Sentiment Analysis

In previous installments to this series, we shared our views on the concept of macro analytics and discussed the importance of exploring alternative key performance indicators (KPIs) to develop foresights on evolving situations. This article shines light on one such alternate KPI known as sentiment analysis.

Foresight can come from any information that aids in informed decision-making about the future with a given level of certainty. As the COVID-19 crisis emerged and developed into a global pandemic, there were signs along the way to prepare as the situation evolved. Social media was generally quick to capture the negative sentiment associated with the early mortality rate, and the trend continued as the infections grew almost exponentially around the world. Countries and local governments were presented the opportunity to leverage this valuable information and be proactive about the preparations by matching decision velocity to the incoming data velocity. In situations like these, sentiment analysis can be a useful tool to analyze, understand and develop actionable strategies from the sentiments of target audience.

So, what is sentiment analysis, and how can we leverage its power to take proactive actions to help resolve or reduce the effects of evolving issues. Simply put, sentiment analysis is a branch of machine learning that seeks to study unstructured response data typically embedded in text responses and assess whether the response is positive, negative or neutral. This is inherently the same exercise as assessing whether a customer has given you a five-star rating or a one-star rating in absence of a quantitative feedback mechanism. Sentiment analysis shines when in addition to extracting negative sentiments, which can be equated to one-star or two-star ratings, the unstructured text data can be studied further to better understand the cause of negative sentiment, giving the researchers deeper insights into issues and thus potential solutions.

DHG offers sentiment analysis as a powerful tool for companies to consider where social media platforms like Facebook, Twitter, etc., are increasingly used to connect and communicate with customers. The unstructured data offered from social media sources can be analyzed using robust and mature sentiment analysis methodologies and techniques.

As an alternative KPI, sentiment analysis has the ability to inform management of what is and is not working. Expanding this scope to peer benchmarking can also drive competitive insights, helping a company to assert itself as positioned relative to its peers. The science of sentiment analysis can lead to the development of innovative new products based on customers’ needs as well as improvement of existing products driven by direct consumer feedback.

Sentiment analysis can be a great tool for marketing teams to study customer behavior and also develop outreach and targeting strategies, but it also offers a potentially robust mechanism to inform operations, finance and other facets of business processes. It is important to note that sentiment analysis does not replace any existing measures of assessing customer satisfaction, such as customer surveys. However, sentiment analysis can help in expansion of these efforts by uncovering areas that may be outside of the purview of surveys, providing timely information when measured on a frequent basis.

When combined with other traditional KPIs, sentiment analysis can provide a holistic view of the business trajectory and potential speed bumps that may occur on the way. As with any analysis, the data is always at risk of misinterpretation as well as subject to bias and other adverse selection. Hence, it is highly recommended to leverage trained data scientists so that potential shortfalls in data quality and statistical techniques are accounted for when conducting these exercises.

Knowledge can be a powerful tool, and a combination of insights from data assets can also provide foresights about what lies ahead. To learn more about sentiment analysis and other insights from data analytics, reach out to us at


Amit Arya
Chief Data Officer


© Dixon Hughes Goodman LLP. All rights reserved.
DHG is registered in the U.S. Patent and Trademark Office to Dixon Hughes Goodman LLP.