Measuring Click and Share Dynamics on Social Media: A Reproducible and Validated Approach

Abstract

Online social conversations have increasingly become the means to find and view information online. In contrast to traditional web surfing models, clicks from social media result from a series of endorsements subject to user memory, past behavior and intermittently divided attention. Understanding click dynamics allows us to leverage those facets to improve relevance, forecast traffic and better manage influence in information dissemination. Unfortunately, data on clicks – even in aggregate – remain proprietary and inaccessible to researchers and scientists in many disciplines. In our work, we aim to allow the study of clicks through a proxy, allowing analysts to fully study click dynamics on Twitter. We focus on the scope of a news content publisher with a large readership and a broad domain of topics. We validate one such proxy, clicks-per-follower (CPF), based on publicly accessible data. We develop a model to compute CPI from public data. We use this method to examine how sharing affects consumption on Twitter: our findings suggest that mass retweeting of a URL does not necessarily translate into a substantial increase in clicks.