Giving you better decision support is the ultimate goal of analytics solutions. This will allow humans to make use of relevant information to make better decisions. Capabilities with regards to decision support can be categorized 5 ways. Each of these categories when deployed answers a different type of question.
What is the plan?
– Planning Analytics
What happened in the past?
– Descriptive Analytics
What was the reason for the occurrence?
– Diagnostic Analytics
What is going to happen in the future?
– Predictive Analytics
What should we do about it?
– Prescriptive Analytics
As seen in the above points, it is like a process. The first point, planning analytics – focusses on what the plan needs to be. Once the plan is identified, you need to figure out what is happening in the business. This is where descriptive analytics comes into the fray. Descriptive analytics allows you to answer the questions pertaining to ‘What happened?’ in all its forms – What were the sales for last month, quarter or even yesterday? Which products had the most number of defects in the last month? What type of customers needed maximum help from customer service? Answering these questions lay a solid foundation for a sound analytics strategy. These initial questions are very important in order for you to set goals and the respective KPIs. This will determine how the enterprise will be managed and measured.
The most common association that we see is between descriptive analytics and data visualization. This association is viewed via dashboards, reports and so on. The rapid adoption of analytics technology can be achieved with captivating visualizations and a user interface that is intuitive so that it can adapt to different types of decision makers. Visualization on its own, however, is only one of the many functionalities of descriptive analytics.
1. Determining business metrics:
The identification of the metrics that are key to evaluate against business goals is a must. Examples of goals that can be set are – to overhead costs, measure productivity, improve revenue via sales, improve operational efficiency so on and so forth. The respective KPIs must be assigned to each goal so that the achievement can be measured and monitored.
2. Data requirement identification:
Data in any organization is stored in multiple sources. This may include – databases, shadow repositories, desktops and records. In order to accurately measure the goal against the KPIs, the required data must be extracted from the correct data source and the organization must maintain a catalogue of this. The metrics must be calculated based on how the business is doing in the present.
3. Data extraction and preparation:
Prior to data analysis, the data needs to be prepared. Cleansing, de-duplication and transformation are just a few examples of the steps involved in data preparation before analysis. This is the most labour intensive and time-consuming step in the entire process. It requires nearly 75% of an analyst’s time. However, it is very critical in order to ensure maximum accuracy.
4. Data analysis:
Models can be created and analysis such as summary statistics, regression analysis and clustering can be done on data in order to measure performance and determine patterns. With the aim of evaluating performance based on historical results, important metrics are calculated and compared to the business goals that have been set. Open source tools like R and Python are used by data scientists to analyze and visualize the data.
Also Read: 5 Ways how Predictive Analytics can help you to get most out of your Business
5. Data presentation:
For the stakeholders to see the results of the analytics, it is presented in the form of graphs and charts. This is exactly where data visualization enters the picture. BI tools such as Power BI, Tableau, etc., give users the ability to present data in a way where people who aren’t analysts can understand easily. There are self-service data viz tools that allow users to create their own visualizations and even alter the output.
It cannot be stressed enough that KPI governance is extremely important to the success of modern descriptive analytics. Today’s business environment is in a constant state of flux. With that in mind organizations should establish and assess a set of changing KPIs. In a study conducted by IDC which involved 150 chief analytics officers, it was found that over a period of 18 months, nearly 70% of them had started measuring and tracking new KPIs for their organizations. This behaviour which is very evident is the most important sign of digital transformation as the readiness to ask new questions and challenge the status quo is showcased.
The Role of Descriptive Analytics in Future Data Analysis
As the use of the results of descriptive analytics by data-drive businesses to enhance their decision-making will continue, data analytics by itself has moved from descriptive, to predictive to prescriptive analytics. It has moved to be a mash-up of simulations, predictions and optimization.
Data analytics’ future does not lie in simply describing what happened, but to accurately predicting what will happen in the future. This statement or claim can be backed up with an example of a GPS navigation system. In this example, descriptive analytics is used to provide directional cues. When it is further reinforced by predictive analytics, the system offers important details like ETA, distance etc. Go one more step and add prescriptive analytics to the mix, you will be able to see the shortest route to your destination based on comparison of multiple routes. Sound a lot like Google Maps to you?
Descriptive analytics will always be the cornerstone based on which analytics is done. It always is the first and most important step in the journey of analytics. Even though it has evolved into predictive analytics and further into prescriptive analytics, the need for descriptive analytics will always remain.
Abhimanyu is a sportsman, an avid reader with a massive interest in sports. He is passionate about digital marketing and loves discussions about Big Data.