
Title: Data Driven Campaigning with Amanda H.
Theory: Data Driven Campaigning
Guest: Amanda H.
Note Researcher: Grace Ge
Key Resources:
Dommett, K. (2019). Data-driven political campaigns in practice: Understanding and regulating diverse data-driven campaigns. Internet Policy Review, 8(4). https://doi.org/10.14763/2019.4.1432
Dommett, K., Barclay, A., & Gibson, R. (2023). Just what is data-driven campaigning? A systematic review. Information, Communication & Society, 27(1), 1–22. https://doi.org/10.1080/1369118X.2023.2166794

Outline of the Theory
- Dommett describes data-driven campaigning as the practice of utilizing extensive data analytics and digital technologies to inform and optimize political campaign strategies.
- This approach involves collecting, analyzing, and using data to tailor campaign messages, target specific voter groups, and predict electoral outcomes.
- The articles emphasizes the importance of data-driven campaigning due to its transformative impact on the efficiency, precision, and effectiveness of political strategies in modern democracies. Below are the key reasons she highlights for its significance: allowing higher percisions in targeted messages and micro-targeting; adapting to the current digital trend of communication "data is the new currency"; and enhancing campaign strategies such as data driven decision making.
Key Points to Focus on
- This episode should focus on the aspects of “purchased data disclosed by individuals” and “purchased inferred data” on social media platforms like TikTok.
- How do platforms like TikTok collect its users’ data? Through what strategies.
- How does TikTok organize and analyze the data? Such as determining which type of data can be sold to political campaigns?
- How does data change the overall landscape of political campaigns?
- What are the potential privacy issues relating to such practice, for example, does TikTok disclose that they will collect the data? Also, do users acknowledge that TikTok will use and sell their data to political campaigns?
- The use of data-driven techniques like micro-targeting has raised concerns about their impact on democratic norm such as: risks of mis-targeting and privacy violations, potential for contradictory or misleading messages to different audiences, challenges in regulating both digital and offline practices effectively.

Guest Bio and Information:
Amanda H. is a data science and machine learning expert with over five years of experience in driving business growth and optimizing operations. She currently leads growth, recommendation, and search strategies at TikTok, while contributing to ByteDance's Douyin platform in e-commerce advertising. Her key contributions include developing a user portrait system for ad targeting, optimizing automated bidding processes, and designing AB/uplift experiments to measure key business metrics like LTV, ROI, and CAC.
Before joining ByteDance, Amanda worked in financial services at RBC, where she advanced risk modeling and predictive analytics strategies. Her technical expertise includes Python, Spark, and SQL, and her experience spans global technology and financial sectors, positioning her as a leader in leveraging data for impactful decision-making.
How will Amanda’s Expertise Helps to Understand Data-Driven Campaign?
- Targeted Ad Campaigns: Amanda’s work on enhancing ad targeting through user portrait systems, lookalike modeling, and uplift modeling at TikTok will be directly relevant. She can help analyze how purchased data can be integrated to refine audience segmentation, improving targeting precision and campaign effectiveness.
- Optimizing Bidding Processes: With her experience in optimizing automated bidding systems, Amanda can help develop strategies for efficiently allocating campaign budgets based on insights derived from purchased data. This would improve campaign ROI and reduce wasted spend by ensuring that ads reach the right users at the optimal time.
- Performance Metrics Analysis: Amanda’s background in measuring and analyzing critical business metrics will be valuable for evaluating how purchased data influences campaign performance. Her experience in building and analyzing predictive models can support continuous optimization and ensure campaigns are data-driven and results-oriented.
- Overall Understanding of Data Purchasing Process: by working at TikTok and analyzing its users’ data, Amanda can potentially provide valuable insights into how the general data collection process works within large-scale platforms. This includes understanding how data is sourced, processed, and made available for purchase, as well as the quality and reliability of the data being offered. Her expertise in advanced data analysis and modeling will allow her to assess how purchased data is integrated with first-party data to enhance targeting and personalization in campaigns. She could also identify any potential biases or limitations in the purchased data, ensuring that data-driven campaigns remain accurate, effective, and ethically sound. This holistic understanding would help optimize the use of purchased data, ensuring it aligns with campaign objectives and delivers actionable insights.

