9 participants
Gender
- Male
- Female
- Non-Binary
Ethnicity
- Asian or Pacific Islander
- Black or African American
- Hispanic or Latino
- White
Age
- 18–24
- 25–34
- 35–50
- 51–65
Area of Residence
- Urban
- Suburban
- Rural
Key Discoveries
Where It Failed
Critical trust-building information was buried. Users expressed deep fears about reaching out, but finding “Privacy and Confidentiality” required navigating a 4-click deep menu path: Menu → FAQ → About Us → What does “confidential” mean.
The Missed Opportunity
Community-specific resources were highly valued but incredibly difficult to find.
What Worked
9 out of 9 users found it “Easy to Very Easy” to find the immediate contact options (Call, Text, Chat). The hero CTA was working.
Friction Points
Qualitative testing revealed three critical points where the design failed to meet users' psychological and informational needs:
The Discoverability Disconnect
Reassurance
Help-Seekers Need Reassurance Before Acting
For a high-stakes crisis line, visible buttons are not enough. Users must feel safe clicking them. Participants turned to FAQs to validate concerns before acting, and when trust questions go unanswered upfront, uncertainty leads to drop-off.
“One of my biggest concerns is that you ask for help and then next thing you know, people call the cops and you have to be entered into an asylum or something.”
Privacy
Privacy Is Highly Important, but Highly Invisible
Users need to know their interactions are secure. Yet finding privacy information required a 4-click journey buried deep in navigation, the opposite of what a hesitant user needs.
“Where the heck, what privacy and confidentiality? So sometimes I can't find stuff like that.”
Community
Community Resources: High Value, Low Discoverability
Community-specific resources for veterans, Asian American, Native Hawaiian, and Pacific Islander communities, and others were highly valued but deeply buried. Users didn't know these dedicated lines existed until they navigated well past the homepage.
“I didn't even know that these specific support groups for these specific communities was an option until I went through a little bit further into the menu.”
Shifting from Content to Intent
The homepage was organized around what 988 is — not what someone in crisis needs to know before they can act. Three separate observations, each from a different data source, pointed to the same root cause.
Stakeholder Observation
During a stakeholder meeting, someone noted that the "What to Expect" page was being used to answer emotional questions: will police be called? Can I stay anonymous? Not just to learn the process. Help-seekers were arriving with psychological intent; the page was organized around institutional content.
Competitive Analysis
I benchmarked other crisis hotlines and found that their navigation and homepages were structured around the help-seeker's journey. Resources were separated by who they served: people in crisis vs. professionals, donors, and providers.
IA Audit
Examining 988's menu confirmed the pattern: help-seeker content and organizational resources were mixed with no distinction. Under "Learn" sat "Our Crisis Centers," content for administrators and providers, not someone reaching out in a moment of distress. Careers, Donate, and professional tools lived in the same space as crisis support.
The Finding
The navigation was organized around what the organization wanted to say, not what a help-seeker needed to find. Separating these two audiences (help-seekers from professionals) became the central IA principle driving the redesign.
To make this concrete, I mapped how a help-seeker looking for emotional support actually navigated the current site and how many steps stood between them and what they needed.
Task Flow — Finding Anxiety Support
“Learn” is surfaced directly in the nav bar. Its dropdown exposes topic-specific resources, including Anxiety, without requiring users to open a hidden menu or navigate through an intermediate page.
Six Questions Every Help-Seeker Asks
I restructured the homepage, eliminating the hidden hamburger menu on desktop and reorganizing the page vertically based on the psychological questions a hesitant user asks themselves.
Defining Success
I defined success as a 10–15 point reduction in bounce rate and a measurable increase in the ratio of homepage visitors who reach the pre-chat survey, the last observable step before a user connects with a counselor. At 1M+ visitors per quarter, a 10-point improvement translates to roughly 100,000 additional people per quarter moving from the homepage to active crisis support.
Section 1 — Hero Area
“How can I connect?”
Kept the primary CTA buttons (Call, Text, Chat, Deaf/HoH) highly visible, as testing proved this was already working. Desktop users now see an exposed navigation menu rather than a hidden hamburger icon.
Section 2 — Trust Building
“What if I'm still afraid to reach out?”
