Key three things for data analytics


Big data is and will be one of the key words for business now and in the future. Then the question arises. What is the purpose of gathering data and analyze them?  We gather a lot of data  and analyze them in order to make a decision in our businesses. Without a decision,  no data has  meaning to  us.  Data should be gathered to make a decision.   The problem is how we can do that? Data is data, of course,  we cannot use the data effectively by just seeing massive amounts of data.  Vendors might provide us data warehouse, statistical models. But they are just tools, although they are important for data analytics.  We can not think each of them exits independently because they should be deeply interconnected each other. So I would like to present three key things to think about data analytics.   Let me explain each of them here.


1. Data

Data is a starting point about analytics.  There are many kinds of data, such as financial data,  temperature of regions and population of each country, etc.  Data has its meaning and structures behind it. Therefore, it is sometimes said that”Let the data speak by itself.”  In the beginning of analytics,  it is very difficult to identify which data has strong power to explain our observations in advance.  So, it is better we include as much data as possible to analyze.

2. Statistical models

Statistical models are interpreters to enable us to understand what the data means to us.   When we gather data, it is usually difficult to understand what the data mean just by seeing them because they are massive volume.  Therefore, we use statistical models so that we can understand what means the data has and identify the mechanism behind the data.  There are many kinds of statistical models.  We should be careful to choose the right one in analyzing data.

3. Outputs

Outputs are critically important for business managers because outputs directly impact their decision in their businesses.  Certainly outputs are provided as numbers or graphs.  All we have to do is to ensure these numbers and graph can support a decision by business managers.  Statistical models can produce only the numbers, not explanations about numbers.  Again, we should ensure the outputs of the models is what we need to make a decision in our businesses.


I always considered these three key things when I analyzed the risk of each borrower in the consumer finance company. When you wonder what are needed for data analytics in your own business,  I recommend  you to consider these key things above.  I am sure you can find what are needed to make your decision supported by data analytics.