AI-Powered
Email Health

During a 1-sprint R&D, the goal was to help users understand and improve their email deliverability in the Keap/Thryv app. Previously, users had little insight into how their emails were performing, making it hard to catch issues before campaigns got suspended.

We ran rapid research to uncover pain points, then designed and tested a new Email Health experience that breaks performance into three actionable sections (Deliverability, Engagement, and Complaints), giving users clear guidance to protect their sender reputation and build long-term success.

A baseline usability test scored the existing experience at a 48.8 SUS, a failing grade, well below the "excellent" target. To understand why, I ran 30-minute in-depth interviews with six users. Three usability issues surfaced clearly.

Most tools manage email health proactively: bounce suppression, soft-bounce thresholds, and email verification. Keap/Thryv lacked these safeguards: no automatic suppression, unclear thresholds, and limited verification, leaving users more exposed to deliverability issues.

A cross-functional team (designers, a content designer, researcher, PM, and customer-success manager) aligned on the core challenges and explored early solutions, focused on empowering users with insight, guidance, and control.

Auditing the current experience exposed the root problems: metrics with no meaning ("average engagement (days)"), weak visual hierarchy, misleading icons, no trend visibility, and no next step when something went wrong. I scoped an MVP around a clear health score and rebuilt the widget and report around it.

Before · current widget

Confusing metrics, unclear guidance, poor visual hierarchy.

After · redesigned widget

Clear health score, organized metrics, actionable insights.

• Clear health score (0–100) for at-a-glance understanding
• Organized categories: Deliverability, Engagement, Complaints
• Visual progress indicators for each metric
• Actionable buttons for direct problem resolution
• Eliminated the confusing engagement-time metric

Before · current report

Confusing 6.4% metric, text-heavy explanations, buried action items.

After · AI-powered insights

Clear deliverability score, AI recommendations, trend visualizations.

• AI-powered insights with personalized recommendations
• Clear deliverability score (0–100) prominently displayed
• Trend visualizations for bounce rates over time
• Detailed bounce-rate breakdown (soft vs. hard)
• Visual hierarchy that highlights the most critical metrics

A walkthrough of the redesigned experience living inside the existing Keap CRM, surfacing the health score, category breakdowns, and AI recommendations without pulling users out of their daily workflow.

To push the experience further, I designed three AI-powered interface concepts showing how intelligence can provide contextual assistance and streamline workflows without pulling users out of their day-to-day.

The prototype demonstrates how these improvements live inside the existing Keap CRM workflow, surfacing actionable insights without disrupting daily tasks. Each report pairs data visualizations with AI-powered recommendations so users can spot and fix issues in place.

• Email-health monitoring within contact workflows
• Contextual insights during campaign management
• Real-time alerts for deliverability issues
• Automated list-cleanup suggestions and in-context AI recommendations

After the 1-sprint R&D, the next step is refining the prototype with the Reporting Center and Email Deliverability teams: assessing feasibility, aligning on priorities, and defining technical requirements for implementation.