Big Data Support

SBG Demand DS• Developed a quantitative understanding of SBG advertisers’ contactability through communicating definitions with business partners, researching methodologies on past projects and using data analytics techniques such as SQL & Python to identify over 800,000 uncontactable advertisers and propose strategies to drive SBG’s north star metric, incremental revenue. • Analyzed the uncontactable advertisers’ profiles, such as their spending patterns, geographical distribution and user activities on ad platforms using SQL, and applied Python to visualize the group’s distinct features to further help business partners to understand the reasons for not able to contact the advertisers through marketing channels.• Designed a machine learning model utilizing logistic regression and random forest to identify the influential factors on advertisers’ contactability. The model recognized variables such as product interfaced-based entry-point exposures, past pitch call experience and the number of marketing channels reached as the positive drivers of contactability, achieving a model accuracy of 0.96. This analysis overturned the long-held belief that the marketing effort is the only driver for advertiser contactability, and instead recommended to the marketing team to focus holistically on multiple factors to engage advertisers.• Presented 5 actionable recommendations such as increasing proactive outreach frequency and expanding the advertisers’ product-based entry points exposures to assist Meta’s marketing team to engage uncontactable advertisers. The proposed strategies generated a $53 million revenue opportunity for SBG.

Partner Business Dashboards Automation Using Tableau & SQL