What is Data Analytics?

Broadly defined, data analytics is a field concerned with collecting, organizing, and analyzing data to gain insights and make informed decisions. Data analytics offers a variety of tools and techniques that can be applied to a wide range of fields and industries, with applications ranging from improving efficiency and productivity to identifying new opportunities for growth and innovation.

There are four basic types of data analytics within the larger field: descriptive analytics, diagnostic analytics, predictive analytics, and prescriptive analytics.

Descriptive Analytics

Descriptive data analytics involves using data to understand and describe what has happened in the past. One of the most accessible types of data analytics, descriptive analytics requires only rudimentary statistical software like Excel to manipulate and visualize data so that key performance indicators (KPIs) can be assessed and trends can be identified.

In retail, descriptive analytics is used to calculate year-over-year growth in sales, which can help a company better understand its performance and identify any patterns and trends in consumer demand. 

Diagnostic Analytics

Diagnostic data analytics is the process of using data to understand why something happened, which is why this type of analytics is alternatively called root cause analysis. To understand the root causes of trends identified in descriptive analytics, diagnostic analytics will go deeper into internal data sets to gain a more nuanced view of a trend or look to external data sets that might suggest correlations between a business’s performance and environmental factors. In the latter case, data mining might be employed in which machine learning algorithms search big data sets for patterns and associations.

To continue with our example from retail, if descriptive analytics identifies year-over-year growth in sales of certain products, diagnostic analytics might then be employed to better understand whether there are links between the kinds of products that are performing well and broader customer sentiment, perhaps by mining social media data.

Predictive Analytics

Predictive data analytics involves using data to make predictions about what may happen in the future. Predictive analytics makes extensive use of machine learning and other statistical techniques to build models that output forecasts from inputs of historical data. Predictive analytics is especially useful for risk assessment and identifying new areas of opportunity.

After determining which kinds of products perform well given a certain customer sentiment through diagnostic analytics, our example retailer might employ predictive analytics to better understand the chances of similar kinds of customer sentiment recurring in the near-, mid-, and long-term.

Prescriptive Analytics

Prescriptive data analytics involves using data to determine the best course of action for the future. As with predictive analytics, machine learning is crucial for prescriptive analytics. With complex machine learning algorithms, a data analyst or data scientist can turn large amounts of data into concrete recommendations for future action — or even automate actions to occur in response to certain inputs.

In retail, a common place you’ll find prescriptive analytics at work is in online product marketing. While predictive analytics can predict the likelihood of certain kinds of customer sentiment returning in the future, prescriptive analytics can employ data mining to identify current consumer sentiment and automatically send personalized marketing emails to segments of a retailer’s customer base it thinks is best poised to respond given its sentiment analysis.

Now that we’ve covered the four main types of data analytics, we’ll move on to discussing how data analytics compares to two fields it’s frequently confused with: business analytics and data science.

In today’s data-centric climate, a frequent source of confusion is the overlap between data analytics, business analytics, and data science. Here’s how we think they compare:

Data analytics vs. business analytics

A search for "data analytics vs business analytics" will yield numerous articles claiming to provide a clear distinction between the two fields. However, it’s often the case that the work labeled as business analytics or business intelligence is similar to what data analysts do. This is particularly true in smaller companies where employees may have a range of responsibilities.

At larger companies, the work of a business analyst may shift to focus on finding ways to increase revenue, decrease costs, expand the market, or enter a new market, while a data analyst may focus more on the daily operations of the organization. In this situation, a business analyst or business intelligence expert needs to have a strong understanding of business. However, even at large enterprises, these roles may not always be clearly defined.

For more, check out our deep dive into the relationship of data analytics and business analytics or the differences between a data analyst and a business analyst.

Data analytics vs. data science

As with business analytics, there is substantial overlap between data analytics and data science. That said, we believe there is a meaningful difference:

Data analytics involves the use of tools and techniques to analyze data. These professionals may work with existing databases and use existing tools and techniques for analysis.

