More is Less? Design Free Sample Strategy via Field Experiment and Double-Debiased Machine Learning - SlideShare

Aug 5, 2026

Free sample strategy has attracted strong attention in digital content industries such as e-books, music, and videos, yet two persistent design questions remain: the optimal quantity of free samples and how to personalize their offering under different contextual conditions. Collaborating with a China-based online reading platform, the study runs a field experiment grounded in Construal Level Theory. Results reveal an inverted U-shaped effect of free sample quantity on purchase decisions, and positive moderation by book popularity and quality when free chapters are provided. The work further builds a personalized strategy using causal forest with double/debiased machine learning and derives managerial implications.

LanguageEnglish

CategoryResearch & Report

FormatPDF

Pages10

Reading Time25 min

AccessFree to download

Abstract

Keywords

Introduction

Digital content markets and free sample strategy

Research gaps in optimal design

Study design and theoretical basis

Construal Level Theory (CLT)

Methodology and model development

Causal forest and double/debiased machine learning

Findings and managerial implications

Association for Information Systems AIS Electronic Library (AISeL) Rising like a Phoenix: Emerging from the Pandemic and Reshaping Human Endeavors with Digital Technologies ICIS 2023 Digital Innovation, Transformation, and Entrepreneurship Dec 11th, 12:00 AM More is Less? Design Free Sample Strategy via Field Experiment and Double/Debiased Machine Learning JIN LIU University of Science and Technology of China, [liujin07@mail.ustc.edu.cn](liujin07@mail.ustc.edu.cn) Hanbing Xue university of science and technology of China, [xuehb@mail.ustc.edu.cn](xuehb@mail.ustc.edu.cn) yongjun li University of Science and Technology of China, [lionli@ustc.edu.cn](lionli@ustc.edu.cn) Follow this and additional works at: [https://aisel.aisnet.org/icis2023](https://aisel.aisnet.org/icis2023) Recommended Citation LIU, JIN; Xue, Hanbing; and li, yongjun, "More is Less? Design Free Sample Strategy via Field Experiment and Double/Debiased Machine Learning" (2023) . Rising like a Phoenix: Emerging from the Pandemic and Reshaping Human Endeavors with Digital Technologies ICIS 2023. 13. [https://aisel.aisnet.org/icis2023/diginnoventren/diginnoventren/13](https://aisel.aisnet.org/icis2023/diginnoventren/diginnoventren/13) This material is brought to you by the International Conference on Information Systems (ICIS) at AIS Electronic Library (AISeL) . It has been accepted for inclusion in Rising like a Phoenix: Emerging from the Pandemic and Reshaping Human Endeavors with Digital Technologies ICIS 2023 by an authorized administrator of AIS Electronic Library (AISeL) . For more information, please [contact](contact elibrary@aisnet.org)[ elibrary@aisnet.org](contact elibrary@aisnet.org). More is Less? Design Free Sample Strategy More is Less? Design Free Sample Strategy via Field Experiment and Double/Debiased Machine Learning Short Paper Jin Liu School of Management, University of Science and Technology of China 96, JinZhai Road, Hefei, Anhui, China [liujin07@mail.ustc.edu.cn](liujin07@mail.ustc.edu.cn) Hanbing Xue School of Management, University of Science and Technology of China 96, JinZhai Road, Hefei, Anhui, China [xuehb@mail.ustc.edu.cn](xuehb@mail.ustc.edu.cn) Yongjun Li School of Management, University of Science and Technology of China 96, JinZhai Road, Hefei, Anhui, China [lionli@ustc.edu.cn](lionli@ustc.edu.cn) Abstract Free sample strategy has attracted considerable interest among practitioners and academics, it has been widely adopted in digital content industries (e.g., e-books, music, and videos). There are two issues that have been the continuous concerning and constantly optimized focus. How many free samples should be taken? How to design a personalized free samples strategy considering the contexts? To better understand these issues, we collaborated with an online reading platform in China to design and conduct afield experiment based on Construal Level Theory (CLT). The results showed an inverted U-shaped relationship between free sample quantity and consumer purchase decisions and also suggested when free chapters were offered, book popularity and quality were also found to positively moderate consumers’purchase decisions. Moreover, by combining the causal forest (CF) technique and the double/debiased machine learning model (DML), we develop a personalized free sample strategy and provide managerial implications. Keywords: Digital Content; Free samples; Field Experiment; Construal Level Theory; Double Machine Learning Introduction From 2021 to 2025, the total revenue of digital content industries (e.g., e-books, movies, and music) is expected to show an annual growth rate of 11.25%, resulting in a projected market volume of US $12,251 million by 2025. Digital content industries present new opportunities. However, the quality and preference fit of digital content products with hedonic properties and high consumer heterogeneity can only be determined after consumers experi

FAQs

What are the two key issues the paper addresses about free samples?

The study focuses on how many free samples to take and how to design a personalized free-sample strategy that accounts for contextual conditions.

What does the field experiment show about the relationship between free sample quantity and purchase decisions?

It finds an inverted U-shaped relationship between free sample quantity and consumer purchase decisions.

How does the paper create a personalized free sample strategy?

It combines causal forest with a double/debiased machine learning model to build a personalized strategy and produce managerial implications.

Aug 5, 2026

Upload by Olivia Brown