Significance of Customer Reviews to Business - Coursework Example

Customer Reviews Name: Class: Unit: Introduction With the rise of internet, customer reviews have become an important business tool. Customers use social media, search engines and websites to spread electronic word of mouth (eWOM) globally in a manner that is not easily forgettable (Filieri, 2015). Research shows that customer reviews are an important influence on the customer purchase decisions as well as attracting and retaining the customers. At the moment, there has been an increase in the number of customers shopping online. As the number of online shops has increased, it has become difficult for the customers to make their purchase decisions based on pictures and product descriptions. This has made them to turn into customers reviews which provide vital information on the products and services being sold (Elwalda & Lu, 2014). Customers are able to compare the reviews and obtain information on the experience on what they are about to make a purchase on. Despite the importance of customer reviews, there have been varying view points on their impacts. There have also been varying viewpoints whether online customer reviews offers the best channel compared to the traditional channels. In addition, online reviews have presented a problem due to bias, manufacturer keeping track of the opinions and numerous reviews received (Forman, Ghose & Wiesenfeld, 2008). This has led to the need for mining and summarising customer reviews. This paper will analyse the varying viewpoints on the customer reviews and findings from industry research and academics. The discussion will outline ways in which the current academic research on customer reviews impacts the business. In addition, the discussion will look at why the companies use mining and summarising of customer reviews and the impacts it has on business. Lastly, the paper will look at the application of customer reviews on Amazon.com and eBay and their use of mining and summarising. 1 Theory 1.1 Significant of customer reviews to business Most of the internet based business models relies heavily on their customers’ feedback. This has been aimed at building loyalty and attracting new customers. Kramer, Guillory & Hancock (2014) claim that most of online business looks at the online reviews as a vital marketing tool which is better than use of TV advertising. In addition, online reviews have been used as a tool for revenue forecasting. Despite this, its efficiency still remains a controversy. This is due to fact that functioning of the online review system is based on voluntary contributions. Online reviews have been seen as a public good since they benefit all consumers and firms. When the reviews are negative, it leads to harmful effect on adoption and diffusion of new products (Elwalda & Lu, 2014). Research shows that it is possible to transfer emotional states through emotional contagion. This is especially with the rise of social media where message can spread to large audience in a short time. This also applies to the use of online reviews. Consumers feelings about a product when expressed online can affect other people behaviours towards it. When a product or service receives a negative review, it is possible for it to affect those reading it (Kramer, Guillory & Hancock, 2014). Internet is a participatory medium which has led to an explosive growth of the internet based reviews. Despite this, most of the reviews that are posted on the online forums are not authenticated (Filieri, 2015). This makes it hard to determine the reliability of the reviews. While some of the reviews may be intentionally misleading, others are incomprehensible and frivolous. This is despite the use of community policing aimed at removing the biased and misleading reviews. This is especially on the review sites such as Trip Advisor (Ayeh, Au & Law, 2013). While some of the reviews on such sites may genuine, others may be contributed with an aim of hurting business reputation. Research shows that there are few cases where the reliability of customer reviews on the online platforms are called into question. This encourages the posting of ambiguous and unreliable reviews. There are no stringent gate keeping procedures that would ensure that the reviews are reliable. This has led to most of the review sites coming up with systems which can detect and verify fake reviews. This is through flagging fake and suspicious looking reviews. There have also been providers who are committed to verifying and publishing reviews for the business providers (Forman, Ghose & Wiesenfeld, 2008). A research done on TripAdvisor showed that the site is fairly reliable. Despite this, the research showed that there are several areas where the reviews made on the site are questionable (Ayeh, Au & Law, 2013). There are cases of rampant misleading reviews on the site. Online customers buy goods which are hard to access before experiencing them. According to Kramer, Guillory & Hancock (2014) reviews make it possible for the customers to make fast and efficient decisions. Research has shown that this leads to a boost in competition to ensure that they offer reliable and quality products and services (Lackermair, Kailer & Kanmaz, 2013). There have been cases of business which have been writing fake reviews and commissioning them to their advantage. This is aimed at unfairly boosting the site positive reviews in relative to competitors. The outcome has been misleading the customers on the product or service attributes and experience. The