To simply understand and answer this question, I believe we should ask what can we understand from the term “Big Data”? Today, Big Data is often described as large volume of data sets where traditional data processing techniques and tools were not capable to manage and measure them. These massive volumes of data will only be useful if you are able to use it effectively and efficiently to address problems that from the past, you would not be able to resolve.
Doug Laney introduced the 3Vs concept in his research publication in 2001. To be considered and has the characteristic of a Big Data, it is believed that the dataset must have this “3Vs” (volume, variety and velocity). Volume refers to the amount of data, variety refers to the number of types of data and velocity refers to the speed of data processing.
Recent years, value and veracity were added to account for unstructured data vs structured data sources. Value refers to economics value of the data and veracity refers to the quality of data.
Today Data has true value and people is trying all means to understand and discover how such value can be created and form an asset for the stake holders. Recent technological breakthroughs have reduced the cost of data storage and possible usage of high-end computation power at a smallest fraction of a cost as compared to Mainframe era at 1950s.
Bon Auxilium looks at Big data in two aspects. 2Cs, Computer and Connectivity. Computer to provide computing power for digitization so that data can be stored and Connectivity as a platform to transfer data in interconnected networks. Due to these 2Cs, Big data will be given a life to infiltrate every part of our daily life and overshadow us with enormous sea of data.
We believe in collecting, managing, categorise, store, analyse, predict and monetisation of your data. There are different types of methodology being practise across this industry but we believe Best practises should follow these three key components:
When doing data analysis, only full and complete analysis of all datasets must be done to tell a full story. Sampling method is not recommended even at quicker throughput that will save vendor overhead cost will come at the expense of customer interest.
Overly focus on one data value may distorted the concept or meaning of information. That is why subject/ domain experts are always a part in our process to interpret data or information.
Data “talk” to each other and it is not a zero-sum game. Isolation or limit the dataset dependent on one another will present very “surface” description and analysis of information. E.g. Customer Satisfaction vs Engagement survey has stronger correlation than Customer Satisfaction vs No. of Companies that the employee worked. Without identifying the relationship, the strength of interpretation will be kept at minimum.
We always associate Big Data with Data Analytics and we often divide them into four type of analytics (Descriptive analysis, Diagnostics analytics, Predictive analysis and Prescriptive analytics) base on complexity and value. So, while Business Intelligence focus on “What happened and what should be changed?” (Descriptive analytics). Data science focus on “Why it happened and what can happen in future?” (Predictive analysis) and then recommending what actions to take? (Prescriptive analytics).
Having to say that, the objectives of Big Data analytics provide information of valuable insights and enabling such information to be used to conceptualise model for decision making. This can be further improved in value creation for maximizing budget spending and precise execution of actions on products and solutions. This has to be aligned to set the expectation during customer engagement.
We do find out that one of the main dangers of doing analysis (common practise without guidance) was when a lot of times and much of the resources were invested during Data Science lifecycle process, Data scientist failed to address a clear storyline through impactful visualisation. Visualisation is utmost importance and necessary. Professional Marketer from Bon Auxilium would be able to provide value added services to provide an end to end Big Data Analytics experiences to our clients.