The Effect Of Big Data Analytic Techniques

Introduction

Over the recent past, great focus has been on big data both in academia and e-commerce industries. Evidence reveals that firms in the e-commerce industry apply big data analytics to increase their value enjoy higher productivity than those firms that do not (McAfee and Brynjolfsson 2012). The internet has promoted the growth of data given the fact that today millions of people share information through internet platforms while a considerable population buy and sell products over the internet. This has promoted the growth of e-commerce by a significant margin. For example, in 2012, sales on website in the UK totaled to £164 billion while 82% of business operated official websites to market their products (Gupta, Sharma and Nayak n.d.).

The use of emerging internet offers e-commerce firms transformative benefits such as dynamic pricing, personalized offers, and real time customer service, as well as facilitating informed decisions influenced by critical insights (Jao 2013). As such, big data in the e-commerce industry plays a crucial role of enabling businesspersons track individual user’s behavior and use this information to make sure the customers make repeated purchases. Therefore, big data analytics helps firms in the e-commerce industry use data more efficiently, which drives higher conversion rates, empowers customers, and enhances decision making. According to Kwon, Lee and Shin (2014), big data analytics is essential to firms in the e-commerce industry in that it enhances cost efficiency: this can be attained by improving the quality of buyer-seller interactions. Therefore, big data analytics is a crucial necessity for

higher-performance online companies in supporting the needs of the business. To such firms, big data analytics also help identify profitable and loyal customers, detect quality problems with the offered goods and services, determine the most effective prices for various products, as well as determine the right amount of inventory to maintain. In agreement, Barrett et al. (2015) describes big data analytics as a platform for growth that every firm in the e-commerce industry should embrace. With more consumers engaging in online buying, big data analytics has emerged a necessity for many firms with the need of communicating with consumers. With big data analytics, online businesses are able to analyze the big data they generate from electronic data interchange, which helps them have a deeper understanding of consumers’ needs (Stucke and Grunes 2016). The results of such analysis can be essential when improving customer service, laying business strategies, and deciding the quality of products and services to offer customers. This study aims at exploring the effect that big data analytics has on e-commerce firms. A special focus will be paid to Amazon.com. Currently, Amazon is the world’s largest online retail shop majoring in sales of electronics, consumer goods, and household items among other items. The following objectives will guide the study:

To explore various perspectives in the definition of big data analytics
To identify the distinguishing features of big data in the e-commerce industry
To investigate different categories of big data in e-commerce
To explore the application of big data analytic techniques and their effect on firms in e-commerce industry

Theoretical Framework

Various authors have investigated the role played by big data analytics in management of online firms. A study conducted by Pavithra et al. (2016), sought to establish the various types of data held by e-commerce firms and how big data analytics helps these organizations make effective use if this data. The study findings indicated that big data analytics has an essential role in helping online firms meet the needs of their customers. Specifically, the study established that data analytics are essential for e-commerce firms when planning in performance improvement as well as increasing reliability in the services offered. Additionally, the study established that big data analytics helps online firms personalize products based on the interests and pricing values showed by individual customers. This helps online firms retain their customers and acquire new ones while enhancing brand image and customer loyalty. Further, the study reveals that big data analytics increases sales opportunities for online companies and provides an opportunity for the firms to improve the quality of their products and services. Therefore, the study concludes that big data analytics should be applied by all online firms aspiring to experience growth. Though effective at establishing the benefits off big data analytics to online firms, the study does not identify the techniques used in big data analytics, which the current study accomplishes.

Another study conducted by Akter and Wamba (2016) explored the various categories of big data applied in e-commerce industry. The study results reveal that online transactional data is among the data types managed by online firms. The study notes that online transactional data deals with selling of goods and provision of different services over the internet. The study gives examples of online firms managing online transactional data to include amazon, eBay, and Expedia. The study further established that online companies use customer browsing and transactional behavior of customers to personalize offers based on the established interest pattern. Though appropriate to the study focus, the study does not investigate how big data affects firms in the online industry, which will be accomplished in the current study.

