Senior Statistical Analysis Manager

• Apr 2022 - Present

Statistical Analysis Manager

• Apr 2012 - Mar 2022

Built predictive models to increase efficiency and effectiveness of donor acquisition and cultivation. Developed reports to automatically track campaign performance and audit donors. • Built an acquisition model to predict likelihood and value of donation. The new selection and asking strategy based on this model achieved 200% lift in response rate and 150% lift in average gift. • Built a retention model to suppress the low value donors from cultivation campaigns. The ROI using the model… Show more Built predictive models to increase efficiency and effectiveness of donor acquisition and cultivation. Developed reports to automatically track campaign performance and audit donors. • Built an acquisition model to predict likelihood and value of donation. The new selection and asking strategy based on this model achieved 200% lift in response rate and 150% lift in average gift. • Built a retention model to suppress the low value donors from cultivation campaigns. The ROI using the model increased by 300%. • Built a lapse prevention model to pinpoint the high risk donors. • Created score card and interactive dashboard to track and forecast campaign performance, make YOY comparison, and monitor budget completion. • Conducted relationship value analysis to audit donors on file and provide insights about the drivers for change. • Created test score card to report performance, significance level, and effect size for marketing campaign tests. It is also a tool to classify, sort, and synthesize multiple tests.

Lead Analyst

• Dec 2011 - Mar 2012

Responsible for evaluating loyalty programs in retail industry. Responsible for evaluating loyalty programs in retail industry.

Sr. Statistician

• Sep 2009 - Nov 2011

Responsible for translating business questions into data analysis questions and identify appropriate methods. Build , apply, and maintain statistical models. Help junior analyst. Create presentations and technical documents • Developed a segmentation system with mixture model in SSAS. Lead a team of four analysts to collect and clean data from census2010, ESRI Tapestry segments, Webtrends, Social Media listening platform, clients database, and demographic data vendors. Developed the… Show more Responsible for translating business questions into data analysis questions and identify appropriate methods. Build , apply, and maintain statistical models. Help junior analyst. Create presentations and technical documents • Developed a segmentation system with mixture model in SSAS. Lead a team of four analysts to collect and clean data from census2010, ESRI Tapestry segments, Webtrends, Social Media listening platform, clients database, and demographic data vendors. Developed the segmentation model in SSAS. Created and presented the profile of each segment. Created the scoring procedure. • Applied multi-level logistic regression as an alternative to traditional logistic regression to predict customer acquisition. This methodology supports a dynamic targeting strategy which automatically adapts to the changes in economics and weather conditions. With this methodology, the model profitability increased by an average of 30% comparing to the traditional logistic regression method. • Applied survival analysis to model the time trend of attrition probability of individual consumers. This methodology made the retention strategy of an energy company more time relevant. This model also helped a client optimize contact time points so that they can reduce contact frequency but keep the volume of responses. • Conducted experimental design to test different contact time points and creative designs for different consumer segments. • Improved the reporting system by automating the reports in SAS, and documenting business rules. Improvements resulted in agreement about the business rules among multiple departments and an approximate 2 day time savings for each report generated. Show less Responsible for translating business questions into data analysis questions and identify appropriate methods. Build , apply, and maintain statistical models. Help junior analyst. Create presentations and technical documents • Developed a segmentation system with mixture model in SSAS. Lead a team of four analysts to collect and clean data from census2010, ESRI Tapestry segments, Webtrends, Social Media listening platform, clients database, and demographic data vendors. Developed the… Show more Responsible for translating business questions into data analysis questions and identify appropriate methods. Build , apply, and maintain statistical models. Help junior analyst. Create presentations and technical documents • Developed a segmentation system with mixture model in SSAS. Lead a team of four analysts to collect and clean data from census2010, ESRI Tapestry segments, Webtrends, Social Media listening platform, clients database, and demographic data vendors. Developed the segmentation model in SSAS. Created and presented the profile of each segment. Created the scoring procedure. • Applied multi-level logistic regression as an alternative to traditional logistic regression to predict customer acquisition. This methodology supports a dynamic targeting strategy which automatically adapts to the changes in economics and weather conditions. With this methodology, the model profitability increased by an average of 30% comparing to the traditional logistic regression method. • Applied survival analysis to model the time trend of attrition probability of individual consumers. This methodology made the retention strategy of an energy company more time relevant. This model also helped a client optimize contact time points so that they can reduce contact frequency but keep the volume of responses. • Conducted experimental design to test different contact time points and creative designs for different consumer segments. • Improved the reporting system by automating the reports in SAS, and documenting business rules. Improvements resulted in agreement about the business rules among multiple departments and an approximate 2 day time savings for each report generated. Show less

