A Transparent, Explainable Index of Brand and Institutional Reputation Inferred from Media and Social-Media Exposure

We present a reputation index that infers the standing of a brand or institution directly from its exposure in news media and social media, without surveys or panels. The index yields three complementary gauges on a 0–100 scale — a Media Reputation Index (MRI), a Social Reputation Index (SRI), and a blended Total — and is designed to be fully transparent, inspectable, and reproducible. Building on Zhang's (2018) media-reputation construct and Mitic's (2017) formalisation of reputation as an influence-weighted average of sentiment, we show that, once simplified, each sub-index is a prominence-weighted mean of sentiment rescaled by an asymmetric logistic function, making it independent of raw volume — a desirable property for cross-brand and longitudinal comparison. We detail the weighting scheme (source tier, mention prominence, article extent for media; audience tier and within-tier engagement for social), an asymmetric sentiment coding (−3 / +1 / +3) that encodes negativity bias and the mere-exposure effect, and companion indicators (volume, polarisation, statistical confidence). We report a calibration study on a full year of a consumer-goods brand and a blind event-study in which the largest movements of the frozen index are matched, after the fact, to concurrent real-world events across deployments in multiple markets. To our knowledge, no public, closed-form, explainable index jointly combines earned media and social media; the contribution is a documented, versioned, and citable methodology.