Gupta, Sharma, and Nayak (n.d) explored the extent to which big data analytics is applied by firms in the e-commerce industry. In particular, the study explored the features of big data that distinguishes it from traditional datasets as well as various technologies that facilitate analytics of consumer data in e-commerce. The study drew from the cases of Amazon, Walmart, and Adidas. Further, the study investigated the challenges that online firms face in analyzing big data drawing from the three organizations. The study results reveal that big data has three characteristics that make it different from traditional datasets. These characteristics are volume, variety, and velocity. These characteristics of big data imply that it cannot be analyzed using traditional databases, which justifies the need for using big data analytic techniques. Among the recommended big data analytic techniques are predictive analytics, social media analytics, and mobile analytics. While concluding, the study highlights privacy as a major challenge hindering the application of big data analytics in e-commerce. Additionally, the study identifies other challenges to include training and staffing, which accrues huge costs to firms in the e-commerce industry. Given the objectives of this study matches the objectives of the current study, the current study will contribute to literature by making special focus on Amazom.com.

Research Approach

This study takes the secondary approach in that it will review existing literature to draw conclusions on how big data analytics has affected the operations of Amazon. Through review of literature, the study will define big data analytics from the perspectives of various authors, identify the factors that distinguish big data in the e-commerce industry, explore various categories of big data in the e-commerce industry and finally explore the extent to which big data analytics is applied in e-commerce and the results effects. Given the study investigates big data in the commerce industry, which includes huge quantities of transactions; it will apply a systematic approach, which will helps establish rigor throughout the study. The systematic approach will also help ensure every objective is well researched and reported.

When identifying relevant publications to be reviewed, the study will form search terms that combine the key words (big data analytics) with other terms and phrases. An example of such search terms is big data analytics AND e-commerce. The search resulted will be limited to articles published between 2010 and 2018. Additionally, only sources with an abstract will be included in the study, which implies that the key words of the generated studies will be used to determine if the article is relevant to be included in the current study. The searched databases are Proquest, Google Scholar, and Ebsco. When analyzing the collected data, the study applies the thematic analysis. Informed by the objectives guiding the study, the researcher identifies key themes under which the sources are analyzed.

Big Data Analytics Definition

Different scholars adopt various definitions for big data. According to Schroeck et al. (2012), big data is multidimensional in definition, which is informed by the varied types of data and analysis; greater information scope; novel types of media data; real time information; new data, which is influenced by new technology; social media data; voluminous data; and the newest buzz world. On the other hand, Edosio (2014) and Russom (2011) write that the widely accepted definition of big data involves three characteristics, which are velocity, volume, and variety. Other authors (IBM 2012; Johnson 2012; and Davenport et al. 2012) study the varied data sources that contribute to big data. Further, there are other authors such as (Fisher et al. 2012; Rouse 2011; and Havens et al. 2012) that focus on storage and analysis requirements of big data when defining big data. According to Gantz and Reinsel (2012), there are three outstanding features of big data, which include the data, the data analytics, and presentation of the analytics results, which helps online firms increase value for their products and services. From these definitions, five major characteristics of big data can be identified. These are velocity, volume, variety, value, and veracity

Volume can be defined as the quantity of the data and exponentially, this quantity has been increasing. Velocity on the other hand can be defined as the speed at which the data is gathered, processed, and analyzed when measured in real time. Variety is the various types of data collected within the environment and in this case in the e-commerce industry. Veracity can be defined as the reliability of the data sources while value is a representation of the strategic, transactional, and informational benefits of the big data.

There are also different definitions of big data analytics. According to Davenport (2012) big data analytics is the quantitative analysis of big data aimed at helping businesses make proper decisions. On the contrary, LaValle (2011) define big data analytics as analysis or strategic analysis that helps business maximize their value while increasing their sustainability. This implies that when analyzing big data, it must be ensured the analysis is in line with the company’s strategic goals. Further, Zhao (2013) defines big data analytics as the analyzing of big data to help the business gain competitive value, drive customer traffic, increase customer loyalty and retention, and improve sales revenue. From these definitions, big data analytics will be defined as the analysis of the data received by firms in e-commerce in order to better serve their customers while improving the quality of the offered products and services.

