Back
Fullscript

Modernizing Auto Refill Portal

Redesigned Fullscript's highest-leverage retention surface into a clear, intuitive experience that patients could navigate without support.

Role

Senior Product Designer

Timeline

3 weeks

Team

2 FEDs, 1 BED, 1 Designer

Scope

1st to 4th Order Patient Conversion

Product DesignWebRetention

My Role

  • Led discovery research to map existing IA and identify friction points
  • Facilitated concept reviews with stakeholders to align on scope
  • Owned data collection and analysis to validate design decisions
  • Partnered closely with engineering on implementation approach
  • Owned Design QA using AI tools and pushed MRs directly to codebase
Old desktop auto-refill UI

Before: Portal

New desktop auto-refill UI

After: Portal

TL;DR

Problem

Patients couldn't answer three basic questions: What's shipping? When? Can I change it?

Approach

Resisted feature creep. Scoped to IA and layout only. Promoted skip/reschedule as alternatives to cancellation.

Outcome

Lifted activations 11%, cut cancellations 3%, halved support tickets, generated ~$600K in incremental revenue.

0

Incremental Revenue

From activation lift alone

Context & Problem

Business Context

Auto-refill is one of Fullscript's highest-leverage retention surfaces. The web portal had accumulated years of drift: dated visuals, fragmented IA, and patterns that made routine edits feel foreign.

User Context

US patients managing supplement subscriptions who needed to view, modify, or pause upcoming orders without confusion.

Constraints & Complexity

No net-new features, foundation work onlyStaged rollout behind flipper for validationHad to support future experimentation

Goals & Success Metrics

Objective

Deliver clarity and reduced complexity, creating a foundation the team could build on for future experiments.

Success Criteria

  • Patients can answer 'What, When, Can I change it?' without digging
  • Reduce support ticket volume
  • Maintain or improve activation and retention metrics

11%

Activation Lift Target

50%

Ticket Reduction

0%

Net Cancellation Increase

Approach

1

Audit

Mapped existing IA, identified where 'view upcoming' and 'edit subscription' flows were scattered.

2

Scoping

Resisted feature creep. Deliberately narrow MVP focused on IA and layout only.

3

IA Rework

Made 'view upcoming' and 'edit subscription' the dominant entry points.

4

Friction Reduction

Promoted reschedule and skip into primary view as alternatives to cancellation.

5

Staged Rollout

Shipped behind flipper with slow rollout to validate before full ramp.

Key Decisions

Critical choices that shaped the final solution

Decision

Scoped to IA and layout only. Visual refresh could come later once foundation was proven.

Tradeoff

Faster ship, cleaner validation, but portal still looked dated.

Decision

Promoted reschedule and skip into primary view as first-class actions.

Tradeoff

Required rethinking action hierarchy, but directly addressed cancellation driver.

Decision

Shipped behind feature flipper with percentage-based rollout, monitoring at each stage.

Tradeoff

Slower deployment but significantly de-risked launch.

The Solution

A restructured portal that answers patients' core questions at a glance: what's shipping, when it's arriving, and how to change it. Reschedule and skip elevated as primary actions to reduce unnecessary cancellations.

Old order detail page

Before: Order Detail

New order detail page

After: Order Detail

Old date picker UI

Before: Date Picker

New date picker with calendar

After: Date Picker

Outcomes & Impact

0

Activations

~2,700 to ~3,000/day from this page

0

Cancellations

From surfacing alternatives

0

CS Tickets

~200/week to ~100/week

0

Revenue Impact

From activation lift alone

Business Impact

The redesign delivered a clean foundation for follow-on experiments: adding products to auto-refills, surfacing recommendations, streamlining management.

User Impact

Patients can now navigate subscriptions without support, with a mental model that matches how they think about recurring orders.

Learnings & Reflection

What Worked

Resisting feature creep paid off. Tight scope meant cleaner attribution and faster validation.

What Worked

Promoting alternatives to cancellation directly moved the cancellation metric.

Insight

Sometimes the best redesign is invisible. Users don't notice good IA, they just stop struggling.

Insight

Building a foundation for future experimentation is as valuable as immediate metric wins.