top of page
Untitled.png

Researcher · Builder · Educator

Assistant Professor of Marketing | HKU Business School

Ph.D., MIT Sloan School of Management | Consultant to Tencent

​

I develop AI and causal-inference methods that make large-scale digital experimentation a more reliable basis for real-world decisions — estimating long-term treatment effects from short-term A/B tests, improving the external validity of online experiments, and using large language models for causal inference and product decisions. Empirically, I study how social media platforms shape advertising effectiveness, social referrals, and content diffusion through large-scale field experiments. This work has appeared in Management Science, Marketing Science, Journal of Marketing, Information Systems Research, and ACM EC.

​

I translate these methods into tools companies use at scale: my work helped launch WeChat's first large-scale A/B test, my methods have been adopted by leading technology companies including Tencent and ByteDance, and presented this work at companies worldwide, such as LinkedIn, Meta, and Kuaishou. I designed Digital Experimentation Methods — one of the first A/B testing courses in Asia and now the most in-demand elective in HKU's MSBA program — which earned the Faculty Teaching Innovation Award.

​

After two and a half years at UW Foster School of Business, I set out to explore Asia's technology landscape — my collaboration with Tencent since 2015 drew me to Hong Kong, where I joined HKU in 2021. I earned my Ph.D. from MIT Sloan School of Management, an MSc from the University of British Columbia, and a BBA from Tsinghua University. From 2020 to 2025, I was a Digital Fellow at Stanford's Digital Economy Lab.

​

Google Scholar | GitHub | Email | CV (PDF)

​

Research Interests: Technology-Enabled Marketing Strategies and Decision-Making
​

Digital Experimentation methods — Long-term treatment effects, external validity, A/B testing at scale

AI for marketing decisions — LLM-based causal inference, product decisions, digital twins

Social media platforms — Advertising effectiveness, content diffusion, algorithmic vs. social recommendations

Journal Publications

​

  • Shan Huang*, Chen Wang, Yuan Yuan, Jinglong Zhao & Jingjing Zhang (2026). Estimating Effects of Long-Term Treatments. Management Science.

    — Methods adopted by Tencent & ByteDance. Presented at Kuaishou, Meta, LinkedIn, Snapchat, etc. ​

  • Yifan Yu, Shan Huang*, Yuchen Liu & Yong Tan (2025). Emotions in Online Content Diffusion. Information Systems Research.

    — Computational methods quantifying how specific emotions drive information spread across large social networks.

  • Shan Huang* & Song Lin* (2024). Do More "Likes" Lead to More Clicks? Journal of Marketing, 89(5), 88-110.
    — Field experiment on WeChat revealing how social cues shape ad click-through (private responses) differently from "likes" (public responses) in social media advertising.

  • Shan Huang*, Sinan Aral, Yu Hu & Erik Brynjolfsson (2020). Social Advertising Effectiveness Across Products. Marketing Science, 39(6), 1142-1165.

— The first field experiments on WeChat (37M users), along which WeChat's first A/B testing system, now running thousands of experiments daily, was built. 

  • Hailiang Chen, Yu Hu & Shan Huang* (2019). Monetary Incentive and Stock Opinions on Social Media. Journal of Management Information Systems, 36(2), 391-417.

— How monetary incentives influence content production on social media platforms. 100+ Google Scholar citations.

 

Conference Proceedings​

 

  • Chen Wang, Shan Huang* & Shichao Han (2024). Enhancing External Validity of Experiments with Ongoing Sampling. ACM Conference on Economics and Computation (EC'24).

— Addresses sample-composition drift in A/B tests. Validated on 600 real experiments. Implemented in Tencent's experimentation system, adopted by ByteDance.

  • Shan Huang* & Yi Ji (2024). Algorithmic vs. Friend-based Recommendations in Shaping Novel Content Engagement. ACM Conference on Economics and Computation (EC'24).
    — Large-scale experimental evidence (2M+ users) on how algorithms and social ties differ in driving content engagement. Presented at Stanford.

  • Shan Huang*, Chen Wang, Yuan Yuan, Jinglong Zhao & Jingjing Zhang (2023). Estimating Effects of Long-Term Treatments. ACM Conference on Economics and Computation (EC'23).

— Conference version of the Management Science paper. Published in both a top management journal and a top CS venue.​

*corresponding author

​

​​Business Cases

​

  • Shan Huang*, Shipeng Yan, Zhenhui Jiang, & Minying Huang (2022), ESG at WeChat Pay to Support SMEs, Asia Case Research Centre

  • Shan Huang*, Xiaoming Yuan, and Minying Huang (2025), Algorithm Innovation at Huawei Cloud, Asia Case Research Centre, forthcoming

​​​​​

Book (in progress)

​​

  • Experimentation at Scale: Methods and Practice from China’s Tech Frontier

— with Yunfei Han* (Bytedance), Jinyong Ma* (Bytedance), Yong Wang* (Tencent), and Kenny Xie* (Google)

​​​

Recent News

​

  • Aug 2026 — Presenting "CausLab: LLM-Driven Multi-Agent Bayesian Framework for Causal Inference" at the 1st International Workshop on AI Data Scientist, KDD 2026 (with Chen Wang, Shichao Han, and Yong Wang).

​

Recent Talks​

​​​​​

  • March 2026: Invited talk at MIS Seminar, Purdue University, on LLM-Driven Multi-Agent Bayesian Framework for Causal Discovery and Inference, IN, USA. 

  • Feb 2026: Invited talk at Marketing Seminar, Purdue University, on Enhancing Externality of Experiments with Ongoing Sampling, IN, USA. 

  • Feb 2026: Invited talk at Marketing Seminar, Notre Dame University, on Enhancing Externality of Experiments with Ongoing Sampling, IN, USA. 

​

Media Coverage​

​

​

​

​

​

​

​

​

​

​

​

​

 

​​

bottom of page