practice also involves suppression of negative reviews through omission. This leads to a false impression of the product to the customer and misleading to the user. The practice has led to a negative impact on the competitors and customers. Moderation of reviews may also lead to some of the genuine negative reviews not being published. This leads to consumers opinion being ignored and giving the wrong picture to those reading the reviews (Filieri, 2015). Use of these practices has been in breach of the existing trade regulation and consumer protection. We live in the age where customers are the important part of the business making it hard to ignore them. Kramer, Guillory & Hancock (2014) explains that the consumers have been empowered and are more demanding. They have the ability to make or break a business and when they lose a trust in the business, they can have devastating impacts. Customer reviews are highly trusted by the modern consumers in some cases more than expert opinions (Forman, Ghose & Wiesenfeld, 2008). In fact, the trust of online reviews has been on rise. It is important to note that most of the people who rely on the online reviews are young, wealthy and technology savvy. At the moment, customers have access to a wide range of information from the internet (Filieri, 2015). To evaluate online reviews, they are looked at as digital form of word of mouth. Although it seems straight forward that having negative reviews have damaging impact on the business, research has given conflicting findings. A research done by Zhu & Zhang (2010) on digital microproducts showed that positive messages led to high sales. This was also supported by a research done on the sales of fragrance and beauty products which showed a positive association between positive ratings and sales (Zhu & Zhang, 2010). In addition, a study done on negative reviews on box office showed that they reduced revenue, in contrast a study done on Yahoo Movie data source which showed no impact (Sun, 2012). This has also been supported by a study done on Amazon reviews. These inconsistencies call for more analysis on the impacts of the online reviews. This is rather than looking at the hypothetical direct link which shows dwindling revenues due to negative reviews (Filieri, 2015). At the moment, the debate continues on the impacts of online customer reviews and the business performance. 1.2 Mining and summarising customer reviews According to Hu & Liu (2004), some of the popular products can get thousand of reviews making it hard for the potential customers to read them. If the customer makes a decision after only reading a few reviews, they may become biased. It is also hard for the manufacturers to keep track of their customers’ opinions. With numerous online reviews, it becomes important to mine the important data from them. Most of the reviews may be long but contains only a few words which give the consumers opinion. Also, some of the reviews are biased making it hard to get an honest opinion. In addition, it is hard for the manufacturers to keep track on the numerous customers’ opinions. This has led to research on mining and summarising these reviews (Jain, Narula & Singh, 2013). To successfully mine reviews, there are sets of steps to be followed. First features of the products which customers have expressed opinions on are identified. This is done using the data mining techniques and natural language processing methods. The review opinion sentence is then identified and a decision is made on whether it is positive or negative. Lastly, a summary is made using the information which has been discovered. Traditionally review mining had been done through text summarisation where the basic points are captured. In addition, there was use of document summaries. With the advancement in methods, sentiment analysis can be used for data mining in the online reviews. Sentiment analysis is also known as opinion mining and effects (Hu & Liu, 2004). Without appropriate customer reviews summary, it is hard for the business to fully understand customer needs and act on them (Hu & Liu, 2004). It is also be hard to understand where to improve on the product by the manufacturers and customers’ rights are affected. The business reputation is at stake if the manufacture cannot read all customers opinions and act on them. The large number of reviews and lack of proper and strong regulation implies that manufacturer cannot remove fake and biased reviews (Lackermair, Kailer & Kanmaz, 2013). In addition, when the business tries to get a response on what others thinks about a product or service, they are met with an enormous data making it hard to find useful information. Based on the human limitations, it is hard to analyse large amount of data. For the business, gaining the customers feedback is important for making their manufacturing and marketing decisions. With an improvement in aspect and opinion extraction, all these issues can be solved. It becomes possible to identify and select only the important and unbiased opinions. This makes it important to have review summary in place for an organisation. Through review summary, an organisation is able to fully benefit from the customer reviews (Hu & Liu, 2004). There has been a lot of research on opinion mining. Most of this research has been based on finding the sentiment which is associated with a given sentence. There have also been researches dedicated to review summarisation (Hu & Liu, 2004). However, customers give their opinions in two main ways. The ways used are subjective and