Sr. Data Analyst

• Oct 2006 - Sep 2009

Built predictive, segmentation, and marketing mix models, implemented scoring system, and wrote technical documents. Analyzed web navigation behaviors. Reported on Direct Mail campaign performance. • Applied clustering analysis to consumer segmentation. The segmented creative outperformed the general creative by 18% in response rate. • Applied market basket analysis to find out which web pages are frequently viewed together. This analysis led to web site optimization for a client… Show more Built predictive, segmentation, and marketing mix models, implemented scoring system, and wrote technical documents. Analyzed web navigation behaviors. Reported on Direct Mail campaign performance. • Applied clustering analysis to consumer segmentation. The segmented creative outperformed the general creative by 18% in response rate. • Applied market basket analysis to find out which web pages are frequently viewed together. This analysis led to web site optimization for a client. • Implemented sequence clustering to classify navigation patterns on a client’s website. • Proposed and consulted with client service team to determine Key Performance Indicators (KPIs) for evaluating web visiting engagement. • Created a variable screening process which can automatically screen more than 1,000 variables. This process dramatically improves the model building process. • Built a marketing mix model with time series analysis. This model is used to simulate KPIs under different media investment scenario. • Implemented value added evaluation with multi-level modeling which quantified the unique contribution of using a new technique. • Collected data from different public sources, including Census Bureau, Bureau of Labor Statistics, SAS, and ESRI. These free public data sources were used in various predictive models to increase prediction and reduce cost. • Improved a scoring process by leveraging SAS macros. This scoring process is easier to implement, automatic in QC, and error preventive. Show less Built predictive, segmentation, and marketing mix models, implemented scoring system, and wrote technical documents. Analyzed web navigation behaviors. Reported on Direct Mail campaign performance. • Applied clustering analysis to consumer segmentation. The segmented creative outperformed the general creative by 18% in response rate. • Applied market basket analysis to find out which web pages are frequently viewed together. This analysis led to web site optimization for a client… Show more Built predictive, segmentation, and marketing mix models, implemented scoring system, and wrote technical documents. Analyzed web navigation behaviors. Reported on Direct Mail campaign performance. • Applied clustering analysis to consumer segmentation. The segmented creative outperformed the general creative by 18% in response rate. • Applied market basket analysis to find out which web pages are frequently viewed together. This analysis led to web site optimization for a client. • Implemented sequence clustering to classify navigation patterns on a client’s website. • Proposed and consulted with client service team to determine Key Performance Indicators (KPIs) for evaluating web visiting engagement. • Created a variable screening process which can automatically screen more than 1,000 variables. This process dramatically improves the model building process. • Built a marketing mix model with time series analysis. This model is used to simulate KPIs under different media investment scenario. • Implemented value added evaluation with multi-level modeling which quantified the unique contribution of using a new technique. • Collected data from different public sources, including Census Bureau, Bureau of Labor Statistics, SAS, and ESRI. These free public data sources were used in various predictive models to increase prediction and reduce cost. • Improved a scoring process by leveraging SAS macros. This scoring process is easier to implement, automatic in QC, and error preventive. Show less

Research Assistant & Teaching Assistant

• Aug 2002 - Aug 2006

• Conducted research about using multi-level modeling to evaluate schools. This research led to my dissertation and a presentation in the annual conference of American Educational Research Association. • Conducted simulation studies to compare different correlation adjustment methods under different sample size and distribution conditions. This research led to a publication in The Journal of Experimental Education (Volume 75, 2007). • Consulted on quantitative research methods and… Show more • Conducted research about using multi-level modeling to evaluate schools. This research led to my dissertation and a presentation in the annual conference of American Educational Research Association. • Conducted simulation studies to compare different correlation adjustment methods under different sample size and distribution conditions. This research led to a publication in The Journal of Experimental Education (Volume 75, 2007). • Consulted on quantitative research methods and conducted data analysis. These activities resulted in two journal publications as the second author. • Gave lectures and answered questions in statistic courses, including ANOVA, multiple regression, and multivariate analysis. Helped students in SAS and SPSS. Greatly improved their statistical knowledge and inspired their passion in learning statistics. Show less • Conducted research about using multi-level modeling to evaluate schools. This research led to my dissertation and a presentation in the annual conference of American Educational Research Association. • Conducted simulation studies to compare different correlation adjustment methods under different sample size and distribution conditions. This research led to a publication in The Journal of Experimental Education (Volume 75, 2007). • Consulted on quantitative research methods and… Show more • Conducted research about using multi-level modeling to evaluate schools. This research led to my dissertation and a presentation in the annual conference of American Educational Research Association. • Conducted simulation studies to compare different correlation adjustment methods under different sample size and distribution conditions. This research led to a publication in The Journal of Experimental Education (Volume 75, 2007). • Consulted on quantitative research methods and conducted data analysis. These activities resulted in two journal publications as the second author. • Gave lectures and answered questions in statistic courses, including ANOVA, multiple regression, and multivariate analysis. Helped students in SAS and SPSS. Greatly improved their statistical knowledge and inspired their passion in learning statistics. Show less