[{"content":" For coding: Code editor with LLM integration: Visual Studio Code , Cursor Git (either command line or with a Github GUI)\nMarkdown editor (for project documentation in Git): Typora , Obsidian Other productivity tools: Markdown-based presentation compiler: Marp RSS feed reader that can generate a custom feed for any regularly updated webpage (e.g. working paper series for a specific topic): Inoreader Boston coffee roaster recommendations: Broadsheet George Howell Gracenote ","permalink":"https://tesarylin.github.io/resources/tools/","summary":"Tools I have settled on or am tinkering with (updated every sometimes).","title":"Recommended Tools for Research Productivity"},{"content":" Download Paper Presentation Video (DISS'25; 27min) Abstract We conduct the first large-scale field experiment on how consent interface design affects privacy choices, using a browser extension to randomize interfaces that consumers face during their natural web browsing. Deliberate obstruction—hiding rejection options behind additional clicks—reduces selection of hidden choices by 70% relative to the baseline rates, while visual manipulations produce modest effects. Surprisingly, these design effects do not systematically advantage popular websites over smaller competitors. During organic browsing, 22% of users close banners without choosing, yet hold varying beliefs about what closure implies, making defaults critical. A structural welfare analysis reveals that current US practices reduce consumer surplus by 23.5% relative to manipulation-free interfaces, but browser-level consent mechanisms improve welfare by 150% compared to any site-by-site approach. The dominant factor is time costs: users spend 6.6 minutes weekly on consent interactions worth $4/week, suggesting that current regulatory focus on banner design may miss larger architectural solutions.\nTABLE 8. Consumer Surplus Under Counterfactual Policies ($, Weekly) Citation Farronato, Chiara, Andrey Fradkin, and Tesary Lin. “Designing consent: Choice architecture and consumer welfare in data sharing.” NBER Working Paper No. w34025 (2025).\n@article{farronato2025designing, title={Designing Consent: Choice Architecture and Consumer Welfare in Data Sharing}, author={Farronato, Chiara and Fradkin, Andrey and Lin, Tesary}, year={2025}, journal={NBER Working Paper}. } Funding Internet Services Grant ","permalink":"https://tesarylin.github.io/research/dark-pattern/","summary":"Our field experiment shows that choice friction dominate user preferences in online data sharing; browser-level privacy controls improve welfare 150% more than banning manipulative consent designs. 📖 Working paper","title":"Designing Consent: Choice Architecture and Consumer Welfare in Data Sharing"},{"content":" Download Paper Abstract As businesses increasingly rely on granular consumer data, the public has increasingly pushed for enhanced regulation to protect consumers’ privacy. We provide a perspective based on the academic marketing literature that evaluates the various benefits and costs of existing and pending government regulations and corporate privacy policies. We make four key points. First, data-based personalized marketing is not automatically harmful. Second, consumers have heterogeneous privacy preferences, and privacy policies may unintentionally favor the preferences of the rich. Third, privacy regulations may stifle innovation by entrepreneurs who are more likely to cater to underserved, niche consumer segments. Fourth, privacy measures may favor large companies who have less need for third-party data and can afford compliance costs. We also discuss technology platforms’ recent proposals for privacy solutions that mitigate some of these harms but, again, in a way that might disadvantage small firms and entrepreneurs.\nCitation Jean-Pierre Dubé, John G. Lynch, Dirk Bergemann, Mert Demirer, Avi Goldfarb, Garrett Johnson, Anja Lambrecht, Tesary Lin, Anna Tuchman, Catherine Tucker. “Frontiers: The Intended and Unintended Consequences of Privacy Regulation for Consumer Marketing.” Marketing Science 44(5) (2025): 975-984.\n@article{dube2025frontiers, title={Frontiers: The Intended and Unintended Consequences of Privacy Regulation for Consumer Marketing}, author={Dub{\\\u0026#39;e}, Jean-Pierre and Lynch, John G and Bergemann, Dirk and Demirer, Mert and Goldfarb, Avi and Johnson, Garrett and Lambrecht, Anja and Lin, Tesary and Tuchman, Anna and Tucker, Catherine}, journal={Marketing Science}, volume={44}, number={5}, pages={975-984}, year={2025}, publisher={INFORMS} } ","permalink":"https://tesarylin.github.io/research/msi/","summary":"A joint position paper based on our prior research findings related to privacy and data regulations. 📘 Accepted: Marketing Science","title":"Frontiers: The Intended and Unintended Consequences of Privacy Regulation for Consumer Marketing"},{"content":" Download Paper Abstract We examine the tradeoff between privacy and personalization for online content by evaluating the impact of YouTube\u0026rsquo;s settlement with the Federal Trade Commission over violating the Children\u0026rsquo;s Online Privacy Protection Act (COPPA). Under the settlement, YouTube removed all forms of personalization for child-directed content starting in January 2020, which included personalized ads and platform features like personalized search and recommendations. We study the resulting impact on 5,066 top American YouTube channels by comparing the child-directed content creators to their non-child-directed counterparts using a difference-in-differences design. On the supply side, child-directed content creators produce 13% less content and pivot towards producing non-child-directed content. Child-directed content creators also invest less in content quality: the proportion of original content falls by 11% and manual captioning falls by 27%, while user content ratings fall by 10%. On the demand side, views of child-directed channels fall by 20%. Consistent with the platform\u0026rsquo;s degraded capacity to match viewers to content, both content creation and content views become more concentrated among top child-directed YouTube channels.