The Future of Reputation Management is Data Capture and Analysis
Thinking back to last year, why does a massive data breach at Equifax force out the CEO but a similar breach at Yahoo is almost shrugged off? How did United Airlines recover so quickly from ‘Fight Club Class’ while Uber continues to suffer from its poor corporate culture? Why do some reputation issues become a crisis while others ebb away?
We live in a digital reputation economy. People decide where to work, who to invest with and what to buy based on a multitude of opinions. More than ever these opinions are being expressed on the internet, leaving a data trail.
Until recently, we have lacked the tools needed to access the vast amount of unstructured data about a company’s reputation for any strategic purpose.
It is widely cited by analysts that about 20% of business-relevant information originates in a structured form; information with a high degree of organisation, such that inclusion in a relational database is seamless and readily searchable by simple, straightforward algorithms. However, 80% of enterprise data is unstructured; text heavy, and information that either does not have a pre-defined data model or is not organised in a pre-defined manner. Crucially, this is where consumer opinion – informing company reputation – lies.
80% of enterprise data is unstructured; text heavy, and information that either does not have a pre-defined data model or is not organised in a pre-defined manner.
With the real-time analytics and machine learning tools now available, companies can finally tap into the rich unstructured data sources – capturing hundreds of thousands of online conversations through social media, news and online publications – to build a truly holistic picture of reputation for the first time. Data can be analysed to produce insights which enable organisations to select the most effective actions to improve customer trust and grow market share, managing their reputation in a significantly more informed and sophisticated way.
We are seeing three real world impacts where Machine Learning (ML) is applied to reputation management, the first being real people metrics. ML technology replaces misleading metrics such as Net Promoter Score (NPS) and Advertising Value Equivalency (AVE), and instead enables companies to access and analyse a wealth of unstructured data to meaningfully predict future growth, predict future customer and stakeholder behaviours, scenario plan (for example, how various crises could impact a company’s reputation), and improve the accuracy of investment models.
The second is within real world investing. The additional insight ML provides has the potential to create a step change in investment. Senior teams are becoming privy to the fact that reputation is a great leading indicator of stakeholders’ future behaviours. Meaning managing reputation needs to be included in their decision making.
Senior teams are becoming privy to the fact that reputation is a great leading indicator of stakeholders’ future behaviours.
Finally, ML is having a huge impact on real world regulations. Regulators are beginning to deploy ML to test scenarios that may impact their sectors. In particular, understanding how customers could react to a number of potential scenarios (e.g. safety failure, fake news, data breach) allows regulators to stress test whether firms have sufficient processes in place.
Until the advent of ML, corporations’ reputation has been more honoured in the breach than in the observance. Now, reputation management is a basic skill of managers, investors and regulators. If nothing else, ML enables corporations to do business better, because they will be punished more severely if they don’t.
Dr Andrew Tucker, is Vice President Data Sciences Director at Reputation Institute the world’s leading research and advisory firm for reputation.