Driven by customer obsession. Proven by growth metrics.

Context

Zendesk reorganized around a new customer segmentation strategy in 2025. A new Digital segment, companies with fewer than 50 employees, was defined. These primarily self-service customers made up a segment representing tens of millions in annual bookings, but the segment had been contracting over multiple consecutive quarters. As the product leader responsible for the Product Led Growth strategy, I set out to reverse the decline.

Problem

Trial signups were healthy, but activation and conversion were declining. New customers entered a blank, configuration-heavy product and dropped off before reaching Zendesk's core “aha moment”, resolving a customer inquiry using the platform. The existing trial and onboarding experience relied on guided tours of features, not on getting users to feel immediate value.

Approach

Outcome

Lesson

AI alone doesn't move metrics — solving a concrete customer problem does. The real work was understanding why trialists were failing, then using AI as the lever to eliminate that friction. We weren't rewarded for using AI. We were rewarded for balancing speed, experimentation, and a genuine customer insight.

Context

As a Principal PM at Amazon Music, I led three growth product teams responsible for global acquisition and retention across 60M+ customers. Amazon Music sat inside one of the world's most powerful distribution ecosystems, yet still faced a significant awareness problem.

Problem

Two gaps, same root cause. Tens of millions of Prime members didn't know Amazon Music was included in their membership. And existing customers weren't upgrading to Amazon Music Unlimited or understanding why it was worth paying for over Spotify or Apple Music. The discovery problem wasn't inside the app. It was before customers ever opened it.

Approach

• →Meet customers where they already are. My biggest bet: embedding an Amazon Music player directly into the amazon.com commerce experience. A customer searching for Billie Eilish merchandise would discover they already had access to Amazon Music and could start listening immediately. I wrote the six-page PRFAQ, secured alignment across Amazon Music, Retail, and Prime leadership, and proposed it as a two-way door decision built for rapid experimentation. It launched after I left Amazon. It's still live on amazon.com today. In Amazon's culture, that means it earned its place.
• →Solve the cold-start problem. We introduced an Artist Taste Collection in the first-time user experience: a short quiz capturing favorite artists that immediately personalized the Home experience with relevant recommendations. Getting personalization right in session one is often the difference between a retained user and a churned one.
• →Remove friction at the upgrade moment. Inspired by Amazon's 1-Click Buy Now, I brought that same pattern into Amazon Music for AMU upgrades, requiring close collaboration with security, compliance, and finance to ensure we weren't eroding customer trust.

Outcome

Lesson

The best growth lever isn't always inside your product. Sometimes the highest-leverage move is finding where your customers already are — and making it frictionless to discover you there.

Context

A large percentage of LendingClub borrowers said they wanted loans to pay off credit card debt, but after receiving funds, many weren't actually paying down their balances. The credit and risk teams proposed a balance transfer product where LendingClub would pay creditors directly instead of depositing cash.

Problem

Two teams with fundamentally different incentives had to build one product together. Risk wanted maximum verification and friction to ensure debt paydown and minimize default risk. Design and Growth wanted a simple, intuitive experience that didn't kill funnel conversion. My job was to find the path through.

Approach

• →Brought both sides together around a shared customer outcome of improving borrowers' financial health, rather than debating features
• →Proposed an iterative rollout: MVP had users manually enter creditor info; risk teams maintained stronger validation controls; design teams accepted initial friction in exchange for fast learning
• →Instrumented the funnel heavily across awareness, offer selection, creditor entry, funding conversion, and repayment behavior
• →Drove subsequent iterations: pre-populating creditor info, simplifying the balance transfer flow, improving offer display relative to traditional cash loans, reducing cognitive overload during loan selection

Outcome

Lesson

Successful product leadership often isn't about choosing between customer experience and risk management — it's about creating alignment around shared outcomes, then using iterative experimentation to build trust across teams.