Optimizing conversion with A/B testing at Sonnet

Sonnet was the first home and auto insurance provider in Canada to allow customers to quote and purchase policies entirely online. As a first mover, Sonnet recognized that staying ahead required not just innovation but continuous refinement of its quoting and account management experiences.

I was hired as UX Manager to lead the UX strategy, guide the team, and implement a systematic approach to optimizing the quoting journey, with a focus on usability, conversion, and customer trust.

The Challenge

Insurance quoting is inherently complex. Users face lengthy forms, unfamiliar terminology, and high-stakes decisions. While Sonnet’s digital experience was functional, analytics and customer feedback showed opportunities to reduce friction, improve clarity, and increase confidence.

The challenge was to establish a repeatable, data-driven process for continuously improving the quoting and account management flows, balancing business objectives, user needs, and regulatory requirements while leading the team in executing the vision.

Approach

I led the UX strategy and team through a structured, iterative process:

Additionally, I aligned with product and engineering stakeholders to ensure priorities matched business goals, and established workflows for design and testing to improve team efficiency and consistency. This approach ensured all work was user-centered, measurable, and aligned with strategic objectives.

Design Objectives

From discovery, three key themes emerged: users lacked clarity, internal teams relied on assumptions, and the quoting flow felt complex. I developed three objectives to guide all improvements:

These objectives became a north star, guiding all team work across research, design, and experimentation.

Selected Experiments & Insights

To validate improvements and measure impact, I designed and led a series of A/B tests across the quoting and account flows. Some representative experiments include:

Home + Auto Bundle CTA

Goal & Hypothesis

Increase adoption of bundled policies by making the CTA more prominent.

Results

• Conversion increased 90.22%

Learnings

• Clear, prominent CTAs drive major conversion gains
• Messaging and placement matter as much as design
• Data-driven testing outperforms assumptions

Save Quote CTA

Goal & Hypothesis

Encourage users to return and complete quotes by improving visibility of “Save Quote.”

Results

• Usage increased, improving conversion by 6.37%
• Performance varied between desktop and mobile

Learnings

• Users value flexibility to pause and return
• Desktop and mobile behaviors can differ, requiring tailored approaches
• Small visibility changes can meaningfully impact conversion

Simplified Quote Selection

Goal & Hypothesis

Simplify decision-making by presenting a single recommended quote option.

Results

• Conversion increased 6.4%

Learnings

• Simplifying choices reduces cognitive load and increases confidence
• Clear recommendations help users make decisions faster

Form Design Enhancements

Goal & Hypothesis

Improve form usability and reduce abandonment rates by simplifying layout and surfacing guidance.

Changes Tested / Validated

• Single-column layout (validated in prototype and usability sessions, not live A/B tested)
• Inline help text and tooltips (tested live where possible)
• Optimized field types
• Step progress indicators
• Inline validation
• Mobile-specific enhancements

Results / Observations

• Prototype testing and usability sessions showed improved scanning and reduced cognitive load with the single-column layout
• Completion rates and confidence improved in usability sessions
• Live-tested enhancements contributed to 3–4% conversion uplift

Learnings

• Layout and guidance improvements reduce friction and increase completion
• Prototype testing can uncover usability wins before live deployment
• Mobile-specific adjustments are critical for consistency

Outcomes

Results Snapshot

• +90% bundle conversion lift
• +6% overall conversion lift (Save Quote)
• +6% conversion lift (quote options)
• +3–4% conversion lift (form enhancements)
• Continuous optimization framework embedded

Impact

Opportunities Identified

While the experiments delivered measurable improvements, research and user insights revealed additional opportunities to further optimize the quoting experience. These insights highlight areas for improvement and refinement:

• Progressive disclosure to simplify long flows
• Natural language inputs for easier comprehension
• Prefill and smart defaults to reduce effort
• Enhanced mobile optimizations (sticky buttons, adaptive keyboards)
• Trust-building microcopy at sensitive fields
• Conditional recommendations and personalization

Reflections

This experience reinforced the value of strategic leadership paired with structured experimentation. By defining clear objectives, embedding testing into the workflow, and guiding the team, we delivered measurable business impact.

Key learnings:

This work demonstrates that strategic UX leadership, combined with A/B testing, can significantly improve user experience and business outcomes.