- Confidential is an intellectual curious and customer - centric self-starter who tackles complex business issues by converting raw data into meaningful and actionable business insights.
- An analyst who is able to discover hidden insights by using a combination of quantitative and qualitative techniques such as regression, correlation, conjoint, factor analysis.
Analytics: SAS, Google Analytics, IBM SPSS, Tableau
Database: MS SQL Server, MySQL
Programming: Intermediate SQL, Beginner JAVA, Beginner PHP
- Export data from database to generate weekly and monthly reports for two product lines, perform sales forecast and present actionable insight and recommendations based on data results
- Provide analytic support to the product managers and sales team
- Analyze the customer purchase history and contribute to the CRM strategy
- Evaluate the progress of the customer acquisition and new product launch
- According to the data mapping, selected and extracted samples from a list of raw datasets, eliminated duplicates, created a one-subject-per-row dataset with PROC SQL.
- With given requirement, established a segmentation for the behavior of customer with PROC FORMAT based on their credit score. Applied this segmentation on all selected samples.
- Defined new variables and assigned different designated values to them based on campaign matrix using CASE expression in PROC SQL, derived product information for each customer.
- Created a Macro with keyword parameters of data cleansing process for future use, including PROC UNIVRIATE, PROC PRINT and DATA step.
- As instructed, split the dataset into 3 different percentage with DATA step.
- Imported raw data into a new dataset with DATA step, defined several new variables to calculate roll-rate (RR) of different time periods with LAG function.
- Performed different calculation of RR for first appearance of every account and rest with First.
- Performed 3-month moving average analysis of RRs. For first appearance of every account, assigned missing value to all RRs.
- For rest of the records, utilized SUM, LAG and N function to achieve moving average calculation. Discrimination of first appearance and rest of the records was implemented by First.
- Based on calculated moving average of RRs, converted original multiple-row-per-account dataset into a one-row-per-account dataset. Only last record of every account was kept with last.
- Selected, extracted and combined variables with PROC SQL. Eliminated duplicates with distinct option.
- For charge-off customers, calculated the time from the beginning of study to the time that they charged-off with INTCK function. For non-charge-off customers, assigned a constant for this variable.
- To find out the potential relation between default behavior and the time, performed logistic regression with PROC LOGISTIC and figured out the regression equation.
- Based on logistic regression analysis, forecasted a list of customers with different delinquency history to see the probability that they would default in the future.
Confidential, Richardson, TX
- Designed e-marketing materials, brochures and advertisements, attracted 500+ followers on social media with low budget
- Customer relationship management, achieved 90% customer satisfaction rate and over 70% customer retention rate
- Created and analyzed sales data and user survey to identify customer demand, developed new marketing strategies to increase the customer base
Digital Marketing Analyst
- Designed company’s website using WordPress
- Monitored the website traffic, analyzed the IT recycling industry and competitors
- Generated the AdWords based on the marketing research and the company’s services
- Achieved an average click-through rate (CTR) of 2.55% by obtained 13,745 impressions and 351 clicks at an average cost-per-click (CPC) of $0.71 after a three-weeks AdWords campaign