\nFigure 6a: Time-Varying Treatment Effects on Log Video Uploads among Made-for-Kids Chanels Citation Johnson, Garrett and Lin, Tesary and Cooper, James C. and Zhong, Liang, COPPAcalypse? The Youtube Settlement\u0026rsquo;s Impact on Kids Content (March 14, 2024). Available at SSRN: https://ssrn.com/abstract=4430334 @article{johnson2024coppacalypse, title={COPPAcalypse? The Youtube Settlement\u0026#39;s Impact on Kids Content}, author={Johnson, Garrett and Lin, Tesary and Cooper, James C and Zhong, Liang}, journal={SSRN Working Paper No. 4430334}, year={2024} } Funding Economics of Digital Services Grant (University of Pennsylvania) Digital Business Institute (Boston University) ","permalink":"https://tesarylin.github.io/research/coppacalypse/","summary":"A COPPA settlement targeting YouTube led to a 13% reduction in YouTube\u0026rsquo;s made-for-kids content; demand responded by concentrating on popular channels. 📚 Accepted: Management Science","title":"COPPAcalypse? The YouTube Settlement’s Impact on Kids Content"},{"content":" Download Paper Presentation Video (EC’23; 21min) Abstract How does choice architecture used during data collection influence the quality of collected data in terms of volume (how many people share) and representativeness (who shares data)? To answer this question, we run a large-scale choice experiment to elicit consumers’ valuation for their Facebook data while randomizing two common choice frames: default and price anchor. An opt-out default decreases valuations by 22% compared to opt-in, while a $0–50 price anchor decreases valuations by 37% compared to a $50–100 anchor. Moreover, some consumer segments are influenced by frames more while having lower average privacy valuations. As a result, conventional frame optimization practices that aim to maximize data volume can exacerbate bias and lower data quality. We demonstrate the magnitude of this volume-bias trade-off in our data and provide a framework to inform optimal choice architecture design.\nFigure 1b: Illustration: How Choice Architecture Affects Sample Data Quality In this example, the proportion of low- and high-income consumers in the full sample is roughly 2:1; consumers share data when their valuation on privacy is lower than the price (compensation) offered by the firm. In the no-frame-optimization condtion (supply curve 1), the sample data is representative. With the volume-maximizing frame (supply curve 2), the sample who share data predominantly consists of low-income consumers. Citation Lin, Tesary, and Avner Strulov-Shlain. \u0026ldquo;Choice architecture, privacy valuations, and selection bias in consumer data.\u0026rdquo; Marketing Science 44(6) (2025): 1217-1459.\n@article{lin2025choice, title={Choice architecture, privacy valuations, and selection bias in consumer data}, author={Lin, Tesary and Strulov-Shlain, Avner}, journal={Marketing Science}, volume={44}, number={6}, pages={1217-1459}, year={2025}, publisher={INFORMS} } Awards and Recognitions 2023 ACM Conference on Economics and Computation, Exemplary Track 2023 Alessandro di Fiore Best Paper Award 2021 Becker Friedman Institute Grant (University of Chicago) ","permalink":"https://tesarylin.github.io/research/choicearch-bias/","summary":"Should companies focus on maximizing consent rates when designing their cookie banners? We show that such practice can exacerbate sample bias in certain situations, and propose a better alternative. 📚 Published: Marketing Science","title":"Choice Architecture, Privacy Valuations, and Selection Bias in Consumer Data"},{"content":" Download Paper Presentation Video (2021 TSE Digital Economics Conference; 15min) Abstract I empirically separate two components in a consumer’s privacy preference. The intrinsic component is a “taste” for privacy, a utility primitive. The instrumental component comes from the consumer’s anticipated economic loss from revealing his private information to the firm, and arises endogenously from a firm’s usage of consumer data. Combining an experiment and a structural model, I measure the revealed preferences separately for each component. Intrinsic preferences have seemingly small mean values, ranging from 0.14 to 2.37 dollars per demographic variable. Meanwhile, they are highly heterogeneous across consumers and categories of data: The valuations of consumers at the right tail often exceed the firm’s valuation of individual consumer data. Consumers’ self-selection into data sharing depends on the respective magnitudes and correlation between the two preference components, and often deviate from a simplistic “low types are more willing to hide” argument. Through counterfactual analysis, I show how this more nuanced selection pattern changes what firms can infer from consumers’ privacy decisions and its implication on effective data buying strategies.\nFigure 5: Willingness-to-Accept Distribution in Consumers’ Intrinsic Preferences for Demographic Information (Estimated via Hierarchical Bayes Model) Citation Lin, Tesary. \u0026ldquo;Valuing intrinsic and instrumental preferences for privacy.\u0026rdquo; Marketing Science 41.4 (2022): 663-681.