Sourced and cleaned datasets to conduct exploratory data analysis (EDA) to find marketing insights and identified online campaign pain points; leveraged findings to inspect customer retention and built out an improved online channel strategy to help company to increase sales conversion rate and reduced campaign investment expenses. Designed and built Power BI dashboard to monitor the products sales performance.
Scraped e-commerce websites’ product attributes via Python; Conducted NLP analysis to gain insights on market and target customers’ needs; Built text classification model to identify new product description quality; Conduct sentiment analytics for new products research and built predictive model for products to support decision making for new marketing strategy.
Wrote reports;Presented insights and brainstormed new market strategies with partners based on the projects’ findings.
Developed brand strategies for China Unicom’s 5G project for 2022 Winter Olympics campaign.
Collected information from public websites to make researches about 5G campaigns from others’ telecommunications companies; Brainstormed market strategies with team members; Wrote creative briefs; Wrote analysis reports; Designed and presented PPTs in weekly team meetings.
Assisted development;negotiation and execution of media plans valued around 1.000.000 RMB for each campaign across multiple digital channels such as social medias for top vehicle company.
Analyzed marketing data such as click and conversion from each campaign and find insights by Python and wrote reports to assist decision-making.
Utilized data-driven ways to help business clients to attract and convert target customers for ALDO’s online business.
Analyzed marketing data by Python and Tableau such as click and revenue to track the performance of each campaign;Visualized data to find insights from data; Wrote reports to support company’s decision-making process.
Scraped the products information and customers’ reviews from e-commercial websites to gain insights to support decision making process for marketing investment and products order for warehouses in different location to reduce unnecessary inventory backlog; // Applied Python packages like Pandas, NLTK and Scikit-learn to conduct EDA, sentiment analysis, opinions extraction and visualization; // Understood business values of each feature and processed unstructured and noisy data. Found what kind of features for the products have good sales performance and customers’ opinions and sentiment to the products in the US market.
Used historic dataset of March Madness competition to predict the winner in 2019. // Understood game value of features, processed and merged different formats datasets to get a data-frame available for model training. Implemented a completed data science pipeline (data processing, features engineering & selection). Trained dataset and evaluated by LogLoss, selected a best fit model from XGBoost, Logistics Regression, Random Forest. Wrote report and presented in final competition.