Turning behavioural data into measurable conversion gains

+8.2% lift on a page that already converted

Role

UX/UI Designer

Team

Solo

Duration

1 month

a cell phone on a table
Overview

Redesigning the carrier screening landing page using AI-powered research, Figma prototyping, and Claude Code — turning behavioural data into measurable conversion gains without disrupting what was already working.


Problem and solution

The carrier screening page was the primary entry point for a high-value product category. At 50.4% conversion it wasn't failing — but drop-off data in Amplitude surfaced real friction: users were entering with intent but leaving before engaging with any CTA.

Rather than redesigning from intuition, I ran an AI-assisted discovery sprint in Amplitude to surface behavioural patterns, generated hi-fi wireframe concepts using Claude and Figma MCP, validated with internal stakeholders, and shipped a refined build directly via Claude Code in VS Code — keeping the entire process lean, fast, and evidence-backed.

Overview

Redesigning the carrier screening landing page using AI-powered research, Figma prototyping, and Claude Code — turning behavioural data into measurable conversion gains without disrupting what was already working.


Problem and solution

The carrier screening page was the primary entry point for a high-value product category. At 50.4% conversion it wasn't failing — but drop-off data in Amplitude surfaced real friction: users were entering with intent but leaving before engaging with any CTA.

Rather than redesigning from intuition, I ran an AI-assisted discovery sprint in Amplitude to surface behavioural patterns, generated hi-fi wireframe concepts using Claude and Figma MCP, validated with internal stakeholders, and shipped a refined build directly via Claude Code in VS Code — keeping the entire process lean, fast, and evidence-backed.

a cell phone on a white block
two cell phones on a gray surface

The page was already converting at 50.4% — not broken, but leaving measurable intent on the table. The update shifted that to 58.6%, an 8.2 percentage point lift on a comparable visitor sample of over 2,200 sessions. That translates to 81 more users clicking through per window — without touching ad spend or traffic volume. The drop-off rate fell by the same margin, from 49.6% down to 41.4%. The consistency between the two figures matters: the same users who stopped dropping off became the users who clicked. That's a behavioural shift, not a statistical coincidence.

a cell phone with a yellow rectangular screen

Business impact

Conversion gains on a page that already works are harder to find than fixing something broken. There's no obvious error to remove — just friction subtle enough that users don't notice it, but measurable enough that Amplitude does.


The 8.2 point lift came from closing that gap: tightening hierarchy, clarifying the next action, and removing the hesitation points the data flagged. Not a redesign — a targeted intervention. The AI-assisted process compressed what would typically be a multi-week cycle into something significantly faster.


Research, wireframing, and build happened in sequence without handoffs — which meant decisions stayed connected to the evidence that motivated them. The result is a page that converts more of the intent that was already there. No additional traffic. No increased spend. Just more of the right users taking the next step.

Other projects

Nathan Rego

Copyright 2025 by Nathan Rego

Nathan Rego

Copyright 2025 by Nathan Rego

Nathan Rego

Copyright 2025 by Nathan Rego