A Sentiment Analysis Framework for Estimating Relational Capital

Abstract

This paper continues research to create a metric for relational capital, a progressively important type of knowledge asset.  Knowledge concerning customers and how to engage with them is increasing at an exponential rate in some organizations due to customer relationship management (CRM) and related software gathering customer data.  As such, understanding relational capital as a critical piece of an organization’s intellectual capital (IC) is ever more important. But measuring the level of relational capital in a firm is difficult.  Several consultancies publish brand equity assessments every year (a closely related concept) but these usually only include the biggest and most valuable brands.  Popular metrics for studying IC either don’t break out the relational capital and other individual components (Tobin’s q) or don’t include relational along with the others reported (VAIC). In previous work, we’ve explored using sentiment analysis as a measure of brand value/relational capital, studying the results of a number of firms in several specific applications (consumer brands, tech brands, media brands, etc.).  Those results have been interesting but were also only exploratory.  At last year’s ECKM conference in Lisbon, we received a suggestion that it was time to take some of these metrics and put them together into a brand equity/relational capital predictor. This paper takes that first step, looking at some sentiment analysis results over time and evaluating their relationship with known brand equity estimates.  Activity level for brand mentions, variance in activity level, sentiment (positive, negative, neutral), influencer levels, and other potential inputs can be fed into some scenarios for predicting relational capital.  The scenarios could include both different variables and different weights, experimenting to determine if an algorithm providing the best fit to the data exists. Once this type of research establishing some benchmarks, additional comprehensive quantitative tests can be conducted in future research, validating and adapting the predictor to different industries and different circumstances.  Consequently, we should be able to develop a much better understanding of the role of relational capital and its contribution to firm success as well as ways to help organizations manage it.