Characteristics of Big Data in the E-Commerce

The e-commerce industry today has a lot of data to analyze given the accessibility and usage of internet, which has promoted online buying. In order to effectively process the capacity of data received daily, online firms have embraced the use of big data analytics. The aim of this section is to explore the characteristics of big data and the effects these characteristics have on the operations of firms in the e-commerce industry.

Over the years, the volume of data online firms receive has increased (Beath et al. 2012). According to Russom (2011) big data analytics requires massive storage and entails a huge number of records, which informs the voluminous nature. Big data is always expressed in petabytes and exabytes. Further, big data is collected from various persons and therefore received in unstructured formats such as videos, audios, images, and data generated from mobile devices. The voluminous nature of big data requires online firms to have an analysis technique that will help with real time analysis so as to ensure the needs of consumers are promptly addressed. This underscores the need for big data analytics in the e-commerce industry.

Big data originates from diverse sources and hence can be structured, unstructured, or semi-structured. Additionally, big data is generated from various sources and in different formats such as texts, audios, images, log files, web tweets, and videos among others. According to Russom (2011),different analytical tools must be used informed by the wide variety in which big data is collected so as to ensure the business effectively responds to the concerns and requests of the customers. This helps in retention and acquisition of more customers. To make decisions, business may need to analyze various aspects of customer data including seasonal and regional buying patterns, customer profiles, and historical buying behavior data. Thus, the variety characteristic of big data justifies the need for online firms to apply big data analytical techniques.

According to Russom (2012), velocity refers to the frequency at which data is generated or delivered. On the contrary, Gentile (2012) defines velocity as the rate at which big data changes and how quickly this data should be used by online firms in order to increase value. Many e-commerce firms receive data at very high frequencies and must respond almost instantly to the customers to feel valued. This has resulted to the use of various techniques to ensure online firms are continually adding value to their businesses. For example, through big data analytic techniques, Amazon has been able to maintain a timely flow of new products and communication with key stakeholders (Davenport 2010). On the other hand, eBay has been able to improve the layout and features of its website through analysis of different data collected on customers. Given that different internet users are constantly sharing their views on products, online firms should have techniques that ensure this data is captured, stored, and process in the shortest time possible so the internet users remain interested in the offered services. This underscores the need for use of big data analytics by firms in the e-commerce industry.

With the different sources of data, reliability demands a rigorous verification of the sources to ensure the quality of data is high. Schroeck et al, (2012) posit that big data analysis requires high quality data so as to enhance to results of the analysis and the implication for the firms in e-commerce. In order to respond to the exact needs of customers, online firms need to verify the sources of their information. This would increase the relevance of the results of the analysis.

Categories of Big Data in the E-Commerce Industry

Online firms receive different categories of data including baskets, visits, orders, keywords, referring links, users, social data, and catalogue browsing. All these types of data can be categorized into four classes: transaction/business activity data, click-stream data, video data, and voice data. Transaction or business activity data is obtained when the business and a customer exchanges information over a period of time. The sources to this type of data are sales transactions, customer profiles that the company maintains, and occurrence of complaints by customers (Chandraskaran et al. 2013). Click-stream data is generated from online and web advertisements as well as from social media contents such as blogs, Facebook postings, and tweets among others. Click-stream data has been a great source of information that e-commerce firms use of make decisions. Video data can be defined as live data that originates from live images. Today, most e-commerce firms do not use click-stream and transaction data in isolation as it is used in association with video data. Thus, video data is considered among the most reliable source of data for e-commerce firms. Finally, voice data originates from call centers, voice calls, and customer service. Over the recent past, this data has played an essential role in analyzing consumer buying behavior and targeting new clients.

