Updated: Aug 18, 2020
Thanks to popularity of electronic devices and systems, data can be collected from buying books to travelling, and from banking to playing games.
Big data is then becoming more important in decision making process. Government, multinational conglomerates and even small companies are now looking at how to utilize big data in their operation and decision making.
According to The Wall Street Journal, some policy makers even track the economic activities from big data instead of traditional statistics from government bureau.
As one of largest professional accounting associations in the globe, ACCA trains its professionals to catch up with the latest business developments and knowledge so that their skills are up-to-date and widely accepted by employers in the world.
Therefore, it structures the route to be professional qualified by including big data and data analytics in Strategic Business Leader (SBL) exam and Ethics and Professional Skills Module (EPSM).
Big data and data analytics in SBL exam covers three main themes –
How can accountants use big data and data analytics to improve business processes and implement strategy?
How big data and data analytics can provide new opportunities and present new risks for businesses?
How important is big data and data analytics in making better and more effective business decisions?
To help you mastering the knowledge and skills in SBL exam, in this article, I will first introduce big data concept and why it is important to business. It will be followed by briefing key data analytics techniques. I will cover how to apply big data knowledge in ACCA SBL exam by an illustration from past paper question in the final section.
What data is big data?
The word ‘data’ is not new to all of us. Denis Kaminskiy defines data is a set of qualitative or quantitative variables, can be structured or unstructured, machine readable or not, digital or analogue, personal or not. Some data is useful in decision-making but many of them are not well captured or not applicable in decision making process in the past.
Big data is high-volume, high-velocity and/or high-variety information assets that demand cost-effective, innovative forms of information processing that enable enhanced insight, decision making, and process automation.
The characteristics that make big data different from other data processing are volume, velocity and variety (3Vs). Here I will explain 3Vs below as they are important in big data conceptual framework.
Volume – Reliability
The first characteristics of big data is high-volume. The sheer scale of the information processed is different from traditional data analysis which demands differently at data processing and storage.
Traditional data analysis can be done by excel spreadsheet, or simple tools in a single computer. But the high-volume characteristics caused big data analytics is not easy to manage under single computer environment.
Big data includes a set of data from a wide range of sources, such as social media, website or survey, that the organisation requires to hire capable people and adequate storage and processing capacity to make use of the datasets.
In its article about big data, ACCA SBL exam team concludes the main thing of volume is additional reliability as from statistician point of views, the more data you have, the more reliable your analysis.
Velocity – Timeliness
The second characteristics makes big data different from other data processing systems is velocity. It is the speed that information moves through the system.
In the past, data analysis based on different batches of data collected in a particular day or week or even month. In the modern world, transactions are conducted and recorded in real time. Data is constantly being processed and analysed so instant feedback is possible to business leaders for decision making.
An example is retail store. The stores now capture all real time transaction data including sales and inventory in their system. In addition, other data such as weather and traffic information can be put into the system so that their customers buying pattern can be analysed in real time basis.
Variety – Relevance
The last characteristics highlighted by Gartner is variety which means the data sources and formats are varied. Traditional data formats are numerical data or text and mainly collected from transactions.
In addition to numeric and text data, big data includes audios, images and videos. It means more processing works needed to transform them into meaningful and useful information to support decision making.
As big data includes both structured and unstructured data, it helps data analysis more relevant than traditional data which mainly covers data from more structured sources.
Justin Ellingwood in DigitalOcean mentioned various organisations or industry experts have suggested more explanations on big data. Some common additions are veracity, variability and value.
Variability and value are easier to understand.
For veracity, it means the variety of sources and complexity of data processing challenge the evaluation of data quality.
Therefore, it is important business leader is sufficiently sceptical of the information coming out from data analytics. He or she needs to challenge or verify the information received.
In view of big data characteristics, single computer system is no longer able to handle the data. In order to handle the storage and computational needs of big data, computer clusters is a solution. As it is a bit technical on computer science, I won’t go over it and you could refer to an article published by DigitalOcean for details on your own interest.
Data analytics techniques
The purpose of data analytics process is to answer business questions such as –
Which products are most profitable?
What is the cheapest way to ship our products from factory to the market?
What is the optimal price to maximize company’s profit?
There are many data analytics techniques to help business leaders resolving these questions. There are essentially three types of data analytics –
Descriptive – What is happening?
Predictive – What could happen in future?
Prescriptive – How should we respond to those potential future events?
Descriptive analytics is an examination of data manually to answer a question ‘What is happening?” by traditional business intelligence (BI) and visualised in various formats such as pie charts, bar charts or tables.
Generally, descriptive analytics allows an analysis of historical information by data aggregation and data mining methods. These methods help to organise the data and visualize them into appropriate charts to identify patterns and relationships.
Predictive analytics, based on Gartner explanation, describes any approach to data mining with following attributes –
Business relevance of the resulting insights
The question to be answered by predictive analytics is “What could happen in future?”. Therefore, it uses all data, both current and historical, to forecast business activity, organisation or people behaviour and trends.
The methods applied in predictive analytics are statistical analysis techniques and analytical queries to create predictive models. The outcome is often a numerical value showing the likelihood of a particular event happening.
A degree of scepticism must be applied because the independent variables put in the predictive model sometimes are good predictors of dependent variable but they may not be the causes.
Prescriptive analytics is our third data analytics techniques introduced. Someone says it is the most powerful form of data analytics because it helps to answer a question on which course of action should be taken among all available choices.
Prescriptive analytics is related to both descriptive and predictive analytics but it is looking for best solution or ‘optimisation’. It demands advance