Designing a Consumer Advocacy Startup From Zero to a Multi-Product Ecosystem
I worked directly with the founders and was responsible for shaping the product from zero → MVP → multi-platform ecosystem used by claimants, advocates, attorneys, and partners.
A blank slate in a high-stakes domain
Turnout began with no existing product foundation — no user flows, architecture, portal, brand, or content. The design challenge was to define and build an entire product ecosystem from the ground up in a highly regulated domain (government benefits), where accuracy, trust, and clarity are critical.
My mission: create a seamless experience connecting multiple stakeholders across four products — User App, Case Portal (advocate-facing), Admin Portal (internal ops), and Marketing Website.
Core challenge: Users are often under financial, health, and emotional strain — the product needed to feel simple, trustworthy, and low-effort, while the backend involved long timelines, multiple agencies, and significant manual operations.
Defined the Jobs To Be Done
Before any wireframes, I established the three core JTBDs that would drive every design decision across the ecosystem.
"Help me navigate a complex benefits system so I can get financial stability without having to understand the bureaucracy behind it."
"Help me manage multiple cases efficiently so I can focus on winning cases instead of administrative tasks."
"Provide a scalable, repeatable process powered by AI and expert oversight."
Foundational discovery
Goal: build understanding from zero in a domain with no existing product, IA, or architecture. I ran stakeholder interviews with advocates, attorneys, SDRs, operations, and founders — mapping the full SSD lifecycle end-to-end.
Key insight 1: ~70% of user confusion occurred before they even spoke with an advocate — pointing to a critical onboarding gap.
Key insight 2: Advocates spent ~40–60% of their time on manual updates → high opportunity for automation. This shaped the Jake AI feed strategy.
Design spike — de-risking before committing to architecture
Before locking in any architecture, I ran a rapid exploration phase to validate the biggest unknowns: case progress models, how Jake AI should communicate, and Case Portal workflows.
- Case tracking models for multi-phase timelines
- Jake: conversational vs. feed-based interaction models
- Case Portal workflow variations
- Technical feasibility with engineering
- Selected non-chat, feed-based model for Jake AI
- Modular architecture that can evolve without rewrites
- Defined system objects: cases, tasks, records, stages, automated actions
MVP scoping — impact vs effort matrix
Using an impact–effort matrix aligned to the three JTBDs, I led scoping decisions to separate what needed to ship from what could wait.
- User onboarding & benefits eligibility funnel
- Basic dashboard with Jake feed v1
- Advocate case management & intake forms
- Core communications & status visibility
- Simple admin tools
- Fully automated document uploads
- Advanced advocate tooling
- Multi-program eligibility scans
- In-app payments
- Rich analytics & notification center
IA & system architecture
Four interconnected products, each with independent scoping for engineering but a shared design system and interaction language.
Turning a government process into a product experience
Designing the Case Feed required transforming a complex legal workflow into a clear, actionable experience that adapts to each claimant's situation. I developed a structured logic system grounded in real SSA policy, operational constraints, and multi-case workload needs.
Outcome: This framework became the backbone for UI decisions, automation strategy, and roadmap prioritization — enabling the team to build smarter before building faster.
Design system & documentation
I owned the foundation of Turnout's component library — optimized for speed, clarity, and WCAG compliance. Reduced UI inconsistencies by 80% and improved build velocity by enabling reusable components.
- Grid, spacing, layout tokens
- Typography and elevation scales (Besley + IBM Plex Sans)
- Iconography system
- Forms, cards, buttons — all variants
- Jake AI components
- Multi-product shared primitives
- Architecture diagrams
- Feed logic tables
- Component usage guidelines
- Interaction patterns
- Accessibility ruleset (WCAG)
Handoff, implementation & QA
I stayed in the build throughout — attending dev standups, conducting design QA on staging, logging UX bugs with severity levels, and ensuring parity across all surfaces.
Impact highlights
Post-launch metrics tracked across funnel conversion, dashboard engagement, Jake feed adoption, advocate task time reduction, and user satisfaction (NPS, CSAT).
Post-launch: Used real data to refine the eligibility funnel, dashboard layout, Jake feed logic, advocate workflows, and navigational clarity — closing the loop between design decisions and measured outcomes.