Challenges of Big Data in Business Marketing-Samples for Students

Question: Discuss about the Challenges of Big Data in Business Marketing. Answer: The term big data refers to the huge amounts of data (both structured as well as unstructured) which is used in business marketing and public relations on a daily basis. With the use of predictive and user behaviour analytics, big data allows businesses to extract and analyse data effectively for effective marketing. It is used to predict business strategies as well as better decision making in marketing (Katal et al. 2013). Big data analytics helps businesses to enable time reductions, cost reductions and optimized product development allowing the user to accomplish tasks related to marketing. For the past 30 years, businesses primarily used the traditional data systems to analyse data such as data warehouses and relational databases. This systems were designed to handle structured data which were organized in records using Structured Query Languages. The systems were capable of reading only 8k and 16k block sizes of data. As the programs were small, processing large volumes of data was inefficient. With the dawn of advanced analytics where statistical data was used for machine learning algorithms, the world of business marketing was introduced to predictive analytics. It is a section of advanced analytics which utilizes statistical techniques to make predictions about events which have not happened yet. Targeting advertisements, analysing behaviour of customers and flagging fraudulent transactions are some of the applications of predictive analytics (Hilbert and Martin 2015). Predictive analytics have gained momentum with big data using text analytics, classification mode ls and deep neural networking to analyse data. The concept of big data gained momentum when Doug Laney, the industry analyst articulated the concept of big data in the three Vs (volume, velocity and variety). Although most companies have the infrastructure to archive data, not many of them have the capacity to process the data due to the usage of conventional data storage systems (such as NAND storage systems). Every year, the volume of data increases substantially with new customers generating pictures and videos on a daily basis. As the sources of data increase, the volumes of data needs to be stored and analysed because these petabytes of data did not exist a decade ago. The need for distributed approach to querying and scalable storage is a challenge for IT structures. Although most companies have the infrastructure to archive data, not many of them have the capacity to process the data. For parallel processing architectures, certain databases are used such as Apache Hadoop and Greenplum for storing and analysing this massive amount of data. Hadoop was developed by Yahoo as an open source platform and utilizes the MapReduce approach that was pioneered by Google for compiling the search indexes. It distributes the data set among multiple servers (the map stage) and recombines the partial data (the reduce stage). The HDFS (distributed file system of Hadoop) is used to store data using multiple computer nodes (Katal et al. 2013). It involves loading data in the HDFS, using MapReduce operations and retrieving the data. Facebook uses the Hadoop infrastructure by storing its data in MySQL database. This data is then analysed to create recommendations based on the interests of the friends. Companies have used conventional batch processes to analyse data which is often slow and not based on real time analysis ("IBM Big Data And Analytics - Marketing And Sales: Industry Use Cases - Kenya" 2018). This works when the incoming data is slower than the processing rate and despite the delay, the data stays useful. With mobile and social applications, this process breaks down as the data streaming services occur in real time and the data stays useful only if the delay stays minimal. The data velocity is crucial as some level of analysis is required while the data is streaming. This is where the concept of big data analytics comes into consideration. Data variety is another aspect of big data. Over the past decade, data structure have evolved to add thousands of formats such as photo, audio, sensor data, documents, GPS data, PDFs and flash. The structure cannot be imposed like conventional analysis systems to keep control over the analysis. With the help of big data, companies such as Google uses smart phone sensor data to determine traffic conditions which was not possible a decade ago (Jin et al. 2015) Together the three Vs determine the analysis conditions and determine the data set which defines the main concept of big data. With the vast number of technologies in the market, enterprises harness their data with the help of a number of analytics tools that make up the big data ecosystem. The core of the ecosystem is handled by the infrastructural technologies. As the databases are getting more and more complex day by day with respect to its volume, velocity and variety, enterprises cannot relate on rational databases which captured data in mere tables and rows. Hadoop, MPP or Massively Parallel Processing Databases and NoSQL are some examples of infrastructural technologies. Unlike infrastructural technologies, analysis technologies are specifically geared towards analysing data such as Analytics Platform, Visualization platforms, Business Intelligence Platforms and Machine learning ("How Big Data Can Improve Manufacturing" 2018). Visualization platforms takes the raw data