What is Data Analytics?

Data analysis is any process that helps to clean, summarize, transform, and model data to discover new and meaningful patterns that can be used to inform decision-making. Once the sole dominion of titans like Facebook and Google, data analysis is now accessible to almost anyone.

Every company has access to data, whether it’s as simple as budget and sales figures or as complex as a vast enterprise data warehouse. And even companies with relatively little data now have the potential to gather more—by optimizing their website, for example, or taking advantage of the analytics tools already built into their social media platforms.

In fact, the sheer quantity of raw digital data available today enables us to gain insights that would’ve been impossible even just a few years ago. But raw data, on its own, can’t tell us much. Before it can yield new information, it needs to be analyzed.

What Are the Types of Data Analysis?

There are as many types of data analysis and each of these kinds of data analysis will comprise several steps, beginning with determining what data is needed, collecting data, data cleaning and analysis, and ultimately, data interpretation and, potentially, visualization.

These are a few of the most prominent data analysis types:

Text Analysis

Sometimes referred to as data mining, text analysis looks for patterns in large sets of written information—the results of customer feedback surveys, for instance, or social media posts. Text analysis can help extract information about key issues hidden inside unwieldy datasets.

Statistical Analysis

This can look at both the overarching characteristics of a numerical dataset as well as meaningful samples or subsets of it. Descriptive statistics offers an overview of data to reveal potential trends (such as growth charts), while inferential statistics looks for significant correlations between data points.

Diagnostic Analysis

Diagnostics drills deeper into the insights gained from statistical analysis to determine causes—to understand why correlations exist. If customers all seem to abandon their purchases at the same point in the checkout process, for example, or if rates of illness correlate to other factors, diagnostic analysis can help identify the reason why.

Predictive Analysis

Those overarching numerics gained from statistical analysis come into play during predictive analysis, where figures are extrapolated or projected beyond the parameters of the existing dataset to help forecast future outcomes.

Prescriptive Analysis

With a deeper understanding of current data in hand, and even a prediction of what lies ahead, one question still remains: what to do next? Prescriptive analysis draws on the total insights gained and uses them to determine the best course of action in a given situation.