Commercial Vehicle (CV) Insurance Estimate Calculator
A multi-phase product evolution, from initial concept to a revenue-validated quote system.
An iterative, 3.25 year arc, Design-led optimization across revenue-critical insurance quote experiences. Built to reduce abandonment, protect lead quality, and improve estimate trust so more users convert into agent-ready opportunities and bound policies.
I originated the concept, drawing initial inspiration from The Zebra, and led the product from CV0 through CV3, defining both the UX strategy and system behavior across iterations. The work progressed through a sequence of controlled changes. First removing friction to unlock completion, then reintroducing structured signal to improve lead traceability and pricing confidence. The CV2 A/B test marked the inflection point, increasing completions 14x and raising estimated value from $27K to $341K, establishing a high-performing baseline that subsequent iterations were designed to protect and extend.
Key Deliverables
Note: Sensitive details redacted. If any public links have expired, originals are available on request. Designs were produced in Adobe XD and Figma, reflecting a legacy-to-modern tooling transition. When inside a flow, use arrow keys at the bottom to browse.
System evolution
CV0 established the initial quote flow structure. CV1 expanded the system but introduced heavy upfront friction through early data capture and a 5-step process. CV2 removed early PII, reduced steps, and introduced A/B testing, reversing the primary failure mode and unlocking completion. CV3 built on that baseline by reintroducing controlled data capture and adding an accuracy framework, without compromising conversion performance.
At a glance
Conversion impact
The CV2 test reversed the funnel’s primary failure mode, increasing completions 14x and raising estimated value from $27K to $341K.
Friction removal
Early PII, excess steps, and qualification barriers were removed to align the experience with user intent and unlock forward momentum.
Revenue system signal
Introduced a measurement layer connecting estimates to final pricing, enabling quote integrity tracking and downstream revenue visibility.
Context
Quotes are a revenue product.
Quote experiences sit at the intersection of acquisition, qualification, and conversion. The work below shows a deliberate progression: first remove friction to increase completions, then add controlled signal to improve lead traceability and quote trust.
Iteration timeline
Phase 1. Conversion lift (A/B test) CV2
Addressed a major abandonment problem right after the first step by reducing friction, cutting steps, and improving clarity and SEO alignment to capture more high-intent demand.
View Phase 1Phase 2. Quote integrity and revenue confidence. CV3
Added a scalable measurement layer to close the expectation gap between digital estimates and agent pricing, while protecting conversion performance with a phased data strategy.
View Phase 2Phase 1
A/B testing a lower-friction quote flow
This phase reversed the primary failure mode of the product. Drop-off was treated as a revenue leak, and the flow was restructured to remove any input not required to generate an estimate, allowing users to move forward before committing personal data.
Primary KPI
Reduce drop-off after step 1, where abandonment was approximately 77%.
Secondary KPIs
Completion rate, time to first step completion, engagement by screen, calls initiated, and SEO alignment.
Test design
Optimizely A/B test with 50/50 split, run to statistical confidence.
Phase 2
From conversion to quote integrity
Improving conversion introduced a new constraint: maintaining lead quality and pricing accuracy without reintroducing early-stage friction. This phase introduced a measurement system and selective data capture strategy to balance both.
Risk-aware rollout
Changes introduced incrementally to avoid degrading a high-performing funnel.
Minimal signal first
Introduced a single lightweight data point to connect sessions to outcomes without impacting completion.
Accuracy as a system metric
Measured estimate-to-final variance to monitor trust, detect gaps, and support revenue reporting.
Scale
Built once, scaled across products and digital surfaces.
The framework established here was designed as a reusable system. The same sequencing logic, friction reduction patterns, and measurement model were applied across additional quote flows, including personal auto, creating consistency while scaling revenue impact.
Disclosure
Key decisions, supported by selectively presented data.
For confidentiality, screenshots and metrics are selectively abstracted. Full details and deeper system walkthroughs are available in interviews.