Data science involves the creation and implementation of new tools, techniques, and methods for collecting and analyzing data. This often involves advanced techniques, such as machine learning, a subdiscipline of artificial intelligence. Data analysts and other data analytics professionals may use methods and tools developed by data scientists. If you are interested in data science, you may want to check out articles on the topic:

• Data science career path

• Data science certificate programs

• Data science bootcamps (Online)

• Data science master’s programs

• How to get a job in data science

If you’re interested in reading more about the difference between data analytics and data science or how a role as a data analyst compares to a data scientist, check out our articles on those topics.

Above, we began looking into the kind of work done by data analysts and other data analytics professionals. In the next section, we’ll dive deeper into the specific data analytics techniques they employ on a daily basis to produce actionable insights for their companies.

Regression analysis

Regression analysis is a kind of data analysis that uses independent, known variables to predict the outcome of dependent variables. In linear regression, independent variables are used to predict the outcome of a single dependent variable.

Regression is a crucial technique for proving or disproving hypotheses about whether two things are correlated or exist in a causal relationship. Say, for example, a motorcycle insurance company wants to better understand if having more years of experience driving a car might impact the likelihood that a motorcyclist gets into an accident to set competitive rates. Regression analysis could be used to determine if a relationship exists and, if so, the nature of this relationship.

For complex forms of regression analysis, data analysts will often use supervised learning, a form of machine learning that uses labeled, structured data sets — where each piece of data is tagged and classified — to train a machine learning model to give a precise and accurate output when fed an input.

Factor analysis

Factor analysis is a kind of data analysis that reduces complexity in a data set by expressing a large number of variables as a smaller number of “factors” that incorporate commonalities identified across the variables.

Factor analysis is an important tool when searching for an actionable signal within a complicated or noisy data set. This is particularly useful for data sets like those produced by quantitative or qualitative polls. Take, for example, a website that wants to aggregate online product reviews. Here, factor analysis could help to take the reviews and numerical scores available in different places on the internet and reduce them to one customer satisfaction score.

Cohort analysis

Cohort analysis is a kind of behavioral analysis that breaks a larger data set into smaller groups of data (cohorts) before analysis, usually based on commonalities or similarities in experience or situation.

Cohort analysis is particularly important for companies looking to better understand and serve different segments of its user or customer base. A social media company might use cohort analysis to break up the data gathered from its users based on which services they use to optimize improvements to user experience.

Cluster analysis

Cluster analysis is a kind of data analysis that involves identifying similarities and differences in an unlabeled data set and sorting this data set accordingly.

Clustering is often used for tasks like market segmentation, online recommendation, search results, and medical imaging analysis. Online content streamers like Netflix and Youtube use sophisticated cluster analysis to identify viewers with similar viewing habits and recommend new content to them.

Time-series analysis

Time-series analysis involves analyzing data points collected over a period of time at regular intervals. 

As would be expected, time-series analysis is particularly useful for understanding change over time, as well as for predicting future events. In finance, time-series analysis can be used to draw meaningful insight from daily closing values of stocks or other securities.

In addition to these techniques, data analysts and other data analytics professionals are required to have a host of other skills. We’ll dive into these next.

To work in data analytics, you must possess a diverse skill set. Though exact requirements will vary depending on the industry, seniority, and responsibilities, key skills generally include:

• Programming: The ability to use Structured Query Language (SQL) to store, manipulate, and retrieve data from databases is essential for data analytics professionals. Increasingly, data analytics professionals are also proficient in R, a statistics-specific programming language, and/or Python, a general-use programming language.

• Data analysis techniques: As detailed above, there are a host of statistical analysis techniques that data analysts must master, including regression analysis, factor analysis, cohort analysis, cluster analysis, and time-series analysis.

• Data analytics process: A data analytics professional must be skilled in data collection, data cleaning, and data mining.

• Data visualization: Companies often expect data analytics candidates to be able to create visually appealing data visualizations using tools such as Tableau.

• Soft skills: In addition to technical skills, data analytics professionals should also possess strong teamwork, leadership, critical thinking, and communication skills.

• (Optional) Machine learning skills: For advanced analysis, some data analysts and other data analytics professionals will employ machine learning methods or utilize machine learning libraries like Pandas or Keras.

We’ve now covered what data analytics is and what you need to be successful in the field — but what kinds of opportunities are out there? In the next section, we’ll look at some real-world data analytics jobs.