\n@article{lin2022valuing, title={Valuing intrinsic and instrumental preferences for privacy}, author={Lin, Tesary}, journal={Marketing Science}, volume={41}, number={4}, pages={663--681}, year={2022}, publisher={INFORMS} } Awards and Recognitions 2022 John Little Best Paper Award 2019 Sheth Foundation ISMS Doctoral Dissertation Award 2018 MSI Alden G. Clayton Doctoral Dissertation Proposal Award ","permalink":"https://tesarylin.github.io/research/dual-preference/","summary":"Why are privacy preferences contextual? This paper empirically identifies intrinsic and instrumental preferences for privacy as a way to explain privacy\u0026rsquo;s context dependence. 📚Published: Marketing Science","title":"Valuing Intrinsic and Instrumental Preferences for Privacy"},{"content":" Download Paper Online Appendix Presentation Video (2021 VQMS) Code: R Shiny App or Python Colab Abstract Consumers interact with firms across multiple devices, browsers, and machines; these interactions are often recorded with different identifiers for the same consumer. The failure to correctly match different identities leads to a fragmented view of exposures and behaviors. This paper studies the identity fragmentation bias, referring to the estimation bias resulted from using fragmented data. Using a formal framework, we decompose the contributing factors of the estimation bias caused by data fragmentation and discuss the direction of bias. Contrary to conventional wisdom, this bias cannot be signed or bounded under standard assumptions. Instead, upward biases and sign reversals can occur even in experimental settings. We then compare several corrective measures, and discuss their respective advantages and caveats.\nFigure 3: Comparison across Different Data and Estimators Note: The horizontal dashed lines indicate the estimated effect sizes, with black, red and blue representing the true (unfragmented), fragmented (naive estimator using fragmented data) and aggregated (stratefied aggregation on fragmented data) estimates. The vertical lines the 95% confidence region. Citation Lin, Tesary, and Sanjog Misra. \u0026ldquo;Frontiers: the identity fragmentation bias.\u0026rdquo; Marketing Science 41.3 (2022): 433-440.\n@article{lin2022frontiers, title={Frontiers: the identity fragmentation bias}, author={Lin, Tesary and Misra, Sanjog}, journal={Marketing Science}, volume={41}, number={3}, pages={433--440}, year={2022}, publisher={INFORMS} } ","permalink":"https://tesarylin.github.io/research/fragmentation/","summary":"Cookies have been crumbling long before the hammer of Privacy Sandbox strikes. This is bad for measurement and inference, but not necessarily in the way you expect. 📚 Published: Marketing Science","title":"Frontiers: Identity Fragmentation Bias"},{"content":"I am happy to provide a recommendation letter to help you succeed—but only if certain conditions are satisfied. These criteria are here to make sure I can write strong and informative letters for students that I recommend. I take pride in placing my students in good programs (in the past, students who got my letter have gotten offers from top PhD programs such as Northwestern, UT Dallas, and USC), and I don’t make endorsements lightly.\nSee below for the criteria:\nLetters for Research Programs: I will gladly write a letter for you if\nYou have worked with me on research projects for 6 months or longer, and While working with me, you have demonstrated a strong interest in research and your potential to become a good researcher. This typically means that you are driven and proactive, are able to propose creative solutions rather than simply following the tasks, and have demonstrated certain technical skills. Letters for Non-Research Programs: I will gladly write a letter for you if\nYou have interacted with me directly, either via working with me or by taking one of my courses, and You have demonstrated that you are a top-performing student either through your research performance (see the section above) or by fulfilling both criteria in my class: You have scored an A, and Your participation score is within the top 15% of the class. ","permalink":"https://tesarylin.github.io/resources/letter/","summary":"I am happy to provide a recommendation letter to help you succeed if certain conditions are satisfied. I use these criteria to make sure I can write strong and informative letters for students that I recommend.","title":"Students Interested in Getting a Recommendation Letter"},{"content":"Course Description This is a course on analytics in digital marketing. The core of marketing is reaching your audience and communicating the value of your brand and products to them, so that you can grow and retain customers. Digitization offers a variety of new data and tools that makes this effort more accessible for large and small companies alike. This course aims to familiarize students with digital marketing analytic tools, as well as the mindset of focusing on incrementality when analyzing the effects of marketing strategies. We will introduce marketing tactics used in different stages of a customer\u0026rsquo;s journey, including advertising, search engine optimization, pricing, and on-site marketing. In the context of these topics, we introduce analytic tools to measure marketing effects and optimize campaign efforts, including experiment design and analysis, targeting campaign design and assessment, recommender models, and attribution modeling.\n","permalink":"https://tesarylin.github.io/teaching/digital/","summary":"My course on analytics in digital marketing, taught at both the undergraduate and the graduate levels.","title":"Digital Marketing Analytics"}]