The article presents a post-pandemic analysis of consumers’ behaviour towards purchasing products using sentiment analysis in Indonesia. It further analyses its implementations on small and medium enterprises.
The pandemic gave a lot of opportunities for a particular group of people despite a surprise in the economy and people’s mobility. Many startups have been booming since the pandemic, and most of the startups are related to online activities. Besides that, customers’ habits have significantly changed. The prediction is that some people’s behaviour will last beyond the pandemic, such as shopping online, learning new things online, and cooking from scratch. The changes in customer behaviour will have an impact not only on online businesses but also on conventional ones. A huge wave of opportunity and a change in customer behaviour is like a double-edged sword. The consequences of this phenomenon are that conventional businesses should quickly adapt, and the supply chain needs to be more agile.
The supply chain is a sequence of activities in the whole process of transporting goods and information from the point of production to the point of consumption and vice versa. On the first hand, materials are transformed into products demanded by the consumers (market). On the other hand, information (feedback) is transferred from customers to producers. Nowadays, millions of people express their thought about various services or products via social media, directly commenting on the business platform and rating online on google. Active feedback is valuable not only for companies to analyse their customers’ satisfaction and the monitoring of competitors but also very useful for consumers who want to research a product or service before making a purchase.
In the past, when someone needed a piece of information, they asked friends or family. If an organisation needed public or consumer opinions, it conducted surveys, opinion polls, and focus groups. Nowadays, if an individual wants to buy a product, they are no longer limited to taking opinions from their friends and family. This is because many user reviews and discussions about the product are available on the internet. An organisation may no longer conduct surveys, opinion polls, or focus groups because an abundance of such information is publicly available.
In general Indonesian customers have changed significantly despite differences in every business sector. Surfing online for shopping is common for all circles (groups of people). Indonesia is one of the countries with a potentially huge market share. The amount of internet and social media users in this maritime country has increased since the two years of the pandemic. A survey on the Indonesian internet users In Indonesia, the number of internet users in the first quarter of 2021-2022 was 220, compared to 175 users before the pandemic. Indonesian netizens described it as an “unsocial society”. According to Microsoft’s 2020 Digital Civility Index (DCI) survey, Indonesia was ranked 29 out of 32 countries, only ahead of Mexico, Russia, and South Africa. Many people think this is a shame, but I think this is an opportunity because Indonesian social media users tend to say something bravely and just the way they are. Indeed, it will be helpful for customer satisfaction and supply chain performance review. All feedback (customers’ opinions) contains sentiment, and sentiment analysis is one of the most popular methods to analyse attitude.
Sentiment Analysis is one of the Natural Language Processing (NLP) techniques which is used to determine whether a text is positive, negative, or neutral. Sentiment analysis can go beyond the polarity (positive, negative, neutral) of a text to detect specific feelings and emotions (anger, happiness, sadness, etc.), urgency (urgent, not urgent), and even intentions (interested or not interested). Text analytics and natural language processing are used to extract and classify data which are expressed by using different vocabularies, slang, and contexts of writing. On the tools used for sentiment analysis, the most used tools for detecting the feelings polarity are Emoticons, LIWC (Linguistic Inquiry and Word Count), SentiStrengh, Senti WordNet, SenticNet, Happiness Index, AFINN, PANAS-t, Sentiment140, NRC (), EWGA and FRN.
The result of sentiment analysis is displayed in graphs, such as pie charts, bar charts, and line graphs. Sales or even everyone in the supply chain who works on the customers’ experience management base can get sentiment analysis results as follows:
In addition to displaying results in graphical form, sentiment analysis tools often show the sources of the data collected. Figure 2 is the author’s experiment with customers’ feedback on a tech company that provides transportation services. Each company can adjust the further analysis to support decision-making based on the actual situation.
Finally, the increase in the quality of consumers’ understanding is expected to facilitate the expansion of the business world, especially micro-enterprises. Furthermore, It will help the Indonesian economy, which is in recovery and growth mode and facing a potential global recession.
Author: Ferderikus Ama Bili Author is an entrepreneur and part-time independent researcher
 Salinca, Andreea. “Business reviews classification using sentiment analysis.” 2015 17th International Symposium on Symbolic and Numeric Algorithms for Scientific Computing (SYNASC). IEEE, 2015.
 Liu, Bing. Sentiment analysis: Mining opinions, sentiments, and emotions. Cambridge university press, 2020.
 Alessia, D., et al. “Approaches, tools and applications for sentiment analysis implementation.” International Journal of Computer Applications 125.3 (2015).