Application of Big Data Analytic Techniques: Focus on Amazon

This technique involves collection of data from social media sites and its analysis to gain new insights. According to Akter and Wamba (2016), social media sites are not mere networks that people connect in that information shared on these sites van influence customers purchase decisions. Data collected from social media sites can be analyzed using two techniques, which are text mining and sentiment analysis (Ylijoki and Porras 2016). Text mining is majorly informed by content on social media sites. This content is filtered using key words, which helps the company monitor the product under evaluation. On the other hand, sentiment analysis applies artificial intelligence or machine learning algorithm to detect sentiments about a particular product. This is essential in evaluating the quality of the products and services offered by a firm.

This is the use of historical data to predict the future behavior of consumers. This is attained through the use of statistical models and machine learning algorithms (Puri 2013). Though predictive analysis was discovered years ago, its adoption is limited because of its high costs and complexities (Ylijoki and Porras 2016). Nonetheless, for firms that are able, predictive analysis helps firms predict consumer behavior faster, which helps in deciding the amount of inventory to maintain as well as the pricing strategies to adopt.

Most of the firms in e-commerce offer thousands of products to consumers, which imply that a consumer cannot navigate through all these products in order to place their orders. The recommender system has allowed online firms quickly identify or predict products that their customers are interested in and recommend them to consumers for purchase. According to Gupta, Sharma and Nayak (n.d), there are two major algorithms that help online firms attain this: collaborative filtering and clustering algorithm.

Collaborative filtering creates a database of historical user preference, which is used to recommend products to consumers. Once a consumer accesses the company’s site, the customer is matched with the preferences database, which helps the customer discover items of interest quickly. Clustering algorithm on the other hand identifies users with similar preferences and cluster them into a group. Once a member of this group accesses the company’s website, product recommendations are made to help the customer promptly find products.

Amazon has been using the recommender system to help its customers navigate through the offered products and place orders easily. For Amazon, the product recommender system helps personalize orders based on the expressed customer experience in the offered products. These products are then adapted to suit the tastes of different customers of real time basis. To attain this, Amazon uses both clustering algorithms and collaborative filtering. The customers are grouped based on the searches they conduct as well as item to item collaborative filtering. Search based analytics uses a consumers history of purchase and rate the items, which creates a search query for consumers to find items. Take an example, if a consumer purchased a DVD called “The long road home”, the product recommender will recommend movies of similar genre, from similar author, or from similar directors. This is among the big data analytics Amazon uses to assist its customers navigate through their products. Amazon also uses big data analytics to manage the prices of its products. This involves using historical data, which may include clickstream, cookies and previous purchases to set prices for

the products or offer subsidized discounts. Though dynamic pricing has helped Amazon attract more customers and acquire new ones, it has been criticized for selling the same product for different prices, which creates a sense of discrimination among employees. Amazon’s dynamic pricing monitors the prices at which competitors offers their products and alters its prices to increase the demand for their products. This has resulted to more sales.

Implications for Firms in E-Commerce

The study results reveal big data analytics as a necessity for every firm in the e-commerce industry. Through application of big data analytics, Amazon has gained a competitive advantage over other online firms, which has made it the world’s largest online shop. Thus, other online firms aspiring to grow should effectively manage big data using big data analytic techniques. Additionally, big data analytics has helped online firms personalize interactions with consumers, which has helped them build their image and brand image. The study has also established that dynamic pricing is a technique that allows Amazon increase its sales revenue. Through this technique, Amazon studies the selling details of its competitors and lowers its prices in order to attract more customers. This implies that online firms should effective apply dynamic pricing strategies. Further, the study has established that through collaborative filtering and clustering algorithms, Amazon has been able to make navigation through the products easier for their customers, which has promoted customer acquisition and retention. Other firms in the e-commerce industry should make effective use of this strategy. Finally, the study has established that predictive analysis is least applied yet it could be beneficial in helping firma manage their inventory and pricing. Therefore, it is imperative for Amazon and other firms in the e-commerce industry to embrace this big data management technique.

