Overview
- Role
- UX Research Co-Lead
- Timeline
- Jan — Jun 2026
- Team
- Strategic Communications & Technology Operations
- Methods
- Interviews · Surveys · Affinity Mapping · Competitive Analysis
TL;DR
I co-led research into why the University of Michigan DPSS intranet had become something staff worked around instead of relied on. Surveys, role-specific interviews and an executive focus group showed that broken trust, not search or content alone, was driving the cycle, and I delivered six strategic priorities, a governance model and a 20-week implementation plan to leadership.
- 553Active users
- 43.5%Daily visit rate
- 9Role segments
- 130+Insights mapped
- 92Survey respondents
Confidentiality
Because of the confidential nature of internal security operations, some project details and deliverables cannot be shared publicly.
Why did a tool used by 553 people stop working for nearly all of them?
The University of Michigan’s Division of Public Safety and Security was transitioning away from Confluence. The existing site—“ComHub”—was plagued by usability issues staff worked around rather than relied upon.
Problem statement
DPSS staff across all roles and campuses cannot reliably access, find, or trust the tools and information they need—resulting in workarounds, disengagement, and operational risk.
My contributions
- Conducted 6 role-specific user interviews
- Synthesized 130+ insights via affinity mapping
- Delivered strategic recommendations to leadership
- Facilitated executive focus group
- Developed personas and navigation priorities
Research approach
- 01DiscoverySurvey data (92 respondents) · 90-day site analytics · 9 role segments
- 02Qualitative6 role-specific interviews · executive focus group · behavioral walkthroughs
- 03SynthesisAffinity diagram (130+ insights → 10 themes) · vendor evaluation matrix
Role segments
- Security Officers
- Patrol / Police
- Dispatch
- IT
- HR
- Admin Assistants
- Assistant Directors
- Strategic Communications
- Executive Cabinet
Three key findings
Fragmented Ecosystem, No Single Source of Truth
Critical systems operated in isolation. BRIEFS had been offline since the July 2024 CrowdStrike incident. Telestaff required duplicate data entry. Cross-unit IT systems failed to integrate—leaving staff to triangulate between disconnected tools with no hub.
Every time I make them sign in or click redundantly, it’s taking time from delivering service.
Documented gaps
- BRIEFS offline 18+ months
- Duplicate data entry required
- No cross-unit IT integration
- Zero unified access point
Failed Trust and Content Governance
No content governance existed at any level. Org charts were approximately one year out of date. Over 75% of roster-listed staff were no longer employed by DPSS. Policy documents were up to ten years old. No channel framework meant no mechanism to flag, archive, or retire stale content.
Less is more. Accuracy is the big thing. Just being able to know if I can rely on it.
- ~1yrOrg chart staleness
- 75%+Roster staff no longer employed
- 10yrOldest policy documents
- 0Content governance framework
Structural Barriers to Access
Long-scroll layout, broken search, and no role-based organization made the intranet structurally hostile to everyday use. Poor mobile experience compounded the problem for field staff. Most critically: Dearborn and Flint campuses were entirely absent from the platform.
The fact that they haven’t received much energy or attention has been noticeable.
- NavigationBroken
- SearchBroken
- MobilePoor
- Dearborn CampusAbsent
- Flint CampusAbsent
A self-reinforcing cycle of failure
Every breakdown fed the next. The key leverage point was not search, not content—it was Trust.
- 01Findability FailuresUsers can't locate what they need
- 02Workarounds AdoptedStaff build personal bypass systems
- 03Content BypassedIntranet goes unused by staff
- 04Content Goes StaleNo views — no incentive to update
- 05Break pointTrust ErodesCritical mass reaches disbelief
- 06More WorkaroundsCycle accelerates, never breaks
The cycle repeats: step 6 feeds back into step 1.
Six strategic priorities
- Role-Based Information Architecture Organize navigation around who you are, not where content happens to live in a folder tree.
- Integrate PowerDMS Centralize policy management with a single auditable, authoritative source used across all campuses.
- Build Dearborn and Flint Presence Regional campus inclusion as a first-class design requirement—not an afterthought or phase two.
- Content Governance Model Assign owners, establish review cycles, implement archive flags and retirement protocols throughout.
- Prioritize Search Functional, role-aware search as a platform-level requirement—not an optional feature to layer on later.
- Set as Default Homepage Drive adoption through habit formation. Make the intranet the ambient workspace, not an afterthought.
Impact & deliverables
Research outputs
- 6Personas
- 24Pages <5 views
- 43%Pages flagged
- 4Vendors evaluated
- 20wkImplementation plan
- 1Governance model
Technical requirements
- SSO with MFA
- Mobile-Responsive
- Role-Based Access
- PowerDMS Integration
- WCAG 2.1 AA
Vendors evaluated
- Simpplr
- LumApps
- Interact
- Appspace
What worked. What I’d change.
Worked well
- Mixed-methods approach
- Affinity mapping at scale
- Executive focus group
- Behavioral walkthroughs
Would change
- Involve content owners earlier
- Include frontline officers from the start
Key takeaway
If you build it the right way, it really becomes such a great training and onboarding system for them. It’s guiding them rather than the human capital we’re investing in for consistency.
The intranet was used by nearly everyone—but wasn’t working for nearly anyone. The opportunity: build a single, trustworthy, role-sensitive environment that reduces friction, supports emergency response, connects regional campuses, and gives leadership a reliable communications channel.
