Building a holistic sentiment score for companies
Creating a Public Perception Score for Automotive Companies through Data Integration
Consumer sentiment refers to the general attitude and perception of consumers towards a company or its products or services. It can be measured through various methods such as surveys, focus groups, social media analysis, and customer feedback. Consumer sentiment can fluctuate over time and can be influenced by a variety of factors such as a company’s reputation, marketing efforts, customer service and competition. Companies with a positive consumer sentiment tend to have more loyal customers, while companies with a negative consumer sentiment may struggle to attract and retain customers. Understanding and monitoring consumer sentiment is vital for companies as it can provide valuable insights into customer satisfaction and areas for improvement.
Although there are multiple articles on the internet on how to perform sentiment analysis of data, a holistic methodology that exactly reflects the public perception of an organization is hard to come by.
In this article, we look at various streams of data collection that are available at our disposal and how we can use them effectively to come up with a holistic sentiment score for a company. The higher the sentiment score, the better the company is perceived by the public. This article presents a methodology to compute the sentiment score for companies in the automotive sector.
Different sources of sentiment data
1.Social media : The most sought after way to gather public opinion on companies is to monitor social media platforms such as Twitter, Reddit, Facebook and Intagram for mentions about the company. There are official APIs and third-party free services that enable us to access and use such data. People post reviews about cars and reactions to major steps taken by OEMs (Original Equipment Manufacturers) on social media and this data can accurately reflect people’s opinion about the companies. Public opinion on other players in the automotive sector like suppliers and dealers are hard to collect through social media data.
2. Customer reviews : Service-based offerings can be better evaluated through reviews and customer comments on car review platforms and reviews on Google. This method of data collection is useful for automotive dealers, mobility service providers, Vehicle-as-a-service (VaaS) providers etc.
3. News monitoring : Monitoring news articles and press releases about the company can provide insight into the company’s reputation and how it is perceived by the public. Several sources like Bing News, Google News, NewsAPI, Financial Times etc. provide APIs that can be used to fetch this data programmatically.
Sentiment Analysis
Sentiment analysis, also known as opinion mining is the use of Natural Language Processing and computational linguistics to identify and extract subjective information. It is often used to determine the overall tone or sentiment of a piece of text. The goal of the sentiment analysis algorithms is to convert the text into a number that reflects the sentiment of the text. There are various rule-based and machine learning based algorithms that perform this step, but in the last few years, deep learning based algorithms have outperformed the previously mentioned ones. Algorithms based on Recurrent Neural Networks (RNNs) and long short-term memory (LSTM) Networks have recently reached human-level accuracy in understanding implied meaning and sarcasms. Refer to this article for a hands-on latest way to approach this topic.
Methodology
Instead of putting all eggs in one basket, it is always a good idea to diversify. The methodology represented here incorporates this strategy. We develop scores that looks into different aspects of consumer sentiment. Then the scores are combined together to form a final sentiment score.
- Social media sentiment (Source 1 — Twitter)
- Social media sentiment (Source 2 — Reddit)
- News sentiment
- Reviews from mobility service related websites
- Area specific sentiment (Sustainability, Innovation etc.)
We also need to consider the type of company that is being evaluated. For instance, social media sentiment and news sentiment could be a good indicator for OEMs whereas reviews might be a good indicator for mobility service providers, dealers etc. We can combine the results from the above aspects in different ways for different type of companies to come up with a holistic score for each organization.
AUTO100 introduction
The automotive industry is undergoing the greatest transformation since its emergence. While some companies are not able to keep up with the changes, many companies increase their turnover despite the challenges. At the same time, new players emerge that raise their enterprise value far beyond what was earlier the norm in the industry. Previously valid mechanisms of corporate success and market value seem to have shifted. Therefore, traditional industry indices are no longer sufficient to realistically reflect company performance. What is needed is an index that reflects the entire automotive ecosystem and considers the future-oriented strategies of the companies.
AUTO100 is a digital product developed by Berylls Digital Ventures that provides an unbiased quantitative view of the automobility industry. AUTO100 uses data from a variety of sources to rank companies in the automobility space in the areas of strategy, relationship, sentiment, sustainability, people and innovation. Each company is ranked on 25 different aspects with a score for each aspect. These scores are combined together to create the final score of the company. What is mentioned in this article is just a glimpse into one such aspect.
If you are interested in checking out this stock index and the ETF based out of it, please feel free to reach out to us through LinkedIn (Clinton Charles, Malte Broxtermann, Florian Peter) or through the Berylls Digital Ventures Website. A big shoutout to Johan Torssell for the support in editing this article. Stay tuned for further updates and publications on this amazing product.
