Business Value of Analytics Assignment Sample

• Subject Name : Accounting and Finance

Use of Churn Prediction Model in the Telecommunication Industry

Abstract on Business Value of Analytics

Churn model helps to predict which customers are likely to cancel the subscription. Customer churn which also refers to customer attrition or a customer stopping the subscription with the company. As losing customers harms the business and the cost of obtaining a new customer is relatively higher than the cost incurred to retain a customer who is likely to leave, the churn model helps to make informed decisions in such scenarios. In this report, the discussion is made on the application of the churn model in telecommunication industry to help the decision-makers in the industry in identifying and retaining the customers who are likely to stop the subscription.

Table of Contents

Overview of the Telecommunication Industry in Australia.

Drivers to Adopt Business Analytics-Churn Prediction Model

Perceived Benefits from Business Analytics Strategy.

Churn Prediction Model

Challenges Faced for applying Churn Prediction model in Telecommunication Industry.

Benefits Achieved through the implementation of Churn Prediction Model

The Drawback of Churn prediction model

Recommendations for improvements.

1. Overview of the Telecommunication Industry in Australia

The selected industry for this report is the telecommunication sector with the main focus on Australia. The development of new technology coupled with consumer demand has made Australia a very competitive market for telecommunications. Growing demand for data-driven services and the emergence of new users for telecommunication resulted in significant investment in this sector. Deloitte 2017, stated that 59% of Australians use four or more devices to access the internet. According to Australia communications and media Authority (2019), Australians’ use of fixed broadband and mobile data has increased by 175% and almost 250% respectively, since 2015, but with 93% of the data delivered via fixed broadband. These are growth rates many times faster than our economy overall. Some of the companies operating in the Telecommunication industry of Australia are Telstra, Optus, Vodafone Australia, VHA,iiNet, TPG, Vocus, and Macquarie telecom. Due to the high competition prevailing in the industry companies have to devise new strategies to avoid losing customers.

2. Drivers to Adopt Business Analytics-Churn Prediction Model

In today's world of competition, Business analytics is proven to be most efficient in terms of implementing data-driven solutions to company progress. Data analytics help the business in making informed decisions and thereby helps the decision-maker in optimizing the resources to achieve higher and sustainable revenue. The availability of data and digitalization enabled to implement data-driven solutions into the real-life problem faced by the company. Data serves as a fuel for the successful running of the business in almost all the sector. The telecommunication industry is one of the rapidly growing industries and it witnesses a large amount of data and has a large customer base. The presence of huge competition among the firms made the industry a highly dynamic and challenging. So decision making is very tough for the firms and this necessitates the firms to make use of data-driven decisions.

According to Coussement, Lessmann &Verstraeten (2017), in a dynamic and competitive market place, customers are considered to be a major asset for the business. Oskarsdottir et al. (2017), mentioned that in a competitive environment where the customer is having numerous choices of service providers and they can easily switch to service or service providers. Such customers are called Churned customers. According to Sharma & Rajan (2017), the main reason for customer churn is dissatisfaction with the product, poor quality, high price, non-availability of required features, and privacy issues. The telecommunication industry is one of the industries where customer churn is seen more often. Oskarsdottir et al. (2017), states customer churn is more in the telecommunication industry due to the presence of tough competition, saturated and dynamic market conditions, and the launching of new attractive offers.

Maria et al. (2016) stated that the use of customer churn prediction models helps the firms to increase revenue and to gain a reputation in the market. Due to advancements in data collection methods and a vast client base, the firms in the telecommunication industry have an abundant amount of data related to customers which can be used to understand the behavior of customers by using various data analytics techniques. Business analysts in the industry need to know the causes of churn and suggest the appropriate management decisions by using advanced data analytics techniques.

