Predictive Analytics: Digital Metrics for Estimating Knowledge Assets

Abstract

Digitization, big data, and follow-on metrics have burgeoned over the last couple of decades.  While the knowledge management (KM) community has embraced some of the advances in data-driven decision-making in business, there remain new applications that are relatively unexplored.  One of these applications is the use of digital metrics to estimate knowledge holdings or intellectual capital (IC).  In a discipline still at the mercy of troublesome metrics, new opportunities to better measure IC would fill a large gap in the existing body of knowledge. Previous work, across several industries and well-known brands, established digital media variables of interest such as volume of mentions, variability of mentions, influencer quality, brand sentiment, and some more platform-specific (X, Facebook, etc.) measures (Erickson, 2023; Erickson & Rothberg, 2023; Erickson, Schmidt & Rothberg, 2020).  These indicators were compared with brand equity from a separate source and methodology, establishing apparent links.  In short, certain digital media variables seemed related to higher brand equity, a proxy for the intellectual capital component of relational capital. More recent research explored the statistical link between digital metrics and brand equity (relational capital) (Erickson & Rothberg, 2024).  On a very small sample, key indicators were shown to predict brand equity values at a very high correlation using different approaches (regression, neural network). This study continues that work, with a larger set of firms.  The firms are drawn from two industries, autos and information technology, both of which have a number of companies with available brand equity estimates.  Moreover, the study adds price/book value ratios, calculated on an annual basis (like brand equity) but which can also be easily estimated for shorter time periods by altering share price.  As a result, in estimating correlation, periodic digital media independent variables can be compared not only with annual brand equity metrics but with matching periodic price/book ratios.  Price/book is an effective proxy for not just relational capital but all intellectual capital of the firm, providing a different approach from a new perspective.