When or What? Understanding Consumer Engagement on Digital Platforms

TL;DR

This study investigates why certain digital content becomes popular by examining TED Talks through topic modeling and engagement metrics. Using Latent Dirichlet Allocation, it identifies 14 topics and compares speakers' topic supply with audiences' engagement across 2006–2022, revealing that timing and contextual factors commonly trump content features in driving views. The analysis combines Chi-square tests, residuals, beta regression, Kruskal-Wallis tests, and correlations to quantify supply-demand gaps and the dynamic evolution of preferences. The results challenge the view that content alone governs popularity, offering practical insights for marketers, platform managers, and creators to optimize engagement strategies by aligning delivery timing with audience salience.

Abstract

Understanding what drives popularity is critical in today's digital service economy, where content creators compete for consumer attention. Prior studies have primarily emphasized the role of content features, yet creators often misjudge what audiences actually value. This study applies Latent Dirichlet Allocation (LDA) modeling to a large corpus of TED Talks, treating the platform as a case of digital service provision in which creators (speakers) and consumers (audiences) interact. By comparing the thematic supply of creators with the demand expressed in audience engagement, we identify persistent mismatches between producer offerings and consumer preferences. Our longitudinal analysis further reveals that temporal dynamics exert a stronger influence on consumer engagement than thematic content, suggesting that when content is delivered may matter more than what is delivered. These findings challenge the dominant assumption that content features are the primary drivers of popularity and highlight the importance of timing and contextual factors in shaping consumer responses. The results provide new insights into consumer attention dynamics on digital platforms and carry practical implications for marketers, platform managers, and content creators seeking to optimize audience engagement strategies.

Figures (6)

• Figure 1: The flowchart of the study. This flowchart summarizes the data processing and analysis procedures, including topic modeling, data normalization, and statistical analysis.
• Figure 2: Distribution of 14 topics in TED talks. The figure provides a comprehensive overview of the distribution and temporal trends of 14 distinct topics. Each panel represents a unique aspect of the dataset analysed, illustrating the relative prominence and variation of topics across different years. Panel (A) illustrates the distribution of various topics within the dataset, using a bar chart where each bar represents the proportion of the total that each topic comprises. The bars are color-coded from dark blue to pale blue to indicate the size of the percentage each topic occupies. Panels (B) focus on single topic, detailing its annual percentage out of the total for each year covered in the study. Line graphs are utilized to depict the trend of each topic over time, highlighting fluctuations and patterns in their relative importance.
• Figure 3: Popularity levels of each topic. The figure exhibits an overview of the mean view counts and temporal trends of 14 distinct topics, showing the shifting popularity levels of each topic over a span of 17 years. Panel (A) depicts the overall mean views of each topic, with each bar represents the logarithm transformed mean view counts that each topic possesses. Panel (B) show popularity trends of each topic across years, detailing the annual click-through rate for single topic. These line graphs display the dynamic changes in popularity of topics, highlighting fluctuation in audiences’ preferences.
• Figure 4: Rankings of average view counts of each topic across years. This figure displays the annual rankings of average view counts for each topic, where each coloured bar represents a different topic. The panels illustrate the rankings of all topics from 2006 to 2022.
• Figure 5: Overview of the difference index. This figure presents the annual difference indexes for each topic, along with the overall difference index per year. Panel (A) illustrate the evolving discrepancies in topic preferences between speakers and audiences over time, with each panel depicting the dynamic variance of a specific topic through the years. Panel (B) displays the annual difference index, which is the cumulative aggregation of difference indexes for all topics within a given year.