3. Perceived Benefits from Business Analytics Strategy

Analytics is part and parcel of the telecom industry and analytics is applied in many ways. Some of the use cases are listed below

Predictive Analytics

Predictive analytics makes use of historical data on customers and draws valuable insights from it and helps in taking an appropriate business decision. Yousef, Fahmy & Mohamed (2017), used predictive analytics to identify the node failure in advance and take precautionary measures for one of the big telecom operators in the Middle East. Use cases of predictive analytics are mentioned below:

Customer segmentation

Mass marketing is very difficult to follow in case of a highly competitive industry like telecommunication. Namvar, Ghazanfari & Naderpour (2017) stated to implement a marketing strategy that has to be designed based on a data-driven segmentation approach and the authors in their study used a two-dimensional framework to segment telecom customers based on behavior and beneficial aspects. For these segments author recommended strategies in such a way, it will increase average revenue per user and decreases marketing expenses.

Customer churn prevention

Acquiring new customers is costlier than retaining old customers. Business analytics makes use of several models to work with customer data and draw insights related to customer behavior and feelings towards a particular product and design strategies to retain the customer and thereby reduce churn.

Lifetime value prediction

By considering customer transaction details, type, and activity discounted value of future profits generated by a customer is calculated. This value will be used to classify customers into different segments.

Recommendation engines

Models are developed based on the customer's purchasing pattern, customer preferences for the products & services, and similar products & services are recommended for the customers.

Customer sentiment analysis

Based on text analysis techniques customer sentiments (positive or negative reaction) for a particular product and services are calculated. Based on these appropriate marketing strategies are framed.

Real-time analytics

Real-time analytics combines the data related to customer profiles, network, location, traffic, and usage to create a 360-degree user-centric view of the product or service.

Network Optimization

For better managing of traffic and smooth functioning of operations daily transaction data are used to make optimized and strategic decisions.

Fraud Detection

The telecommunication industry more often comes across fraudulent activities like illegal access, the use of fake profiles, cloning, etc. So the fraud detection technique is useful. The Telecom fraud detection model developed by Zheng, Zhou, Sheng, Xue & Chen (2018), was applied to detect telecom fraud in two commercial banks and resulted in the prevention of a loss amount of 10 million RMB I twelve weeks.

4. Churn Prediction Model

A churn prediction model a business analytics strategy employed by the telecommunication industry is explained in this section

The Need to Integrate Data from Multiple Business Applications or Data Sources

Telecommunication service providers store huge data related to operations, finance customers, and constituency every day. These data need to be integrated into one place and so that proper analysis can be made to draw reasonable insights. To enhance customer experience which helps in avoiding churn and improvement in Return on Investment proper insights have to be drawn from integrated data (Subramanian & Palaniappan, 2015). The authors proposed a model using logistic regression and decision trees to increase customer experience.

Lack of Visibility Into the Company’s Operations, Finances, and Other Areas

Lack of visibility in data for different segments of the company may hinder the organization's ability to use data for drawing insights. Data visibility can be made possible by proper data integration, data security, and deleting old or redundant data.

The Need to Access Relevant Business Data Quickly and Efficiently

To compete with other firms in the industry individual firm should efficiently use their data in the fastest way to expand subscriptions. New technology trends impact the telecom industry by forcing them to adopt quick accessibility of data and proper handling of data to maintain their market leadership. Proximus is one of the leading telecoms company in Belgium. It was facing huge competition from other companies. To maintain its number one position and to meet the growing demand of customers and increase the customer base it needed to move faster. For this, it introduced the new program called Excite which involves the migration of data and application to the cloud for accessing data quickly and efficiently.

Increasing Volume of Users Requiring and Accessing Information and More End-Users Requiring Analytical Capabilities

Due to the advancement in generating, collecting, and storing data, availability of data sources, online transactions a huge amount of data is gathered and it becomes difficult to avoid generating data. Billions of people are using smartphones and doing online transactions. A survey by Perrin (2015), showed 42 percent of people in America use smartphones several times a day and 21 percent of people are found to be available online always. Given this huge amount of data, the telecom industry is required to use it to draw insights that require accessing information and the analytical capability of end-users.