To directly combat fear of institutionalization and police involvement, I pulled Privacy Concerns, Is This Right For Me, and Top FAQs out of hidden menus and placed them immediately below the hero.
Design Decision
Considered
Chosen
Considered: The current site has a "Get in Touch" section with reassuring copy: "The 988 Lifeline is for everyone... Our skilled, judgment-free counselors are here to provide compassionate support, 24/7/365." It was warm, accessible, and already written.
Why: During testing, participants said they didn't know you don't have to be suicidal to use 988. That insight reframed the problem: users weren't just unsure how to reach out; many were unsure they were allowed to. A user who self-selects out before even reading the CTAs is a different failure mode than one who can't find the phone number. Eligibility uncertainty needed to be addressed first.
Section 3 — Expectation Setting
“How is this going to work?”
GA data revealed that users who navigated to the "What to Expect" page engaged for 49 seconds with only a 38% bounce rate, far outperforming the homepage's 65% bounce rate. This quantitatively proved the content is highly effective at retaining hesitant users, justifying its prominent placement on the redesigned homepage.
Section 4 — Social Proof
“Will this work?”
Research revealed that fear and uncertainty were the primary barriers to action, not awareness of the service. Participants needed validation that reaching out was worth it. A Testimonial Widget featuring "Stories of Hope and Recovery" directly addresses this by showing that others in similar situations reached out and were helped — reducing the perceived risk of making contact.
Section 5 — Community Resources
“Who else is this for?”
GA data showed that when users navigated to the "Help Someone Else" page, they engaged for 49 seconds with a 35% bounce rate, making it the site's most engaging content yet out of reach for over 1 million homepage visitors. Surfacing community-specific resources on the homepage was one of the clearest opportunities to close the gap between what users needed and what the page offered.
Section 6 — Safety Net
“I'm not ready yet — what else can I do?”
My Safety Plan, Safe Space, and Kind Mind are Vibrant Emotional Health products (the parent organization behind 988 Lifeline), not currently available on 988Lifeline.org. Recognizing that a meaningful portion of visitors aren't ready to call but still need somewhere to go, I proposed integrating these tools as a safety net at the bottom of the page. This recommendation goes beyond the homepage redesign: it surfaces a cross-product opportunity for Vibrant to extend its existing tools to the 988 audience and reduce the gap between landing on the site and finding meaningful support.
Current
Proposed
Section 1
“How can I connect?”
Hero CTAs kept prominent.
Section 2
“I still have concerns.”
This section is to help reduce uncertainty and remove barriers to taking action.
Section 3
“How is this going to work?”
What to Expect shows users exactly what happens when they reach out, reducing uncertainty.
Section 4
“Will this work?”
Testimonial widget provides social proof that reaching out helps.
Section 5
“Who else is this for?”
Community-specific resources given dedicated homepage real estate instead of buried sub-pages.
Section 6
“I'm not ready — what else can I do?”
Safety net at the bottom: My Safety Plan, Safe Space, and Kind Mind for users not yet ready to call.
Next Steps & Reflection
Because implementing changes on a national healthcare site requires strict federal approval, design is only the first step.
- 1
Stakeholder Alignment
High-fidelity recommendations are currently undergoing approval with SAMHSA and Vibrant Emotional Health. The proposed community resources section gives SAMHSA the dedicated homepage real estate needed to surface specialized lines as federal priorities evolve.
- 2
Engineering Evaluation
Collaborate with the development team to evaluate level of effort and plan phased roadmap iterations.
- 3
Further UX Research
Conduct a dedicated Card Sorting exercise to fundamentally restructure and simplify the complex global menu.
- 4
Phased Implementation
Execute roadmap iterations informed by ongoing research and stakeholder feedback.
Roadmap — Agile Progression
Stakeholder Alignment
SAMHSA & Vibrant Emotional Health sign-off on high-fidelity recommendations
Engineering Evaluation
Level-of-effort assessment and phased delivery planning with dev team
Card Sorting
Restructure global navigation grounded in participants' own mental models
Phased Implementation
Iterative rollout informed by ongoing research and stakeholder feedback
Reflection
If I approached this again, I would invest in card sorting earlier — before rather than after usability testing — to ground the new IA in participants' own mental models rather than deriving it from observed friction alone. The redesign is a well-reasoned hypothesis; the next step is stress-testing it.