Data analytics and its different terms can be confusing, especially for beginners. Here's a beginners guide to explain the the 4 main types of data analytics:
Descriptive Analytics (What Happened?)
Let's say you work at an online shop that sells games. As part of your job scope, you create and present a report summarising the sales for Quarter 1 (i.e. January, February and March).
Sound familiar? Most, if not all of us would have done this at some point in time, but did you know that what you just did was descriptive analytics?
Descriptive analytics is the use of data to identify patterns and trends. As one of the simplest forms of data analytics, descriptive analytics is one of the most popular, and is well-used across all functions and industries.
Diagnostic Analytics (Why Did This Happen?)
While creating your report, you visualise your data using charts and graphs and notice that even though there was no change in your sales and marketing strategy, sales in March was much higher than in January and February.
You want to know why, so you go through your data again and notice that sales significantly increase in the months of March, June, September, November and December.
From this, you infer that the increase in sales is correlated with the school holidays.
This is known as diagnostic analytics, where you identify correlations based on patterns in your data. By better understanding how each variable is connected, you get a better understanding of your data, giving you a better chance at replicating - or even surpassing - successes and avoiding mistakes.
Predictive Analytics (What Will Happen?)
Predictive analytics is the use of data to forecast future outcomes. Perhaps the most well-known function of data analytics, predictive analytics is well integrated into our daily lives - without it, we would not have access to incredibly accurate weather forecasts and other services.
Back to our example of your online shop. There's a new console game coming out soon, and you want to use the increased traffic to your website to promote other games available on the same console.
You go back to your data, and notice that new games tend to sell out quickly, and bundle sales and discount coupons are incredibly effective at increasing sales.
With predictive analytics, you can answer the following questions:
Which is more effective at increasing sales: A 5% or 10% discount?
Would the bundle be more attractive if I have 4 or 5 games?
Will I have higher profits and higher brand loyalty if I open preorders for customers with an account on my website?
In short, predictive analytics increases your chance of success by forecasting potential outcomes, allowing you to choose the path with higher likelihood of success.
Prescriptive Analytics (How Do You Reach Your Objective?)
You have your objective (to increase sales on games of the same console), but you don't really know what to do to reach that goal.
You decide to use machine learning and artificial intelligence to perform prescriptive analytics, which tells you that sales will increase by 30% and customer loyalty will increase if you send a follow up email with recommendations of similar games 1 month after their original purchase.
This is known as prescriptive analytics, which suggests the best course of action based on historical data and different variables.
Prescriptive vs Predictive Analytics
Prescriptive analytics and predictive analytics are often mixed up, and can be challenging to differentiate. We'll use another example to illustrate their difference.
You want to drive from point A to point B. The night before, you open up your preferred traffic app to check the estimated travel time - an hour by route 1. The app uses predictive analytics to forecast your estimated travel time.
On the actual day, you open up your app again to set your driving route. Halfway through, the app suddenly changes your route due to a traffic jam in route 1. This time, your app is using prescriptive analytics as it changes your route in real-time to minimise driving time.
Descriptive analytics reports what's going on (i.e. trends and patterns).
Diagnostic analytics identifies the cause of these trends.
Predictive analytics forecasts future outcomes.
Prescriptive analytics uses these forecasts and different variables to identify the best course of action.
Want to find out how to use each type of analytics?
Learn about the data analytics tools needed to perform descriptive, diagnostic, predictive and prescriptive analytics in the Wiley Certified Data Analytics course. Develop your data analytics skills and gain a Wiley endorsed data analytics certificate in only 11 classes. Sign up today!
Written by: Chloe Thio (Gen Infiniti Academy)