and presents it in a multidimensional visual format. Analytics and business intelligence platforms analyses data and presents it throu gh visualizations in a timely manner. The applications platform of big data takes the analysed data and presents it to end users in an optimised format. In the health sector, for example, neurosurgeons can check neurological information with the help of Mintlabs and offer diagnosis and treatment. Avansera is used in the retail sector for providing companies with the food purchasing variables (such as price flexibility). Big data analytics generates data from various sources such as streaming data, public sources and social media (Kaisler 2013). The data is often available in an unstructured format, hence proper analysis is required to make effective business decisions. As the volume of data increases, large companies struggle to find solutions to make this data useful for managing, storing and analysing it for utilization purposes. A survey conducted in 2011 states that over 55% of the organization projects that undertook big data analysis were left incomplete (Abawajy 2015). This survey was backed up by another survey which showed that most of the big data projects that were undertaken by big corporations were either incomplete or not successful. Several challenges need to be undertaken to make big data analysis optimal. The first challenge revolves around the storage of data. According to the Digital Universe Report, it is estimated that the amount of data that is stored by big IT corporations doubles every two years. The digital universe will grow from 130 Exabyte to over 40000 Exabyte (a factor of 300) from 2005 to 2020. This is approximately 5300 gigabytes per person in 2020. By 2020, the enterprises will have the responsibility for over 80% of the accessible information (Fan et al. 2014). Moreover, most of the data stays in an unstructured format which does not reside in databases. Examples of unstructured data are videos, audios and pictures which are very cumbersome to detect and analyse. The problem escalates when the data reserves combine the unstructured data from separate sources leading to errors. The challenges range from logic conflicts, inconsistent data, missing data and duplicates. To deal with this overwhelming challenge, companies resort in adopting several technologies and tools s uch as Hadoop, NoSQL databases and machine learning to go through huge amount of data. Other companies use EMC Isilcon (a clustered storage system) for better management of big data. For fast indexing, SSDs or Solid state drives can be used. Another solution is to use Cloud computing services and storage systems for storing big data. The inability to adopt different technologies poses as the second challenge for adopting big data in businesses. No matter what data is collected and stored for big data analysis, the appropriate tools and professionals should be selected properly for maximum efficiency. The ecosystem of big data analysis tools such as Hadoop is not easy to use and manage. Feedbacks should be taken seriously as it boosts engagement among the workers. The reasons for adopting the technologies and their prospective benefits should be clarified to the professionals carefully to mitigate this issue. While the tools are hailed for their ability to analyse massive volumes of data, the technology is still new and several professionals are not yet accustomed with it (Abawajy 2015). Moreover, the analysis tools require huge amounts of internal resources which make it difficult for the companies to justify using the tools instead of trying to solve the actual data problem. As the job is multidisciplinary, it i s important for data scientists to have varied skills. Another problem that is often faced by companies is the scarcity of data scientists to analyse the amount of data being produced (Fan et al. 2014). Many companies often overcome this problem by providing their own training resources and assigning the management team. The third challenge is based on the quality of data. As big companies employ data analysis tools for better decision making, efforts should be made to determine whether the data is accurate or not. Testing should be made a priority for ensuring the high reliance on data. Often companies get data from several sources which do not always agree with each other. For example, the numbers in the Enterprise Resource Planning (ERP) and the sales figure of the ecommerce system might be different (Liebowitz and Jay 2013). Data governance is a process by which the records and data sources are made sure that they are accurate and secure. To oversee data governance challenges, certain policies and procedures are required. Companies can invest in data management to increase accuracy of data and simplify data governance. This process is often time consuming and requires expensive tools to prevent ill-advised decisions. The fourth challenge is a big concern for companies who endorse big data stores. Securing the complex volumes of data is necessary for ensuring that the data does not fall in the wrong hands. APTs or Advanced persistent threats and hacker attacks are common in companies with big data reserves. Most companies fail to understand that the measures they employ for making the data secure are not adequate. According to a survey by IDG, more than 40% of the people interviewed agreed that proper security measures (such as data encryption, data segregation and access control) were not ensured for the big data repositories (Fan et