Dig deeper into The Crucial Role of Business Performance with our selection of articles.

References

• Akter, S. & Wamba, S.F. 2016, "Big data analytics in E-commerce: a systematic review and agenda for future research", Electronic Markets, vol. 26, no. 2, pp. 173-194.
• Barrett, M., Davidson, E., Prabhu, J. and Vargo, S.L., 2015. Service innovation in the digital age: key contributions and future directions. MIS quarterly, 39(1), pp.135-154.
• Beath, C., Becerra-Fernandez, I., Ross, J. and Short, J., 2012. Finding value in the information explosion. MIT Sloan Management Review, 53(4), p.18.
• Chandrasekaran, S., Levin, R., Patel, H. and Roberts, R., 2013. Winning with IT in consumer packaged goods: Seven trends transforming the role of the CIO. McKinsey & Company, pp.1-8.
• Davenport, T.H., Barth, P. and Bean, R., 2012. How'big data'is different. MIT Sloan Management Review.
• Edosio, U.Z., 2014. Big data Analytics and its Application in E-commerce. E-Commerce Technologies, 1.
• Fisher, D., DeLine, R., Czerwinski, M. and Drucker, S., 2012. Interactions with big data analytics. interactions, 19(3), pp.50-59.
• Gantz, J. and Reinsel, D., 2012. The digital universe in 2020: Big data, bigger digital shadows, and biggest growth in the far east. IDC iView: IDC Analyze the future, 2007(2012), pp.1-16.
• Gentile, B., (2012). Top 5 myths about big data. [Online] Available at:
• 2012] Gupta, U.K., Sharma, B. and Nayak, M., Big Data Analytics And Its Application In E-Commerce. Havens, T.C., Bezdek, J.C., Leckie, C., Hall, L.O. and Palaniswami, M., 2012. Fuzzy c-means algorithms for very large data. IEEE Transactions on Fuzzy Systems, 20(6), pp.1130-1146.
• BM, (2012). What is big data? [Online] Available at:
• Jao, J., (2013).Why big data Is A must In ecommerce. [Online] Available at:
• 22 August 2018]. Johnson, J. E. (2012). Big data + big analytics = big opportunity. Financial Executive, 28, 50–53. Kwon, O., Lee, N. and Shin, B., 2014. Data quality management, data usage experience and acquisition intention of big data analytics. International Journal of Information Management, 34(3), pp.387-394.
• LaValle, S., Lesser, E., Shockley, R., Hopkins, M.S. and Kruschwitz, N., 2011. Big data, analytics and the path from insights to value. MIT sloan management review, 52(2), p.21. McAfee, A., Brynjolfsson, E., Davenport, T.H., Patil, D.J. and Barton, D., (2012). Big data. The management revolution. Harvard Business Review, 90(10), 61–67.
• Pavithra, B. Niranjanmurthy, M. Kamal, J. & Mani F. 2016, “The study of big data analytics in e-commerce”, International Journal of Advanced Research Research in Computer and Communication Engineering, vol. 5, special issue 2. Rouse, M., (2011). big data (Big Data). [Online] Available at:
• August 2018] Russom, P., 2011. Big data analytics. TDWI best practices report, fourth quarter, 19(4), pp.1-34. Schroeck,M., Shockley, R., Smart, J., Romero-Morales, D., & Tufano, P. P. (2012). Analytics: The real-world use of big data. NY, USA: IBM Institute for Business Value.
• Stucke, M.E. and Grunes, A.P., 2016. Big data and competition policy (p. 15). Oxford University Press. Ylijoki, O. & Porras, J. 2016, "Perspectives to Definition of Big Data: A Mapping Study and Discussion", Journal of Innovation Management, vol. 4, no. 1, pp. 69-91.
• Zhao, D., 2013. Frontiers of big data business analytics: patterns and cases in online marketing. Big data and business analytics, 43.