# Leonard Kinzinger

> Leonard (Leo) Kinzinger is a Ph.D. candidate in Digital Marketing at the Technical University of Munich (TUM), Chair of Digital Marketing, TUM School of Management, advised by Prof. Dr. Jochen Hartmann. In fall 2026 he is a visiting researcher at the Digital Twin Lab at Columbia Business School, hosted by Oded Netzer and Olivier Toubia. His research combines generative AI methods with marketing science, with a focus on LLM-based digital twins for market research.

- Website: https://leonardkinzinger.de/
- Email: leonard.kinzinger@tum.de
- TUM profile: https://www.msl.mgt.tum.de/dm/team/kinzinger-leonard/
- Google Scholar: https://scholar.google.com/citations?user=xUOkE4wAAAAJ
- LinkedIn: https://www.linkedin.com/in/leonardkinzinger
- Based in Munich, Germany

## Research interests

- LLM-based digital twins and synthetic respondents for market research
- Conditioning language models on socio-economic microdata to simulate individual respondents
- Visual features and temporal dynamics in video advertising
- Generative AI for music marketing and AI-generated music
- Audio watermarking

## Publications

- [Valuable Visual Variety: How Temporal Dynamics of Visual Features in Video Ads Affect Clicks and Purchases](https://papers.ssrn.com/sol3/papers.cfm?abstract_id=7356279). Jan Ole Krugmann, Leonard Kinzinger (equal contribution), Keno Tetzlaff, Jochen Hartmann. Forthcoming in Journal of the Academy of Marketing Science.

## Working papers

- [Synthetic Personalities: How Well Can LLMs Mimic Individual Respondents Using Socio-Economic Microdata?](https://arxiv.org/abs/2606.04592). Leonard Kinzinger, Jochen Hartmann. Revise and resubmit at International Journal of Research in Marketing. Builds individual-level digital twins from the German Socio-Economic Panel (SOEP), scoring more than 2.1 million twin responses from 500 participants on 183 held-out questions; the best configuration reaches 78.8% accuracy.
- [Latent Audio Watermarking for Robustness to Neural Codec Resynthesis](https://arxiv.org/abs/2609.25830). Lovro Brulec, Sahil Karawade, Leonard Kinzinger. Under review at ICASSP.

## Work in progress

- LLM-Based Digital Twins: Near Perfect or Just Better Than Chance. With Oded Netzer and Jochen Hartmann.
- GenAI for Music Marketing. With Shihong Zhang and Jochen Hartmann.
- Consumer Responses to AI Music. With B. Roder and Jochen Hartmann.

## Education

- Ph.D. in Digital Marketing, Technical University of Munich (2024 to present)
- Visiting Researcher, Digital Twin Lab, Columbia Business School (fall 2026)
- M.Sc. Management and Technology, Technical University of Munich (2022 to 2024), grade 1.3 (top 5%), master’s thesis graded 1.0 at the TUM GenAI Lab
- B.Sc. Industrial Engineering and Management, Technical University of Berlin (2017 to 2022), grade 1.6 (top 15%)

## Honors and awards

- 2026: Finalist, PhD Best Paper Award, Artificial Intelligence in Management (AIM) Conference, USC Marshall
- 2025: German Market Research Award, Young Researcher of the Year (BVM / VMÖ / SWISS INSIGHTS), category Master’s thesis, €2,000 prize, for the thesis "Exploring the In-Context Learning Ability of LLMs to Create Realistic Synthetic Personalities". Announcement: https://www.bvm.org/unser-beruf/preis-der-deutschen-marktforschung/nachwuchsforscherin-bvmvmoeswiss-insights/rueckblick-nachwuchsforscherin-bvmvmoeswiss-insights/
- 2023: TUMfast Selected Talents Scholarship, TUM School of Management

## Teaching

- Marketing Mobility with Generative AI, seminar at the TUM School of Management (2025 to present), co-taught with the Professorship of Mobility Policy with industry partners such as Munich Airport, MVG, Deutsche Bahn, Flix and MINGA.

## Service

- Ad hoc reviewer for Management Science and the International Journal of Research in Marketing