№1. Descriptive Analytics (What Happened?)
The purpose of descriptive analytics, as the name implies, is simply to report what happened in the past. It doesn’t try to explain why something happened, and it doesn’t try to build cause and effect relationships. The main goal is to present a digestible picture.
Google Analytics is a fantastic example of descriptive analytics in action. It gives you a snapshot of what’s going on with your site. For example, how many visitors you’ve had over a period of time or where they came from. Similarly, systems like HubSpot will show you how many people opened a particular email or participated in a campaign.
But then there are two main methods that come into play in descriptive analytics; data aggregation and data mining. The process of collecting and presenting data in an aggregate format is known as data aggregation.
As a result, descriptive analytics compresses huge amounts of data into a clear basic summary of what happened. As we will see later, this is often the starting point for deeper analysis.
№ 2. Diagnostic analytics (Why did this happen?).
This is the type of analytics that tries to find out why something happened by digging deeper. The main goal of diagnostic analytics is to find anomalies in your data and respond to them. For example, if your descriptive analysis shows a 20% drop in sales in March, you’ll want to find out why. Diagnostic analytics basically helps you do that.
Diagnostic analytics applications.
Using this type of analytics, the analyst looks for any new data sources that might provide more information about the causes of the sales decline. They may go further and discover that despite a large number of website visitors and a large number of “add to cart” actions, only a small percentage of visitors actually make a purchase. Further investigation may reveal that the majority of customers dropped out during the time they entered their shipping address.
№3. Predictive analytics (What happens in the future?)
Predictive analytics seeks to anticipate what will happen in the future. However, this is possible based on past patterns and trends to help estimate the likelihood of a future business event or outcome.
Predictive models literally create predictions based on relationships between a set of variables. For example, you can use the correlation between seasonality and sales numbers to predict when sales will drop. So, if your predictive model predicts that sales will drop in the summer, you can use that information to create a summer-themed advertising campaign or cut spending elsewhere to make up for the seasonal drop.
On the other hand, you may be running a restaurant and want to know how many takeout orders you’ll get on a typical Saturday night. The results of this type of analytics can help you decide whether to hire an additional delivery driver.
№ 4. Prescriptive analytics (What’s the best course of action?).
To help determine the best course of action, prescriptive analytics examines what happened in the past, why it happened, and what might happen in the future. In other words, prescriptive analytics explains how best to take advantage of descriptive, diagnostic, and predictive analytics.
Nevertheless, it is the most complex type of analysis. This is because it involves many things, including machine learning algorithms, statistical approaches, and computational modeling procedures.
Essentially, a prescriptive model evaluates all the different choice models or paths a corporation might take, as well as their likely consequences. This allows us to visualize how each set of decisions might affect the future, as well as quantify the impact of a particular decision. Moving forward, the organization will be able to determine the best routes based on every conceivable scenario and consequence.
Predictive models are similarly used to calculate all the different “routes” a corporation can take to achieve its goals; with the best option in sight. And knowing which actions to take for the best chance of success is a tremendous asset to any company. So it’s not surprising that prescriptive analytics plays such a big role in business.