Your business does not need to be big to get benefits from big data solution. In fact, it is not your business size but the quantity and quality of data your business possesses, business issues your company is facing and your company’s available resources that determine whether you should invest in them or not.
- Read more: Business Intelligence technologies
Table of Contents
Why your business might need them
According to Kim Stevenson, the CIO of Intel, 80-90% of organizational data are raw and unpatterned, they contain a lot of “hidden potentials” to generate business competitive advantages and the pain point is that often, we miss them. Business data is everywhere, internal and external. Internally, those data come from enterprise systems, financial transactions, sales information, communication among staff and etc.
- Read more: Big Data Trend 2017
Externally, they come from your suppliers, business partners, customers, social platforms and multiple devices around you. Today, due to their diversity, complexity and large volume, traditional software and database solution fails to process and gather them all in one place. That’s when Big Data and analytics step in. They facilitate companies with skills, tools, technologies and processes to generate actionable insights that help them shape their operation decision and build strategies, for example, regarding their customer preference or market penetration. Moreover, as organizations are becoming more data-driven, their tendency to execute those decisions will be higher than before, with higher successes.
How big is big data?
Big data is closely associated with unstructured data. Examples of unstructured data are email contents, comments on social media, blog entries, voicemails, those that do not fit into any data type and can not be managed effectively and affordably using normal relational database techniques.
Big data also works with structured data; they are data that you can put in column and row. We often hear that Big Data with a large scale of data but how large the dataset is it to be ‘big’? IDC concluded that for the data store to be ‘big’, its size has to be at least 100 terabytes, which is equal to 450 million of 200-page books. Sound scary? That’s why it exceeds the manipulation capacity of traditional tools and commonly, companies will need data scientists to try tons of different data models before they can give an insightful visual conclusion on that enormous flood of data across machines. However, from another point of view, size can not be used to define big data, in this fast-evolving world of technology, what is big today can become small tomorrow.
It ‘s easy to get into a traffic jam with big data; however, you can avoid that headache by following this route.
First destination: the “Know your problems”
Understand what you want to solve and what you want to achieve is important since later on they will help you consider your resources, select the right experts and choose relevant technologies. In addition, it is also necessary to define whether the raising issue affects directly your business strategy and can be solved with a certain budget. The common practice recommended by experts is that the company should rank and prioritize high-value business problems. Another option is to select at least one customer-focus problem and at least one operational focus problem.
Second destination: the “Self-assessment”
Where are you now with the solution? What are you trying to achieve? Will the expected results outreach what you will invest? By doing self-assessment, you will be able to identify what you already have to address the issue, such as technologies, data source, reports, dashboards, spreadsheets, etc. Those will support a lot in the later review and analysis process with solution or service providers.
The last destination: the “Build your force”
Your force will be incomplete without the participation of the top men and women. The level of executive buy-in also determines the project feasibility since they are important to project motivators and sponsor.
Next, an ideal tree of a hybrid team should have four branches. One is the business representative who thoroughly understands the use case. The second branch is an experienced project manager who is able to oversee the whole project.
Next is a data scientist who will is capable of unveiling data pattern secrets. Last but not least, it is a skilful developer that will help model and visualize data using various techniques. When selecting technology for your Big Data application, we might come to many technical questions like: “What solution should we choose”, “Which database to use: SQL or NoSQL?”, “Should we put it to the cloud?”, “Which architecture to use: MPP or ETP?”, “Should we build one or utilizing prepackaged software?” and many more. In the case your company do not have enough resource, consider selecting an experienced companion in the market to see where are the hidden potentials is not a bad idea.