comparative sentences. A subjective sentence is based on a praise or deprecation on a product. This is unlike the use of comparative sentence which gives an opinion based on comparison between products (Jain, Narula & Singh, 2013). Opinion mining is a task of the Natural Language Processing (NLP) that analysis on how to analyse opinion on the written text. Through NLP, the web becomes a vital platform to gather information and opinions on a given subject. Customer reviews are in the natural language and in unstructured format. The task of analysing and extracting the required data is thus hard and requires use of specific techniques (Hu & Liu, 2004). To understand opinion mining, it can be looked at in three levels. The levels are document, sentence and aspect. A document level sentiment looks at an opiniated document. This is a review based on an entire page in a given document. In this case, an entire document is used for expressing a single opinion (Jain, Narula & Singh, 2013). Through sentence level sentiment analysis, an entire sentence contains a single opinion. Despite this, it is important to note that not every sentence acts as a subjective sentence. Research shows that despite the fact that mining opinion at both document levels, it is hard to determine the exact needs of the people. To gain the exact and clear opinion, it is important to use the aspect level. Aspect level analysis is also defined as the aspect based opinion mining. This works through identifying and extracting aspects of the product then extracting their opinions and determine their polarities (Hu & Liu, 2004). The review data to be mined comes mainly from online shopping sites, weblogs social media and review sites. A weblog is mostly run by an individual or a group and contains reviews made by authors on products. It might contain reviews on a product or services. A reviews site contains reviews which are given by different authors on experience based on different products and services. Each product or service has their own page or multiple pages where they are collected and published. On the online shopping site, customers are given a platform where they can place their feedbacks. This is related to the product on shopping, experience or any other opinion. Most of the popular online shopping sites rely heavily on customer reviews for their publicity. Customers can use a free format or detailed review in giving their feedback (Jain, Narula & Singh, 2013). Through an opinion summarization system, it is possible to input the product name and the web page with the product reviews and get the summary as the output (Hu & Liu, 2004). This is attained by downloading the reviews which are then stored in the database. The frequent features are extracted and opinion obtained through use of the frequent features (Jain, Narula & Singh, 2013). Through use of WordNet, it becomes possible for the system to identify the semantic orientations. The system then finds the infrequent features through use of extracted opinion words. This is followed by the identification of the opinion in each sentence and compiling a summary. Despite the importance of opinion mining in customer reviews, there are several challenges. First, there is implicit sentiment which occurs in some of the reviews. This is an objective polar utterance which may be used to refer to a negative evaluation. There are also instances where the phrase polarity is contextual based. Contextual polarity leads to a challenge when mining data and can impact the outcome (Hu & Liu, 2004). 2 Application of product reviews by Amazon and eBay Two of the most successful online companies are eBay and Amazon. eBay is one of largest online shops with its presence in over 25 countries. On the other hand, Amazon was formed in 1993 as an online bookstore and had grown its presence globally (Wang, Zhu & Chen, 2008). At the moment, consumers can purchase anything from the two online scores in different categories. The online shops are known for their low prices, selection, quality and convenience. Online business models depends highly on the customers reviews. This is due to fact that most of the consumers analyse the customer reviews before making a purchase decision. The main problem is that the number of reviews posted is high which makes it hard to analyse them individually. This has made online shops such as eBay and Amazon to utilise review mining techniques to summarise and aggregate the customers’ feedback (Jain, Narula & Singh, 2013). Both eBay and Amazon are some of the sites that contain a lot of user generated reviews. Due to the amount of data that is contained in these reviews, it is hard for the business and customers to use it fully in decision making. Amazon and eBay utilised review mining to extract the semantic orientation as well as the product aspects from customer feedbacks (Lupo, 2015). Research has proved that about 50% of the online shoppers take time to read the reviews before making a decision to purchase. In fact, it found out that about 26% of the customers at Amazon read reviews before making the final decision on purchase (Mudambi & Schuff, 2010). For Amazon, users are allowed to post reviews after making a purchase. These reviews can be accessed by anyone and also includes a star rating and a time stamp (Amazon, 2011). The system then archives all reviews made and comes up with an aggregate review which is determined by the respondent opinions. The review system that is implemented by Amazon is open and supports anonymity (Wang, Zhu & Chen, 2008). This