Project Overview

Unified marketing engineering and operational (Ops) processes to maximize results from large-scale PPL projects collaborating with 1,100+ influencers annually. Technically automated the fragmented influencer selection, contract creation, and performance tracking processes to minimize operational resources and establish a company-wide data-driven marketing system.

The Challenge

• No Operational Process: Influencer recruitment through contract execution relied entirely on manual work, creating bottlenecks during large campaigns
• Lack of Data-Driven Metrics: Beyond simple impression-based evaluation, no framework existed to measure actual contribution to brand search volume and lead acquisition
• Tech-Ops Disconnect: Ad data and marketing execution processes weren't connected, preventing real-time performance analysis and rapid media mix adjustments

Strategy / Solution

1. Performance Scoring System: Developed a model combining channel influence, historical lead efficiency, and variability (MAD) to quantify performance weights (Branding/DB scores).
2. Marketing Ops Standardization & Automation: Built an E2E operations system including selection criteria, rate negotiation, and auto-generated e-signatures and contracts to systematize the entire workflow.
3. Metrics Data Pipeline Development: Classified traffic categories through GA analysis and developed ML-based attribution models to quantify specific channel impacts, establishing objective KPI management.
4. Hyper-Automation Implementation: Self-developed a bulk cold email program with A/B testing and read tracking, reducing outreach time by 98% and maximizing operational productivity.
5. Unified Analytics & Dashboard: Consolidated fragmented data into BigQuery and developed real-time dashboards for continuous performance monitoring without weekly reports.

"The essence of marketing engineering is breaking operational limits through technology and turning every process into a data-provable system."

Execution

• Tech Stack: Python, SQL, BigQuery, Sora API, PHP, GA4/GTM
• Key Activities: Managing 1,198 influencer pool, developing MAD calculation logic, establishing long-form and live content guidelines
• Collaboration Scope: Served as a 'bridge role' spanning PM, design, and development — executing all processes in-house without outsourcing

Results