Rapid Company Growth or A Recent or Pending Merger/acquisition

Data analytics helps in understanding what works for business and what is not. As the saying goes what is not measured cannot be managed. If any company holds data properly then it can have many benefits. By drawing actionable insight from the data actionable strategy can be designed which helps in the growth of the company

Introduction of New Products/services

Launching new products in the market often fails which leads to a waste of time and resources. This may be due to poor communication or failure in product design. If the telecom company wants to launch a new product it can rely on predictive analytics for the successful launching of the product. A model can be build using predictive analytics and this model can be used along with business logic for the successful launching of the product.

Upgrades within the IT Environment

Upgrading helps in meeting proper data required for the analytics and this insight can be used to design marketing strategy.

5. Challenges Faced for applying Churn Prediction model in Telecommunication Industry

The biggest challenge is the use of proper tools and techniques based on the situation, data availability, and requirements. The accuracy of the technique is very important. Poor data results in poor prediction The availability of accurate data and manpower required to carry out analysis is another challenge.

To retain the customers who are ready to churn the business analyst should able to predict customer who is going to churn and plan appropriate actions that will have higher retention impact. So customer relationship managers should be ready to act and engage with these customers. This retention goal is extremely challenging.

Hashmi, Butt,& Iqbal (2013) reviewed and concluded in their study on churn prediction in telecommunication that the data unavailability and availability of noisy data ate the data quality challenges face while building the churn model. Other challenges mentioned by them were the problem in building a precise model due to the data imbalance and the presence of a large number of dimensions. Ahmad, Jafar & Aljoumaa (2019) in their study on telecom churn prediction using the Syriatel dataset encountered dataset problems like huge volume, variation in the data structure, unbalanced dataset, Presence of extensive features, and missing values.

6. Benefits Achieved through the implementation of Churn Prediction Model

Over the years the telecommunication industry has undergone major changes. A few years back it was one of the fastest-growing industry but now the industry is highly saturated and highly competitive Requirements of customers are diverse, complex, and difficult to understand. They always better services at an affordable price. So the firm needs to understand customer sensitiveness. In this context data-driven decisions helps in generating higher revenue. customer Churn is one of the problems faced by the telecom industry today. The cost of retaining a customer is lesser than the cost of bringing new customers so it is very important to find factors responsible for customer churn. So the application of the churn prediction model provides benefit to the telecom industry by identifying factors responsible for customer churn and it identifies customers who are going to churn in near future. So the firm can target these customers and make business decisions. Petkovski et al. (2016) identified factors responsible for customer churn in the telecommunication services of Macedonia by building and comparing different models. Their results predicted customer churn with high accuracy. The churn prediction model benefits the company by way of identifying churners and nonchurners.

7. The Drawback of Churn Prediction Model

Telecommunication data for churn analysis comes with more dimensions and the data will be highly imbalanced leading poor churn prediction. The majority of the Churn prediction model predicts customer churn or risk by using existing customer data which is static. A commonly used method of churn prediction is the logistic regression technique. Though these methods offer some insights and identify risky customers. They are less accurate and sometimes leads to wastage of money. Sometimes there will be commonalities between churn and non churn features. This leads to an increase in the error rate of the classification model used (Amin et al.,2018)

8. Recommendations for Improvements

To improve prediction from churn analysis advanced data analytics tools should be used. Some researchers suggested using ensemble techniques. These techniques uses a combination of classifiers and give one aggregate model. Churn model will give better prediction if the data used to predict is better. Lu et al. (2014) suggested using a boosting technique to separate the customers into two clusters and use a prediction algorithm to these clusters separately and they indicated these techniques classify churn and non-churn customers accurately compared to other methods.

The churn model should use real-time data to predict the risk of a customer churning. The use of static data may result in inaccurate classification of customers leading to wastage of resources, time, and money.

