Newfold Digital Impact Dashboard

Breaking Vendor Lock-in: The $3M/Month Shift

Project: 'DESTINY' Infrastructure Overhaul

Role: Product Manager, Platform Strategy

The Ecosystem Context

• Newfold operates multiple hosting brands serving millions of users.
• Administrative backend depended on cPanel.
• Price hikes and roadmap constraints threatened margins.

What I Noticed

It wasn't just that our operational costs were high; it was structural. When I analyzed our P&L, I realized a massive chunk of revenue was effectively a pass-through to cPanel. We were trapped in vendor lock-in, paying a premium to have less control over our own customer experience.

Strategic Analysis (Build vs. Buy)

• Option A: Continue scaling licensing fees with growth.
• Option B: Build 'DESTINY' in-house.
• Build cost amortized within 14 months.
• Decision: build for long-term unit economics.

The Path I Chose

I defined the 0-to-1 vision for 'DESTINY' to replace cPanel entirely. I prioritized a roadmap focused on parity-plus, matching core features while adding AI-driven tools. The goal was to convert a variable cost into a fixed asset and secure margins.

Execution Roadmap

• Phase 1: Parity for essential cPanel functions.
• Phase 2: Innovation with AI site builder + staging.
• Phase 3: Phased migration by cohort risk.

The Human Outcome

The projection is massive: $3M monthly savings straight to the bottom line. But the operational win is independence. We now control our product experience and can ship updates on our timeline, not someone else's.

Future Vision

• Launch vertical-specific hosting experiences.
• White-label DESTINY for other hosting providers.
• Turn a cost center into a revenue generator.

Scaling Empathy with AI: The Agentic RAG System

Project: Agentic RAG Support Automation

Role: Product Manager, AI & Automation

The Support Landscape

• 50,000+ tickets per month.
• High CPC due to human agents handling repetitive tasks.
• Long wait times drove customer dissatisfaction.

What I Noticed

It wasn't just that our support costs were high; it was that customers felt ignored. We were paying millions to provide a slow experience that frustrated both agents and customers, causing churn on the human side and the business side.

Technical Architecture

• LLM + RAG to read internal knowledge base.
• MCP gives AI hands to execute API calls.
• Human-in-the-loop handoffs for low sentiment scores.

The Path I Chose

We needed a technical agent, not another chatbot. I architected an Agentic RAG system and gave it hands via MCP so it could reset permissions or clear cache. Actions were grounded in documentation to avoid hallucinations.

Implementation Strategy

• Targeted high-friction, low-empathy tickets first.
• Rolled out to 5% traffic to train retrieval accuracy.
• Feedback loop updated the knowledge base weekly.

The Human Outcome

We saved $175k each month, but the real win was trust recovery. A 3 AM outage could be solved in seconds, driving a 15% churn reduction. We automated 60% of tickets and freed human agents for complex problems.

Scalability

• Expanding to billing and sales inquiries.
• RAG system accelerates human agent onboarding by 40%.

Pricing Psychology: Driving Premium Adoption

Project: Hosting Tier Redefinition

Role: Product Manager, Growth

Current Market Positioning

• Commoditized market competing on price and specs.
• Customers don’t know what “Unmetered Bandwidth” means.
• Users defaulted to the cheapest plan, hurting ARPU.

What I Noticed

We were selling server specs while customers were buying business dreams. Non-technical users froze at checkout, defaulting to the cheapest plan because we spoke a different language than they did.

User Research Insights

• Outcomes matter: “My site should be fast.”
• Users pay for peace of mind but don’t know where it lives.
• Specs needed translation into outcomes.

The Path I Chose

I shifted from specification-first to persona-first, redefining tiers as Low vs High Performance and labeling for user types. The checkout highlighted outcomes like speed, security, and growth so users could self-select easily.

A/B Testing & Iteration

• Test A: Spec-heavy comparison table.
• Test B: Persona table with “Recommended for You.”
• Result: 20% immediate lift in premium selection.
• Refined copy to emphasize “High Performance.”

The Human Outcome

We saw a 28% increase in premium sign-ups. Customers bought better-fit plans, reducing buyer's remorse and improving long-term value. Clarity became the conversion tool.

Key Learnings

• Clarity beats choice overload.
• Higher LTV when premium adoption happens Day 1.

The End of the "Blank Page": GenAI Site Builder

Project: AI-Driven Website Builder

Role: Product Manager, AI Initiatives

The Activation Problem

• Users bought hosting but dropped at “Open Builder.”
• Blank canvas created anxiety around design and content.
• Activation drop-off killed publish conversion.

What I Noticed

Users opened the builder, stared at a blank screen, and quit. They had intent but lacked skill. We were selling the canvas without teaching them how to paint.

The Psychology of Creation

• Creator mode is high friction.
• Editor mode is low friction.
• Goal: move users to editing a draft.

The Path I Chose

I led the strategy for an AI-driven builder that asked users to describe instead of drag. We generated an instant first draft of layout, copy, and imagery so users could edit rather than create from scratch.

Product Flow

• Answer 3 questions: name, industry, vibe.
• AI generates a full site in under 60 seconds.
• Magic Editor lets users tweak before publish.

The Human Outcome

Time-to-publish plummeted. Users felt empowered, and publication conversion improved by 20%. We removed the creative barrier and helped more small businesses go live.

Broader Impact

• Published sites churn 4x less than empty domains.
• AI Builder used to upsell AI SEO and AI Content add-ons.