Two data analysts smiling at the camera, holding some documents

What Are the Different Types of Data Analysis?

Emily Stevens

The most successful businesses and organizations are those that constantly learn and adapt. No matter what industry you’re operating in, it’s essential to understand what has happened in the past, what’s going on now, and to anticipate what might happen in the future. So how do companies do that?

The answer lies in data analytics. Most companies are collecting data all the time—but, in its raw form, this data doesn’t really mean anything. It’s what you do with the data that counts. Data analytics is the process of analyzing raw data in order to draw out patterns, trends, and insights that can tell you something meaningful about a particular area of the business. These insights are then used to make smart, data-driven decisions. 

The kinds of insights you get from your data depends on the type of analysis you perform. In data analytics and data science, there are four main types of analysis: Descriptive, diagnostic, predictive, and prescriptive. In this post, we’ll explain each of the four different types of analysis and consider why they’re useful. If you’re interested in a particular type of analysis, jump straight to the relevant section using the clickable menu below. 

  1. Descriptive analytics
  2. Diagnostic analytics
  3. Predictive analytics
  4. Prescriptive analytics
  5. Key takeaways and further reading

So, what are the four main types of data analysis? Let’s find out. 

1. Descriptive analytics: What happened?

Descriptive analytics looks at what has happened in the past. As the name suggests, the purpose of descriptive analytics is to simply describe what has happened; it doesn’t try to explain why this might have happened or to establish cause-and-effect relationships. The aim is solely to provide an easily digestible snapshot. 

Google Analytics is a good example of descriptive analytics in action; it provides a simple overview of what’s been going on with your website, showing you how many people visited in a given time period, for example, or where your visitors came from. Similarly, tools like HubSpot will show you how many people opened a particular email or engaged with a certain campaign. 

A screenshot taken from Google Analytics, showing descriptive analytics for a website

There are two main techniques used in descriptive analytics: Data aggregation and data mining. Data aggregation is the process of gathering data and presenting it in a summarized format. Let’s imagine an ecommerce company collects all kinds of data relating to their customers and people who visit their website. The aggregate data, or summarized data, would provide an overview of this wider dataset—such as the average customer age, for example, or the average number of purchases made. 

Data mining is the analysis part. This is when the analyst explores the data in order to uncover any patterns or trends. The outcome of descriptive analysis is a visual representation of the data—as a bar graph, for example, or a pie chart. 

So: Descriptive analytics condenses large volumes of data into a clear, simple overview of what has happened. This is often the starting point for more in-depth analysis, as we’ll now explore.

2. Diagnostic analytics: Why did it happen?

Diagnostic analytics seeks to delve deeper in order to understand why something happened. The main purpose of diagnostic analytics is to identify and respond to anomalies within your data. For example: If your descriptive analysis shows that there was a 20% drop in sales for the month of March, you’ll want to find out why. The next logical step is to perform a diagnostic analysis. 

In order to get to the root cause, the analyst will start by identifying any additional data sources that might offer further insight into why the drop in sales occurred. They might drill down to find that, despite a healthy volume of website visitors and a good number of “add to cart” actions, very few customers proceeded to actually check out and make a purchase. Upon further inspection, it comes to light that the majority of customers abandoned ship at the point of filling out their delivery address. Now we’re getting somewhere! It’s starting to look like there was a problem with the address form; perhaps it wasn’t loading properly on mobile, or was simply too long and frustrating. With a little bit of digging, you’re closer to finding an explanation for your data anomaly. 

Diagnostic analytics isn’t just about fixing problems, though; you can also use it to see what’s driving positive results. Perhaps the data tells you that website traffic was through the roof in October—a whopping 60% increase compared to the previous month! When you drill down, it seems that this spike in traffic corresponds to a celebrity mentioning one of your skincare products in their Instagram story. This opens your eyes to the power of influencer marketing, giving you something to think about for your future marketing strategy. 

When running diagnostic analytics, there are a number of different techniques that you might employ, such as probability theory, regression analysis, filtering, and time-series analysis. You can learn more about each of these techniques in our introduction to data analytics.

So: While descriptive analytics looks at what happened, diagnostic analytics explores why it happened. 

3. Predictive analytics: What is likely to happen in the future?

Predictive analytics seeks to predict what is likely to happen in the future. Based on past patterns and trends, data analysts can devise predictive models which estimate the likelihood of a future event or outcome. This is especially useful as it enables businesses to plan ahead. 

