Designing clarity into a legally complex compliance platform
Termageddon is a compliance platform that generates and maintains legally defensible privacy policies across websites and applications. When I joined, the core questionnaire was long, legally dense, and heavily dependent on support from the founders. I redesigned the login, dashboard, and multi-step questionnaire into a clearer, section-based workflow and embedded a context-aware AI companion to guide completion. The redesign preserved legal rigor while meaningfully reducing support burden and moving the product closer to true self-serve.
Impact
What drove our strategy...
Goals
- Make legally complex compliance workflows clear and structured
- Reduce reliance on founder-led support through embedded guidance
- Preserve legal accuracy while improving self-serve completion
User Frustrations
- Long, required questionnaires without contextual explanation
- Uncertainty about how answers affect legal defensibility
- Legal terminology that was difficult for non-experts to interpret
- Dependence on support to complete high-stakes workflows
Personas
- Small Business Owners – Responsible for staying compliant but not fluent in legal terminology. They needed clarity, reassurance, and structured guidance.
- Law Firms and Agency Partners – Managing compliance across multiple clients. They required reliable workflows that preserved legal rigor without adding unnecessary friction.
Where do we start?
User Flows
The legacy product showed three core issues:
- Structural inconsistency across login, dashboard, and questionnaire
- Long, required workflows without contextual explanation
- High support dependency for questionnaire completion
The questionnaire was the primary friction point. Users were required to translate their business practices into compliance terminology, with little explanation of how their answers affected legal outcomes. Validation was rigid but not instructive, leading many to stall, guess, or seek reassurance. Because compliance accuracy could not be compromised, the challenge was not to remove complexity, but to structure it.
Auditing the Existing Workflow
Our first step was analyzing the current experience end to end. The system was feature-rich but structurally dense, with long questionnaire forms, layered navigation, and limited inline explanation. Validation and incomplete states were surfaced late, increasing uncertainty and support dependency. The redesign would focus on clarifying hierarchy, guiding completion progressively, and making legal rigor more navigable.
User Research
Observation
Support requests clustered around the questionnaire workflow, particularly where users were required to interpret legal terminology without context.
Recommendation
- Add contextual explanation for required fields
- Clarify how answers influence compliance outputs
- Reduce reliance on external documentation
"I just want to know if I'm answering this correctly. I don't want to mess something up legally."
"We keep answering the same clarification emails. The product should explain this better."
"I don't know which laws apply to me. I just want to make sure I'm covered."
"I can get through it, but it takes a lot of second guessing."
Observation
Different user types approached the system with varying levels of expertise, but the workflow treated them uniformly.
Recommendation
- Structure flows intentionally around user roles
- Support delegation and partner-based completion
- Preserve legal rigor while reducing cognitive load
"I understand the questions, but I'm not sure what counts for my specific business."
What did we
learn?
Legal complexity cannot be removed, but it can be structured
The questionnaire required exhaustive, legally precise inputs. The issue was not the number of fields, but the lack of hierarchy and explanation. Structuring questions into clearer sections and surfacing why inputs mattered reduced cognitive overload without compromising accuracy.
Uncertainty drives support more than interface friction
Users were rarely confused about navigation. They were unsure whether their answers were legally correct. Providing contextual guidance and embedded assistance reduced anxiety and prevented repetitive clarification requests.
Scalability requires embedded support, not better documentation
External help articles were not enough. High-stakes workflows require support inside the experience. Integrating a context-aware AI companion transformed the product from founder-assisted to meaningfully self-serve.
Wireframes and Iteration
The first major shift was structural.
We reorganized the questionnaire into clearer thematic sections rather than long continuous forms. This allowed users to build a mental model of progress and reduced the feeling of being overwhelmed.
Key improvements:
Structured Sections
We reorganized the questionnaire into clear thematic groupings with visible progress indicators. This reduced overwhelm and helped users understand where they were and what remained.
Review Before Publish
Inline summaries and confirmation states allowed users to review their responses before generating policies. This increased confidence in high-stakes decisions.
Saved Progress
Saved progress and improved return states supported longer workflows and reduced abandonment during multi-step completion.
That insight led to the second major decision...
Introducing the AI Companion
Redesign alone would not eliminate support dependency, so we introduced a context-aware AI companion embedded directly within the questionnaire workflow.
It was aware of:
- The page the user was on
- The fields currently visible
- The broader state of questionnaire completion
Its role was not to replace the workflow, but to act as embedded support.
The companion could:
- Explain why specific questions were required
- Clarify unfamiliar terminology
- Guide users through decision points
- Provide suggestions relevant to the current section
The AI companion absorbed many repetitive clarification requests that previously required direct human intervention. It reduced anxiety, improved confidence, and supported more independent completion.
Visual Design
The visual redesign prioritized clarity and hierarchy by introducing:
- Consistent layout structures across login, dashboard, and questionnaire
- Clearer typographic hierarchy and spacing
- Improved contrast for dense form environments
- Distinct section headers and progress states
- More intentional dashboard organization for managing licenses and entities
The goal was a stable and structured interface that communicates reliability and control.
Dashboard
License Management
Status Filtering and Bulk Actions
License Detail and Questionnaire Sections
Embed Codes and Sharing
Setup Flow
Account Setup Start
Business Information
Policy Configuration
Questionnaire Entry
Questionnaire Progress
Review and Confirm
Install Instructions
Setup Complete
Measuring and Validating Designs
We evaluated the redesign through operational trends and qualitative feedback following release. The focus was on reducing preventable support, improving independent completion, and validating that embedded guidance meaningfully changed user behavior.
Support & Completion
- Reduction in questionnaire-related support requests
- Fewer clarification emails during onboarding
Embedded Guidance
- Regular use of the AI companion within active workflows
- Improved completion without direct founder intervention
Outcome Highlights
- Reduced questionnaire-related support volume
- Increased independent workflow completion
- Strengthened confidence in legally complex decisions
Results and Reflection
The most meaningful gains came from structuring complexity rather than attempting to remove it.
In legally governed systems, clarity is leverage. Translating compliance logic into understandable language and embedding guidance directly into the workflow reduced uncertainty and improved independence.
Support volume proved to be a signal of cognitive overload, not user error.
The next opportunity lies in automation. Detecting data collection behaviors and pre-populating questionnaire inputs would further reduce friction and shift the system from reactive to intelligently assisted.
