The term data analytics seems pretty popular these days. It is no surprise that with the amount of data generated every single minute, analytics are now focused on handling these new big datasets and making use of them to give us insights. This is where the role of data science in big data analytics, machine learning, and deep learning comes into play. The more data you have, the more intelligent your predictions can be. After all, DOMO in its Data Never Sleeps 7.0, estimates there are approximately 4.5 Million Google searches conducted every single minute.
In actual fact, data analytics is such a generic term which means, a process to examine data in order to derive and conclude findings. There could be many techniques to analyze the data and it depends on what your goals are.
Here are the 4 main types of analytics:
1. Descriptive Analytics
This is the most basic type of analytics, often use in business analytics to identify what had happened based on a set of data. How many sales are made, what month recorded the highest sales, which region has the highest profit –those kinda questions. You can even use Pivot Tables to help you with these.
2. Diagnostic Analytics
Once you have answers to what had happened, you can move on to diagnosing why it happened. Highest sales in July, potentially attributed to rigorous summer marketing campaigns. The technique could involve drilling down and doing correlation with multiple sets of data to provide context to the findings.
3. Predictive Analytics
Enter one form of advanced analytics called Predictive analytics where it provides a glimpse of what might happen in the future. This can be done through algorithms, statistical modeling or even based on the understanding of historical trends. For example, sales prediction is estimated to be 10% lower during winter based on the historical mean.
4. Prescriptive
simulation algorithm - Action course based on what might happen
Probably the most sought-after skill and the reason why data scientists are highly paid for, this type of analytics can provide the next course of action to what might happen. This type of analytics involves leveraging on widely available data, constructing advanced simulation and machine learning to find patterns and build algorithms. For example, scheduling stock order and optimizing store inventory to maximize sales for a specific period or automatic product recommendation to customers based on historical purchases and product browsing.
Now that you’ve got a sense of the 4 types of analytics out there, take a look at the task you’re handling and see what category it falls under. To create more value in your work, perhaps it’s time to create more proactive solutions? Or perhaps you might want to begin by building a stronger foundation in understanding your data?