al. 2014). The tools that analyse the data collects information from various sources making it highly vulnerable to cyber-attacks. As more and more data are produced, the vulnerability risks increases simultaneously ("Tata Consultancy Services | Technology, Digital Solutions, Consulting" 2018). Moreover the data sources are often internal (such as marketing and finan ce) as well as external (social media data). Companies need to figure out a solution around this to make big data analytics safe and secure. The fifth challenge rises from organizational resistance. According to a survey conducted by NewVantage Partners, more than 85% of the surveyed individuals agreed that the companies are adopting a data driven culture and over 30% agreed that they had been successful in their attempts already. The surveyed people mentioned three obstacles primarily for adopting the cultural shift (Chen, CL Philip and Chun-Yang Zhang 2014). The first is organizational alignment that is insufficient. The second obstacle is the lack of understanding of the middle management teams. The third obstacle is the lack of understanding of the business procedures. To adopt big data in companies, strong leaders are necessary for understanding the possibilities of the opportunities that comes with big data. People familiar with the entertainment industry might know how Netflix used big data to ensure success of its TV series by predicting user behaviours. Since its debut in 2007, Netflix had six years to gather information for making its first original production (House of Cards) a sure-shot success. They created certain data points (called events) to analyse when a user watched shows, which devices they used, whether they re-watched any shows, searches and ratings ("Netflix United Kingdom Watch TV Programmes Online, Watch Films Online" 2018). These data was gathered by the data scientists to create certain insights that was utilized by Netflix to take informed decisions. Despite the challenges faced by it, big data makes sense for impacting the economic development. With the help of big data, companies can choose a suitable location for promoting their businesses. This is done by comparing demographics, information about labour force, consumer spending, GIS maps, talent pool data and industry data ("Intel: Tablet, 2 In 1, Laptop, Desktop, Smartphone, Server, Embedded" 2018). To make critical decisions, big data helps to give website users important data to search for locations. To prosper economically, time management is important. Going through huge amounts of data can be time consuming and overwhelming. To stay competitive, big data helps to offer reliable and up-to-date information in real time providing the data handler ample time to focus on other works. Big data also benefits stakeholders by providing the analysis of data in an organized and user-friendly format giving them an overview on how much investment they should make on a particular org anisation. Though there are several challenges ahead of big data, the technology needs to be given a chance it deserves. From training to recruitment and from budgeting to strategizing, big data comes with a number of challenges as well as possibilities. Just like internet, big data has the capability to change the world. There are several successful cases of big data that proves this otherwise and chances are that big data will be implemented by thousands of users with the growing number of unorganized data in the near future References Katal, Avita, Mohammad Wazid, and R. H. Goudar. "Big data: issues, challenges, tools and good practices." InContemporary Computing (IC3), 2013 Sixth International Conference on, pp. 404-409. IEEE, 2013. Kaisler, Stephen, Frank Armour, J. Alberto Espinosa, and William Money. "Big data: Issues and challenges moving forward." InSystem sciences (HICSS), 2013 46th Hawaii international conference on, pp. 995-1004. IEEE, 2013. Chen, CL Philip, and Chun-Yang Zhang. "Data-intensive applications, challenges, techniques and technologies: A survey on Big Data."Information Sciences275 (2014): 314-347. Fan, Jianqing, Fang Han, and Han Liu. "Challenges of big data analysis."National science review1, no. 2 (2014): 293-314. Liebowitz, Jay, ed.Big data and business analytics. CRC press, 2013. Hassanien, Aboul Ella, Ahmad Taher Azar, Vaclav Snasel, Janusz Kacprzyk, and J. Abawajy.Big Data in Complex Systems. Springer, Heidelberg, 2015. Yin, Shen, and Okyay Kaynak. "Big data for modern industry: challenges and trends [point of view]."Proceedings of the IEEE103, no. 2 (2015): 143-146. "Netflix United Kingdom Watch TV Programmes Online, Watch Films Online". 2018.Netflix.Com. https://www.netflix.com/sg/. "IBM Big Data And Analytics - Marketing And Sales: Industry Use Cases - Kenya". 2018.Ibm.Com. https://www.ibm.com/big-data/ke/en/big-data-and-analytics/marketing/industries/index.html. "How Big Data Can Improve Manufacturing". 2018.Mckinsey Company. https://www.mckinsey.com/business-functions/operations/our-insights/how-big-data-can-improve-manufacturing. "Tata Consultancy Services | Technology, Digital Solutions, Consulting". 2018.Tcs.Com. https://www.tcs.com/. "Intel: Tablet, 2 In 1, Laptop, Desktop, Smartphone, Server, Embedded". 2018.Intel. https://www.intel.in/. Jin, Xiaolong, Benjamin W. Wah, Xueqi Cheng, and Yuanzhuo Wang. "Significance and challenges of big data research."Big Data Research2, no. 2 (2015): 59-64. Hilbert, Martin. "Big data for development: A review of promises and challenges."Development Policy Review34, no. 1 (2016): 135-174.