What are the four types of data analytics and how do you use them?

We all know how valuable data is in today’s economy, but in reality, hardly any value can be extracted from raw data sets. And to transform your data into an asset, you need to implement various data analytics techniques.

Data analytics includes a wide range of different processes, so let’s look at the different types.

What are the 4 types of data analytics?

In general terms, the types of data analytics can be thought of as layers that make up a cake. Each subsequent layer placed on top builds upon the work of the last.

1. Descriptive analytics

Descriptive analytics is the most fundamental type of data analytics, and only describes the process that has happened so far – not the reasons behind it or causal relationships between variables. For example, years worth of transactional data analysis could show that there are higher sales for a particular product in the months of October, November, and December.

2. Diagnostic analytics

Descriptive analytics is used to explain the reasons behind changes and patterns in data, and can be used to establish causal relationships. This could involve comparing concurrent trends and discovering correlations between variables. For example, higher sales during October, November, and December may be explained by the fact they coincide with the festive shopping period, when customers are more likely to buy a particular product as a gift.

Diagnostic analytics often involves two stages: ‘discoveries and alerts’, and ‘queries and drill-downs’. Discoveries use datasets that have already been collected, such as identifying the app with the most downloads, while alerts warn of potential issues before they occur, such as server downtime in which e-commerce may be affected. Queries and drill downs investigate past data to draw conclusions, such as exactly how infrastructure disruption causes lower sales.

3. Predictive analytics

Predictive analytics uses data to forecast future occurrences. Just as diagnostic analytics builds upon descriptive analytics, predictive analytics uses the conclusions from diagnostic analytics to create statistical models that make predictions. You can then use these models to design strategies based on likely scenarios.

For example, using the magnitude of sales uplift that has occurred between October and December over several years, you can predict the magnitude of uplift for the upcoming year.

4. Prescriptive analytics

Prescriptive analytics is the top of the layer cake, and is perhaps the most useful type of data analytics. Given likely scenarios, it uses variables and predicted outcomes to suggest actions. Optimization and random testing are used to discover how potential actions affect results, and identify those with the highest chance of the desired outcomes.

For example, using the predicted sales volume between October and December, you could forecast the server load, number of customer service agents, and quantity of stock needed to fulfill orders.

And what about cognitive analytics?

If descriptive analytics is the simplest form of data analytics, cognitive analytics is the most complex. This is the proposed fifth stage of data analytics, going beyond prescriptive analytics.

Cognitive analytics is an anticipated paradigm shift that involves applying human-like intelligence techniques, such as semantics, AI, and machine learning, to software. This enables the models to learn from both data and interactions with humans to become more effective over time.

Arguably, cognitive analytics falls under the term ‘advanced analytics’. The difference between ‘analytics’ and ‘advanced analytics’ is that advanced analytics includes the application of automation and AI to more complex datasets, ultimately producing more accurate results, and more sophisticated insights.