Data analytics is fast rising in popularity and few are aware that there are different types of analytics with different uses. As outlined in a previous post, your business goals decide the data analysis tools you use. This also applies to the type of analytical technique you use. Let’s look at the four types of data analytics and how they are used. The graph below shows their potential value to a business vs complexity of use
1. Descriptive Data Analytics
Descriptive data analytics is by far the most common type of data analytical techniques today, accounting for about 80 percent of business data analytics. Viewed as the introductory aspect of data analytics, it’s designed to gain basic information: who, what, when, where, how many?
It simply pinpoints a problem which is usually visualised in simple charts, dashboards, and scorecards. Other data analytical techniques are then required to pull further insights from data.
For example, this technique may be applied to find out why total online sales for a certain product in 2017 dropped. It can show that there were zero sales in a particular month, picked up in another, plateaued subsequently.
It won’t show, for example, why there were no sales in that specific month. Sso other analytical techniques are used to address this limitation.
2. Diagnostic Data Analytics
As the name implies, this technique matches historical data with other data to answer ‘why did this happen?’ questions. It usually applies techniques like drill-down, data discovery, data mining and correlations and often results in a dashboard.
It can be used, for example, in a social media marketing campaign, where it can be used to assess page views, number of posts, fans and other social media performance metrics. There can be thousands of online mentions that can be distilled into a single view to see what worked in your past campaigns and what didn’t.
One common drawback with using this tool is that patterns can be misinterpreted as the cause of certain problems in an organisation. While answering what and why questions via analytics works for some businesses, others may gain more valuable insights from examining future trends.
3. Predictive Data Analytics
This technique analyses data to make predictions about uncertain future events. This is possible by using predictive models which are created by applying statistical analysis techniques to automated machine learning algorithms.These models assign scores or numbers based on the likelihood of an event occurring. Being able to make informed predictions helps organisations to be more proactive and prepared for outcomes in future. Practical applications in the industry include fraud detection, optimising marketing campaigns, reducing risk and so on.
For example, by examining patterns found in historical and transactional data a business can identify risks and prioritise them based on their impact. Another popular use of predictive data analysis is analysing unstructured text from social media or call centre notes to gauge user sentiments.
4. Prescriptive Data Analytics
By using AI and big data techniques, this provides predictions and recommends actions based on the predictions. It helps answer questions like ‘What happens if we do so and so?’ or ‘What is the best course action in this situation?’
Prescriptive analytics employs a combo of tools and techniques (algorithms, machine learning and computational modelling procedures) to process hybrid data (structured and unstructured data) and business rules.
While its impact is invaluable, prescriptive data analysis is difficult to administer and many companies can’t use them yet for daily business transactions.
Which method works best for your business?
For many businesses, as was pointed out, descriptive data analytics would suffice. Regardless before choosing a technique for your data analytics, it is important to ask questions such as: What are my objectives? How deep do I want to go into this data? Am I skilled enough or have access to skills that can do this? Do I need even need this?
Once decided, it is important to follow the rest of the steps prescribed here.