Predictive models use the relationship between a set of variables to make predictions; for example, you might use the correlation between seasonality and sales figures to predict when sales are likely to drop. If your predictive model tells you that sales are likely to go down in summer, you might use this information to come up with a summer-related promotional campaign, or to decrease expenditure elsewhere to make up for the seasonal dip. Perhaps you own a restaurant and want to predict how many takeaway orders you’re likely to get on a typical Saturday night. Based on what your predictive model tells you, you might decide to get an extra delivery driver on hand. 

In addition to forecasting, predictive analytics is also used for classification. A commonly used classification algorithm is logistic regression, which is used to predict a binary outcome based on a set of independent variables. For example: A credit card company might use a predictive model, and specifically logistic regression, to predict whether or not a given customer will default on their payments—in other words, to classify them in one of two categories: “will default” or “will not default”. Based on these predictions of what category the customer will fall into, the company can quickly assess who might be a good candidate for a credit card. You can learn more about logistic regression and other types of regression analysis here.

Machine learning is a branch of predictive analytics. Just as humans use predictive analytics to devise models and forecast future outcomes, machine learning models are designed to recognize patterns in the data and automatically evolve in order to make accurate predictions. You can learn more about the key similarities and differences between (human-led) predictive analytics and machine learning here.

As you can see, predictive analytics is used to forecast all sorts of future outcomes, and while it can never be one hundred percent accurate, it does eliminate much of the guesswork. This is crucial when it comes to making business decisions and determining the most appropriate course of action. 

So: Predictive analytics builds on what happened in the past and why to predict what is likely to happen in the future. 

4. Prescriptive analytics: What’s the best course of action?

Prescriptive analytics looks at what has happened, why it happened, and what might happen in order to determine what should be done next. In other words, prescriptive analytics shows you how you can best take advantage of the future outcomes that have been predicted. What steps can you take to avoid a future problem? What can you do to capitalize on an emerging trend?

Prescriptive analytics is, without doubt, the most complex type of analysis, involving algorithms, machine learning, statistical methods, and computational modeling procedures. Essentially, a prescriptive model considers all the possible decision patterns or pathways a company might take, and their likely outcomes. This enables you to see how each combination of conditions and decisions might impact the future, and allows you to measure the impact a certain decision might have. Based on all the possible scenarios and potential outcomes, the company can decide what is the best “route” or action to take. 

An oft-cited example of prescriptive analytics in action is maps and traffic apps. When figuring out the best way to get you from A to B, Google Maps will consider all the possible modes of transport (e.g. bus, walking, or driving), the current traffic conditions and possible roadworks in order to calculate the best route. In much the same way, prescriptive models are used to calculate all the possible “routes” a company might take to reach their goals in order to determine the best possible option. Knowing what actions to take for the best chances of success is a major advantage for any type of organization, so it’s no wonder that prescriptive analytics has a huge role to play in business. 

So: Prescriptive analytics looks at what has happened, why it happened, and what might happen in order to determine the best course of action for the future. 

5. Key takeaways and further reading

In some ways, data analytics is a bit like a treasure hunt; based on clues and insights from the past, you can work out what your next move should be. With the right type of analysis, all kinds of businesses and organizations can use their data to make smarter decisions, invest more wisely, improve internal processes, and ultimately increase their chances of success. To summarize, there are four main types of data analysis to be aware of:

  1. Descriptive analytics: What happened?
  2. Diagnostic analytics: Why did it happen?
  3. Predictive analytics: What is likely to happen in the future?
  4. Prescriptive analytics: What is the best course of action to take?

Now you’re familiar with the different types of analysis, you can start to explore specific analysis techniques, such as time series analysis, cohort analysis, and regression—to name just a few! We explore some of the most useful data analysis techniques in this guide.

Ready for a hands-on introduction to the field? Give this free, five-day data analytics short course a go! And, if you’d like to learn more about data analytics and what it takes to start a career in the field, check out the following:

What You Should Do Now

  1. Get a hands-on introduction to data analytics with a free, 5-day data analytics short course.
  2. Take a deeper dive into the world of data analytics with our Intro to Data Analytics Course.
  3. Talk to a program advisor to discuss career change and find out if data analytics is right for you.
  4. Learn about our graduates, see their portfolio projects, and find out where they’re at now.

If you enjoyed this article then so will your friends, why not share it...

Emily Stevens

Emily Stevens

Managing Editor at CareerFoundry

Originally from England, Emily moved to Berlin after studying French and German at university. She has spent the last five years working in tech startups, immersed in the world of UX and design thinking. In addition to writing for the CareerFoundry blog, Emily has been a regular contributor to several industry-leading design publications, including the InVision blog, UX Planet, and Adobe XD Ideas.