Hi, my name is Erfan and I am a quantitative marketing Ph.D. candidate at the University of Washington.
My research focuses on applying rigorous machine learning and econometric methods to consequential problems in digital marketing and public policy. I bring expertise in modern machine learning, causal inference, and econometric techniques to study questions that matter to both practitioners and decision makers. I am committed to applying and designing novel methodological tools that help firms improve their marketing practice, and to studying and understanding important public policy problems. These two lines of work are exemplified in my publications, where I have developed novel experimentation methods for firms to set prices and promotions, and have studied how access barriers shape marketing effectiveness in healthcare markets. My work bridges methodological innovation with real-world impact, using novel technological approaches to address challenges central to both business practice and policy.
My job market paper addresses a practical challenge firms face when optimizing unstructured treatments like text and images. While generative models make creating these forms scalable, pre-trained models are often not aligned to business objectives and require quality data to optimize performance. Firms face a fundamental tradeoff between observational and experimental data. Historical data is abundant but subject to confounding from spurious correlations, while experiments provide unbiased measurements but are costly and potentially noisy at small scale. In my work, I first quantify the value of each data source and then propose an inverse variance weighted method to effectively integrate the two sources, which outperforms using either alone. I then study what happens when firms face limited experimentation budgets and compare different experimentation strategies. My results show that the optimal strategy depends heavily on the setting and experimentation window, and that the two data sources can be cleverly combined to select which cases to test.
I am on the academic job market for positions starting July/September 2027 and look forward to contributing my research and teaching expertise to a leading academic institution.
Email: loghmani@uw.edu
Curriculum Vitae
Research Interests
Substantive Areas: Digital Marketing, Experimentation, Public Policy, Healthcare Marketing
Methods: Deep Learning, Causal Inference, Active Learning, Machine Learning, Econometrics
Job Market Paper
Published Papers
- Loghmani, E. and Goli, A. (2025). “Investigating the Impact of Advertising on Smoking Cessation: The Role of DTC Prescription Drug Advertising,” Marketing Science, DOI: mksc.2024.0848
- Loghmani, E. (2025). Aligning Language Models with Observational Data: Opportunities and Risks from a Causal Perspective, Accepted at NeurIPS 2025 Workshop on CauScien: Uncovering Causality in Science, Available at OpenReview and arXiv
- Jain, L., Li, Z., Loghmani, E., Mason, B., Yoganarasimhan, H. (2024). “Effective Adaptive Exploration of Prices and Promotions in Choice-Based Demand Models,” Marketing Science, DOI: mksc.2023.0322
- Gastinger, J., Huang, S., Galkin, M., Loghmani, E., Parviz, A., Poursafaei, F., Danovitch, J., Rossi, E., Koutis, I., Stuckenschmidt, H., Rabbany, R., Rabusseau, G. (2024). “TGB 2.0: A Benchmark for Learning on Temporal Knowledge Graphs and Heterogeneous Graphs,” Accepted at Datasets and Benchmarks Track of the NeuRIPS 2024 Conference, Available at arXiv
- Loghmani, E., Fazli, M. (2025). “Effect of Choosing Loss Function when Using T-batching for Representation Learning on Dynamic Networks,” Information Sciences, DOI: j.ins.2025.122755
- Fazli, M., Alian, P., Owfi, A., Loghmani, E. (2024). “RPS: Portfolio Asset Selection using Graph based Representation Learning,” Intelligent Systems with Applications, 200348. DOI: j.iswa.2024.200348
Work in Progress
- Loghmani, E., Goli, A., Akchurina, D. Understanding Substitution in Mobile Phone Usage: Evidence from Behavior Sequence Modeling
Education
Advisors: Prof. Ali Goli, Prof. Amin Sayedi
Advisor: Prof. MohammadAmin Fazli
Honors and Awards
Presentations
Aligning Language Models with Observational Data: Opportunities and Risks from a Causal Perspective
- 2025 Poster Presentation at the Twenty-Sixth ACM Conference on Economics and Computation (EC’25)
- 2025 Poster Presentation at the NeurIPS 2025 CauScien: Uncovering Causality in Science Workshop
Discussion for: “Review Manipulation and Platform Policy: Evidence from Online Travel Agencies”
- 2025 Haring Symposium @ Kelly School of Business
Investigating the Impact of Advertising on Smoking Cessation: The Role of Direct-to-Consumer Prescription Drug Advertising
- 2025 Tobacco Online Policy Seminar (TOPS) — Recording available on YouTube
- 2024 Program In Health Economics And Outcomes Research Methodologies (PHEnOM) at UW CHOICE
- 2024 UW-UBC Joint Marketing Conference
Teaching Experience
Service
Work Experience
As a Machine Learning Engineering Intern on Instacart’s Economics team, I focused on pricing experimentation and policy evaluation. I developed off-policy evaluation methods to address challenges posed by large experimentation state spaces, enabling more reliable assessment of pricing policies. I also built and deployed an automated LLM-based experimentation monitoring system that improved the team’s ability to track and interpret experiments accurately and efficiently.
During my internship at Stackline, I applied econometric and causal inference methods to analyze how different application components impacted user behavior. I uncovered causal relationships that helped identify the value of specific product features. As part of this work, I identified the causal effects of loyalty programs on customer purchasing decisions, both within the same product category and across different categories.
At Rooberah.co, I played a key role in the development of a Software as a Service (SaaS) platform aimed at boosting online store sales through the utilization of experimentation and machine learning techniques like recommender systems. My primary responsibility involved designing and implementing innovative features.
At Pushe.co, I started as a backend developer, utilizing the Django web framework to create robust web applications. Later, I joined the Data team, where I designed and implemented machine learning methods for fraud detection and CTR prediction.
Volunteer Experience
I joined Outdoors for All to support individuals with disabilities in accessing nature, providing the instruction and company needed to help them feel confident and safe on the water.
Interests and Hobbies
I love staying active. Volleyball is my favorite sport because I really enjoy the teamwork involved. When I want to relax, I like paddleboarding to connect with nature. I also spend time rock climbing to build my strength and reach new heights, along with hiking and swimming.
Academic References
Joint Letter
- Prof. Ali Goli — Assistant Professor of Marketing, Simon Business School, University of Rochester
- Prof. Amin Sayedi — Professor of Marketing, Foster School of Business, University of Washington
Solo Letters
- Prof. Hema Yoganarasimhan — Professor of Marketing, Michael G. Foster Faculty Fellow, University of Washington
- Prof. Shirsho Biswas — Assistant Professor of Marketing, Foster School of Business, University of Washington
- Prof. Lalit Jain — Assistant Professor of Marketing, Foster School of Business, University of Washington