As the telecom industry comes across huge data daily. Big data analytics is a boon to draw insights and the industry should rely on real-time data analytics.

Accurately predicting customers who are going to churn is very important. So focus should be on the appropriate selection of methods which results in accurate prediction and effective action can be taken by implementing proper risk management and marketing strategies. This will ultimately result in the increasing revenue of the firm.

References for Business Value of Analytics

Ahmad,A.K.,Jafar,A., & Aljoumaa,K.(2019). Customer churn prediction in telecom using machine learning in the big data platform.Journal of Big Data,6(28). https://doi.org/10.1186/s40537-019-0191-6

Amin, A., Al-obeidat, F.,Shah, B., Adnan,A., Loo, J., & Anwar,S.(2019). Customer churn prediction in telecommunication industry using data certainty.Journal of Business Research,94,290-301 https://doi.org/10.1016/j.jbusres.2018.03.003.

Australia Communications and Media Authority. (2019). Communications Report 2017-18. Retrieved from https://www.acma.gov.au/~/media/Research%20and%20Analysis/

Coussement, K., Lessmann, S., & Verstraeten, G. (2017). A comparative analysis of data

preparation algorithms for customer churn prediction: A case study in the telecommunication industry. Decision Support Systems, 95, 27–36. http://dx.doi.org/10.

Deloitte. (2017) Media Consumer Survey 2017: Australian media and digital preferences 6th edition, Deloitte, Sydney, p 4, available via http://landing.deloitte.com.au/rs/761-IBL- 328/images/tmt-media-consumer-survey-2017-INB_pdf

Hashmi, N., Butt, N.A., & Iqbal, M.(2013). Customer churn prediction in Telecommunication a decade review and classification.IJCI,10.Retrieved from https://www.researchgate.net/publication/257920014

Lu, N., Lu, J & Zhang, G.2014. A Customer Churn Prediction Model in Telecom Industry Using Boosting. IEEE Transactions on Industrial Informatics,10(2),1659-1665.doi: 10.1109/TII.2012.2224355

Maria, O., Verbeke, W., Sarraute, C., Baesens, B., & Vanthienen, J. (2016). A comparative

study of social network classifiers for predicting churn in the telecommunication

industry. IEEE/ACM International Conference on Advances in Social Networks Analysis

and Mining (ASONAM) (pp. 1151–1158).

 Namvar, A., Ghazanfari, M., & Naderpour, M. (2017). A customer segmentation framework for targeted marketing in telecommunication.12th International Conference on Intelligent Systems and Knowledge Engineering (ISKE), Nanjing, pp. 1-6. doi: 10.1109/ISKE.2017.8258803.

Perrin, A.(2015). One-fifth of Americans report going online almost constantly.Pew Research Center, December 2015 (survey conducted July 2015).

1016/j.dss.2016.11.007.

Sharma, R. R., & Rajan, S. (2017). Evaluating prediction of customer churn behavior

based on artificial bee colony algorithm. International Journal Of Engineering And

Computer Science, 6(1), 20017–20021.Retrieved from http://ijecs.in/issue/v6- i1/32ijecs.pdfhttp://dx.doi.org/10.18535/ijecs/v6i1.32.

Subramanian, P., & Palaniappan, S.(2015).Telecom data integration and analytics-proposed model to enhance customer experience.International Journal of Conceptions and Information Technology,3(3),2345-9808.Retrieved from http://wairco.org/IJCCIT/October2015Paper4.pdf2

Zheng, Y., Zhou, X., Sheng, W., Xue, Y., and Chen, S.(2018). Generative adversarial network- based telecom fraud detection at the receiving bank.Neural Networks,102,78-86. https://doi.org/10.1016/j.neunet.2018.02.015

Remember, at the center of any academic work, lies clarity and evidence. Should you need further assistance, do look up to our Accounting and Finance Assignment Help