Show Intro and Questions:
Welcome to Wonks and War Rooms, where political communication theory meets on the ground strategy. I'm your host, Elizabeth Dubois. I'm an Associate Professor and University Research Chair in Politics, Communication and Technology at the University of Ottawa. My pronouns are she/her, and today we're talking about intersection of technology, data, and innovation. Today, we’re diving into the increasingly important world of data-driven campaigning, with a special focus on purchased data, particularly in platforms like TikTok. Our Guest today is Amanda H. Amanda, can you introduce yourself, please?
For Katherine Dommett, a prominent researcher, she defines data-driven campaigning as a process that integrates data collection, analytics, and digital technologies to optimize communication and resource allocation within political campaigns. This approach involves using real-time metrics, such as social media analytics and A/B testing, to assess the effectiveness of messages and interventions, ultimately aiming to influence voter behavior and increase campaign efficiency. She defines the data sources and categorizes them into four main section, today, we are focusing on the purchased data both disclosed from individuals and inferred by media platforms. So, Amanda, you’ve worked extensively with data-driven campaigns at TikTok. To start, can you give us a high-level overview of how data purchasing fits into the broader landscape of digital-focused campaigning today?
We already discussed that political campaigns can purchase data from different platforms and can be coming from various sources. What are some of the key types of data that campaigners typically purchase, and how do they integrate this data into their overall campaign strategy?
In your experience, how does the use of purchased data differ in commercial sectors, like e-commerce, compared to political campaigns?
You’ve worked with recommendation systems at TikTok. How do you think similar algorithms could be applied to political campaigns, particularly when it comes to segmenting voter populations and personalizing messages?
One of the challenges in political campaigns is understanding voter behavior. How do machine learning techniques, like the ones you’ve implemented at TikTok, help campaign teams predict and influence voter decisions using purchased data?
Another challenge is that given the potential impact of purchased data on voter targeting, how does TikTok ensure that the data used in political campaigns is both effective and ethical?
With the vast amount of data available, how do you recommend managing privacy concerns, especially in the context of political campaigns using purchased data for targeting?
Pop Quiz: Amanda, in your opinion, what do you see as the future of data-driven political campaigning, especially with the increasing role of social media platforms like TikTok in shaping public opinion?

Show Notes
In this episode of Wonks and War Rooms, host Elizabeth Dubois sits down with Amanda H., a data scientist and machine learning expert at TikTok and ByteDance, to explore the intersection of technology, data, and innovation in modern political campaigning. Together, they dive into the concept of data-driven campaigning, focusing on the use of purchased data and its role in shaping political strategies on platforms like TikTok. Elizabeth begins by introducing data-driven campaigning as a process that integrates data collection, analytics, and digital technologies to optimize communication and resource allocation, as defined by researcher Katherine Dommett. With Amanda’s expertise in leveraging data and recommendation systems, they explore how purchased data is utilized in political campaigns to target audiences, personalize messaging, and enhance overall campaign effectiveness. As well as the types of data campaigns typically purchase, how this data is integrated into broader strategies, and how approaches developed for commercial use at TikTok, such as segmentation and personalized recommendations, can be applied to political contexts. The discussion also addresses ethical considerations and privacy concerns, emphasizing the importance of balancing effectiveness with responsible data use.
Additional Resources
To learn more about big data in political campaigns, consult Political Campaigns and Big Data by David W. Nikerson and Todd Rogers
To learn more about the potential privacy and surveillance issues with social media and other corporations data collection, check out Wonks and War Rooms Season 3 Episode 8 Surveillance Capitalism with Vass Bednar
Also wondering about the Canadian laws on data usage by political parties? Check out Canadian Election Act and Bill C-76, the Elections Modernization Act.
Create Your Own Website With Webador