Context
By 2025, Legalstart’s acquisition engine needed a strategic evolution. As Head of Growth Design, I took ownership of the pre-payment funnels—the core revenue driver of the company. The challenge was twofold: migrating complex legacy flows to a new identity while simultaneously architecting an experimentation engine capable of tackling churn and maximizing upsell conversion in a highly competitive market.
Project metadata
• Role : Head of Growth Design & Conversion (Squad Lead).
• Squad : 1 Product manager, 2 front-end developers.
• The North Star : Maximize CVR & upsell revenue while reducing churn.
• The result : +1.03 pts increase in overall conversion rate (CVR) across all funnels.
Stakeholder management & squad leardership
Transitioning into this hybrid role, I acted as both the strategic lead and the operational bridge between design, product, and engineering
• Agile governance : I spearheaded the squad’s rituals, setting monthly north star metrics and prioritizing a high-impact backlog. I ensured a high cadence of experimentation (1-2 A/B tests launched per sprint).
• Cross-functional alignment : I negotiated roadmaps to balance long-term UX integrity with aggressive short-term growth goals, managing the developers' velocity and the PM's strategic requirements.
• Data-informed prioritization : I transformed the team's culture into a "test & learn" powerhouse, where every design decision was filtered through potential ROI and technical feasibility.
High-velocity strategy: "Test or kill"
I established a "fake it until you make it" culture to prioritize speed of learning over pixel perfection.
• Dirty iterations : We launched unpolished MVPs to validate behavioral hypotheses in real-time. If a test failed, it was killed immediately. If it won, it was instantly scaled across the entire ecosystem.
• The learning loop : This "dirty" approach allowed us to gather real-market data in days rather than months, drastically reducing our time-to-market.
• Dirty iterations : We launched unpolished MVPs to validate behavioral hypotheses in real-time. If a test failed, it was killed immediately. If it won, it was instantly scaled across the entire ecosystem.
• The learning loop : This "dirty" approach allowed us to gather real-market data in days rather than months, drastically reducing our time-to-market.
AI-powered operations: Scaling quality
Managing tens of thousands of potential funnel permutations (questions x products x upsells) required a shift from manual work to AI-first operations.
• Automated QA with Claude : We implemented an AI-driven system to audit our code and logic. What used to take days of manual checking was reduced to minutes, ensuring that every "dirty" iteration remained technically sound across all paths.
• Autonomous health checks : To safeguard our revenue, I spearheaded the creation of an AI-automated auditing tool. It runs multiple daily health checks, instantly alerting the squad to any performance regression or technical glitch.
• Autonomous health checks : To safeguard our revenue, I spearheaded the creation of an AI-automated auditing tool. It runs multiple daily health checks, instantly alerting the squad to any performance regression or technical glitch.
The business impact
By merging aggressive growth design with AI-driven velocity and cross-functional leadership, we transformed the acquisition flow into a high-performing profit center.
• Conversion milestone : Achieved a +1.03% point uplift in total CVR, directly impacting the company's bottom line.
• Upsell optimization : Significantly increased the attachment rates for high-margin subscriptions through optimized, psychologically-triggered flows.
• Operational excellence : Automated the most complex parts of the QA process, allowing the squad to focus on high-level strategy and creative experimentation.
.User flow pre-payment for modification