has exposed the system to biased, malicious and unfair reviews. This has led to concerns on the credibility of some reviews given for products sold. This is a problem that made Amazon to come up with a feature where other users can evaluate the existing reviews on a product and give their opinions. The main weakness of the Amazon review system is the fact that it does not consider the aging of the reviews and can also lead to product discrimination through unfair ratings (Mudambi & Schuff, 2010). Amazon has been working hard to improve its online reviews platform (Mudambi & Schuff, 2010). This includes banning reviews that are written by customers who have been offered an incentive. This is due to fact that incentivised reviews put the system credibility into jeopardy. Some of the customers writing paid reviews may lack honesty leading to a compromised system. Incentivised reviews are only allowed through their credible system which asks only the trusted reviewers to post reviews for a fee (Leswing, 2016). eBay is one of the most reputable brand in the UK and globally mostly due to the positive customer reviews that the company has been able to gain online. Creating trust and satisfaction online has been highly promoted by use of customer review mining techniques. For example, eBay users are able to rate their customers and sellers using positive, neutral or negative feedback (Resnick et al., 2006). The online shop has an established Feedback dispute console where a seller can dispute a negative feedback. There is also a system for mutual feedback withdrawal when the involved partners agree. When the user violates eBay feedback policies such as use of offensive language, the feedback may also be removed. In addition, eBay has a third party review process using independent reviewers (Bhattacharjee & Goel, 2005). The system utilised by eBay aggregates an average on the given positive and negative reviews. The system also displays the number of the number of positive, negative and neutral reviews and time. It is important to note that eBay customer feedback system combines both seller and buyers reviews. The reviews are only separated in the comments area. Thus, when a buyer is making a decision, they are supposed to look at the merchant who has the best experience in selling based on the reviews and aggregate ratings. Items categories also determine the merchant reputation. For example, a customer may fail to buy from sellers who have no reputation or positive reviews in selling a given category of products (Lupo, 2015). An advantage of the feedback system is the fact that it can track the user based on the type of reviews that they leave. A user who leaves a lot of negative reviews on products becomes less reliable than a user who leaves an average score (Cabral & Hortacsu, 2010). In eBay, the feedback score is calculated based on a point system. For a positive rating, both buyers and sellers get one point, neutral rating is worth zero points and a negative point if one is rated negative. When all these points are aggregated, it becomes possible to determine the final feedback score (Resnick et al., 2006). To build a score, one has to receive a total of 10 ratings. There is also use of feedback percentage which determines the number of buyers in percentage who posts a positive rating with seller. Through the detailer seller ratings, it is possible for the buyers to make ratings on specific aspects such as the item description, shipping and the handling expenses. In addition, other members are allowed to comment on other customer reviews. Amazon and eBay uses the feedback system to ensure that the trading partners and customers are satisfied. Customer feedback forms an integral part of both companies where users can give feedback on their level of satisfaction on products and services. Both organisations have made reforms to their feedback systems to ensure that they are in line with the customers’ trends and reduce bias. Despite this, it is important to note that the system used by Amazon has its main focus on seller feedback while eBay looks at both buyers and sellers (Lupo, 2015). At Amazon, their customer feedback rating gives information on the merchant ratings. Both positive and negative feedbacks are utilised in giving the percentage of the total. Through use of Amazon system, it is possible to gain ratings on the past 12 months and also look at the current feedback (Lupo, 2015). Sellers who have gained a lot of positive reviews are able to gain placement advantages. At eBay, gaining a lot of positive feedback and selling a lot of items lead to a qualification in the power seller program. This makes it possible for the user to gain promotional fees among other benefits. This is due to fact that both online shops values customer feedback (Lupo, 2015). The companies know that an online marketplace can only be trusted and become safe if customers are allowed to give their feedbacks (Cabral & Hortacsu, 2010). The two online shopping companies have been able to utilise customer reviews to create trust in online shopping. Data mining has enabled the companies to gather the large amount of data contained in these feedbacks and use it to improve their services as well as inform potential buyers (Lupo, 2015). If Amazon and eBay were analysing customer feedbacks manually, it would have led to wastage of time, loss of accuracy and reduced reliability. It is evident that with the advancement of technology, companies can no longer ignore the electronic word of mouth especially through customer