Welcome to our study on Generative AI effectiveness in email ✨

Our research focuses on how Generative AI can personalize various categories of email, like marketing, transactional, and even phishing email. This study aims to evaluate the effectiveness of AI-personalized emails, comparing user engagement and conversion between a generic mass email and an AI-personalized email.

This study is led by researchers from Utrecht University, aiming to bridge the gap between theoretical AI capabilities and practical applications. This research informs whether Large Language Models have the skills to effectively personalize email in a real-world context and experiment.


About This Study

Context

Exploring Generative AI's role in effective email communication

Generative AI (GenAI) is transforming how organizations communicate. This study investigates the use of large language models (LLMs), such as ChatGPT, to personalize a wide range of email categories—marketing, transactional, and even informational emails. The partnership with KPMG allows us to test these methods in a real-world corporate environment.

Email personalization has long been known to enhance engagement and conversion rates, but achieving this level of customization can be resource-intensive. By leveraging AI, we aim to automate and enhance the personalization process while evaluating its actual effectiveness compared to standard, generic emails.


Goal and Scientific Contribution

Evaluating the effectiveness of AI in personalized email communication

This study aims to assess how well AI-personalized emails perform in terms of user engagement (opens, clicks) and conversion rates (actions taken). By comparing generic emails with AI-personalized versions, we seek to uncover whether personalization leads to better outcomes.

Our findings will contribute to a deeper understanding of how generative AI can enhance communication strategies. This research bridges theoretical AI capabilities with practical applications, providing actionable insights for organizations aiming to optimize email communication.


Method

Employing randomized controlled trials in a corporate setting

The study uses randomized controlled trials (RCTs) to measure the effectiveness of AI-personalized emails. Recipients are divided into a control group (receiving generic emails) and multiple treatment groups (receiving AI-personalized emails generated by different LLMs). Engagement and conversion metrics are then tracked and analyzed to determine the impact of personalization.

Personalization is achieved by utilizing employee data such as job title, department and office location. Data security and privacy are prioritized, with all data stored securely and used solely for the purposes of this study.


Participant Information

Your privacy and rights are our priority

Participation in this study is voluntary, and we are committed to protecting your personal data in accordance with GDPR guidelines. Minimal data is collected for personalization purposes, and no identifiable information is shared with third parties without anonymization. Participants have full control over their data throughout the study.


Processing of Your Data

We collect and process data, such as your name, job title, and interests, to personalize email communication. This data is securely stored and used exclusively for research purposes. Sources include publicly available platforms and internal company systems.


Accessing Your Data

You can access, review, and update your data through a secure portal. Your rights to data privacy and management are central to this process.

Access Your Data

Privacy Policy

For detailed information about how your data is handled, read our privacy policy.

Read the Privacy Policy

Data Protection Impact Assessment (DPIA)

Learn how we ensure your data is handled responsibly through our DPIA.

View the DPIA

Data Management Plan (DMP)

Understand the procedures for managing and protecting your data throughout the study.

View the DMP

Opting Out of the Study

Your participation is entirely voluntary

If you wish to withdraw from the study, you can opt out at any time. Simply click the link below and follow the instructions to have your data removed.

Opt Out Here

Meet the Research Team

Get to know the experts behind the project

Have questions about the study, methodology, or your data rights? Contact the lead researcher at c.e.w.brachten@students.uu.nl or reach out to Utrecht University's privacy contact at privacy-beta@uu.nl.


C.E.W. (Mike) Brachten

Lead Researcher

Mike Brachten is a Master’s student in Business Informatics at Utrecht University. His research explores practical applications of generative AI in business, with a focus on bridging the gap between theoretical AI capabilities and real-world challenges.

Dr. R.L. (Slinger) Jansen

Project Supervisor

Dr. Slinger Jansen, Associate Professor at Utrecht University, specializes in software product management and ecosystems. His expertise in bridging academia and practice supports the study's innovative methodology.

Drs. N.A. (Nico) Brand

Second Examiner

Drs. Nico Brand, Assistant Professor at Utrecht University, focuses on software and process science. His expertise in digital transformation ensures a comprehensive approach to the study's research design.