Warwick Analytics has been published in the latest volume of the peer reviewed journal Applied Marketing Analytics.
The paper ‘The Coming Democratisation of Emotions Analytics’ looks at the growing volume, diversity and complexity of customer feedback data and the increasing limitations with sentiment analysis. As a result, analysts are looking for the next steps in the capabilities of text analysis. We discuss these advancements later in this paper but first we look at the current limitations and opportunities that have acted as the drivers for a more emotional approach to AI text analytics.
Whilst sentiment analysis has served text analytics well for a long time, the challenges and opportunities being presented are becoming more significant. Machine learning has promised much in the way of assisting text analytics to uncover more hidden customer data but the skill required and complexity has thus far proved a barrier to the models which can unlock true sentiment.
Now there are techniques and methodologies appearing which can democratise data science, more specifically voice of customer data, to enable business analysts rather than data scientists to turn customer sentiment not just into charts but into actions, satisfaction and profit. $3.1 trillion is IBM’s estimate4 of the yearly cost of poor quality data, in the US alone, in 2016. The Harvard Business Review5 also highlights the following statistics:
50% — the amount of time that knowledge workers waste6 in hidden data factories, hunting for data, finding and correcting errors, and searching for confirmatory sources for data they don’t trust.
60% — the estimated fraction of time that data scientists spend cleaning and organizing data, according to CrowdFlower7.
Reducing these costs requires a new way of thinking. The latest in AI Emotions Analytics looks much deeper at the origins and overall content of the data and solves many more root causes of issues. The benefits of improving data quality with this latest technology go far beyond reduced costs though. Improving data quality is a gift that keeps giving — employing the right analytics and the right level of automation vs human in the loop will become a self-fulfilling cycle producing more efficiencies every day, enabling firms to more easily pursue other data strategies.