reviews. The online reviews have helped both companies to understand customers’ attitudes towards their products and services. This has also helped in developing the appropriate marketing strategies and eliminating rogue traders from their platforms. Online purchase involves an interaction of the buyer with an e-vendor whose authenticity is hard to determine. This risk has made the need for credible online reviews and building of trust. For the online communities, trust is built through sharing of knowledge and experience (Elwalda & Lu, 2014). In fact, most of the online customers for both eBay and Amazon trusts customer reviews than experts opinions. Conclusion To sum up, customer reviews forms an important part in modern business. This is especially with the rise of internet where the reviews can be done online. It is clear that there is a positive correlation between positive online customer reviews and an increase in sales. Customers shopping especially in the online platforms depend on reviews to make a purchase decision. Despite this, the volume of the information is high which makes it hard to read everything. This has led to the use of customer review mining techniques and summarising. At the moment, there are systems in place which helps in mining and summarising customers reviews. With the increase in the number of online shops, it has become important to look at customer reviews on products and services before buying them. This is due to fact that the buyer has to rely on an e-vendor who they have no prior knowledge or physical contact. Amazon and eBay are two online business models that have been able to utilise customer reviews to enhance their services and trust. In fact, the companies have been able to address the issue of trust which is very vital in online shopping through listening to their customers. Both online shops have customer review systems which mine data to ensure that all information is captured and analysed for user benefits. This has reduced time wastage and enhanced accuracy and reliability of the customer reviews. It has become possible for the two companies to understand their customers more and come up with new strategies. References Amazon (2011). Customer reviews. Available at: https://www.amazon.com/eBay- Inc/product-reviews/B004SIIBGU (Accessed: 1 February 2017). Ayeh, J. K., Au, N., & Law, R. (2013). “Do we believe in TripAdvisor?” Examining credibility perceptions and online travelers’ attitude toward using user-generated content. Journal of Travel Research, 52(4), 437-452. Bhattacharjee, R., & Goel, A. (2005, August). Avoiding ballot stuffing in ebay-like reputation systems. In Proceedings of the 2005 ACM SIGCOMM workshop on Economics of peer-to-peer systems (pp. 133-137). ACM. Cabral, L., & Hortacsu, A. (2010). The dynamics of seller reputation: Evidence from eBay. The Journal of Industrial Economics, 58(1), 54-78. Elwalda, A., & Lu, K. (2014). The influence of online customer reviews on purchase intention: the role of non-numerical factors. In Proceedings of the LCBR European Marketing Conference 2014. Filieri, R. (2015). What makes online reviews helpful? A diagnosticity-adoption framework to explain informational and normative influences in e-WOM. Journal of Business Research, 68(6), 1261-1270. Forman, C., Ghose, A., & Wiesenfeld, B. (2008). Examining the relationship between reviews and sales: The role of reviewer identity disclosure in electronic markets. Information Systems Research, 19(3), 291-313. Hu, M., & Liu, B. (2004, August). Mining and summarizing customer reviews. In Proceedings of the tenth ACM SIGKDD international conference on Knowledge discovery and data mining (pp. 168-177). ACM. Jain, V., Narula, G. S., & Singh, M. (2013). Implementation of data mining in online shopping system using tanagra tool. International Journal of Computer Scienceand Engineering, 2(1), 47-58. Kramer, A. D., Guillory, J. E., & Hancock, J. T. (2014). Experimental evidence of massive- scale emotional contagion through social networks. Proceedings of the National Academy of Sciences, 111(24), 8788-8790. Lackermair, G., Kailer, D., & Kanmaz, K. (2013). Importance of Online Product Reviews from a Consumer's Perspective. Advances in Economics and Business, 1(1), 1-5. Leswing, K. (2016). Amazon is banning most reviews that were written in exchange for a free product. Available at: http://www.businessinsider.com/amazon-bans-incentivized- customer-reviews-2016-10 (Accessed: 1 February 2017). Lupo, J. (2015). Comparing Amazon and eBay feedback mechanisms. Available at: https://www.feedbackfive.com/blog/amazon-feedback-vs-ebay-feedback/ (Accessed: 1 February 2017). Mudambi, S. M., & Schuff, D. (2010). What makes a helpful review? A study of customer reviews on Amazon. com. MIS Quarterly Vol. 34 No. 1, pp. 185-200 Resnick, P., Zeckhauser, R., Swanson, J., & Lockwood, K. (2006). The value of reputation on eBay: A controlled experiment. Experimental economics, 9(2), 79-101. Sun, M. (2012). How does the variance of product ratings matter?. Management Science, 58(4), 696-707. Wang, B. C., Zhu, W. Y., & Chen, L. J. (2008, December). Improving the Amazon review system by exploiting the credibility and time-decay of public reviews. In Proceedings of the 2008 IEEE/WIC/ACM International Conference on Web Intelligence and Intelligent Agent Technology-Volume 03 (pp. 123-126). IEEE Computer Society. Zhu, F., & Zhang, X. (2010). Impact of online consumer reviews on sales: The moderating role of product and consumer characteristics. Journal of marketing, 74(2), 133-148. Read More