“Big data has all the answers!” Fair enough, but what should I ask?

Date:  29 November 2023

My last post ended in the words: ”find the questions relevant to your organization and only then can you start utilizing big data and analytics successfully’.

One should really pay attention to figuring out the questions that could bring useful results. In “The Hitchhiker’s Guide to the Galaxy,” the poorly-formulated ultimate question of life, the universe, and everything received an answer of “42” despite the use of a supercomputer surely capable of handling big data in all its forms.

Traditionally, there are four types of analytics, all of which answer different kinds of questions: descriptive, diagnostic, predictive, and prescriptive analytics. Before analytics can be done in the first place, data needs to be collected. Moreover, in order for analytics to have an effect on an organization, its results should be put into action. Next, examples of relevant questions to be asked of data are described under the different types of analytics.

Data collection (before analytics)

It is important to understand the data an organization has in the first place; i.e. what you are measuring and what you should measure. Equally important is to understand that having the data is just the beginning. Data collected can be either a valuable gold mine or a showcase of wasted resources. If you do not start utilizing the data you have collected, the latter will be true.

Analytics for past (descriptive analytics)

Describing past events, descriptive analytics, is the most common type of analytics and many organizations perceive analytics to be just that. Once you have the data from operations from for example last month, you want to know what has happened. With such information you understand better how your organization has performed. You might also spot pressure points and high performing areas through this data.

Analytics for diagnosis (diagnostic analytics)

The second type of analytics aims to increase understanding of why things have happened the way they have. Typically, in hospital operations for example, a myriad of factors may affect the results. In such an environment, finding out the true causes of any event can be very cumbersome. Consequently, various rules of thumb are used to try to fix acute problems, meaning that the real causes are never discovered. This puts organizations in fire-fighting mode – but these fires could be put out for good if one could first understand the causes and fix them instead of the symptoms.

Analytics for future (predictive analytics)

The third type of analytics, predictive analytics, is the key to enabling the transition of operations from reactive to proactive actions. It is so much easier to prepare for the future when you know what you will be facing. The question that many want to answer is “what will happen?” Not everything can be predicted accurately, but in a repetitive environment, such as hospital operations, many events reoccur and are relatively predictable to a certain extent.

Analytics for action (prescriptive analytics)

In the fourth type of analytics, it is possible to support decision-making by giving suggestions for future actions through data, i.e. what you should do. In certain areas, it is also possible to automate decision-making, although one has to be very careful in attempting to do that, especially in a hospital environment. Through utilizing prescriptive analytics, an organization has come a long way from understanding what has happened in the past to understanding what to do in the future to become even better.

Operational excellence (after analytics)

As previously stated, analytics itself is not enough – you also need knowledge and most importantly, action. The biggest value for an organization comes from situations in which  you understand not only the big picture of operations, but also what is going to happen, and then you act accordingly in a proactive manner. Therefore, in order to achieve high-quality operations, organizations (or their management at least) have to ask themselves one more question: based on the knowledge acquired through analytics, what should our best practices be? After all, analytics itself does not bring value for an organization – it is the knowledge acquired through analytics and put into use.

Emil Ackerman, member of the Harnessing Big Data SIG Steering Committee and Chief Data Officer for Pirkanmaa Wellbeing Services County, Finland.

Guest writer

A blog from a member of the Harnessing Big Data SIG.


Wellbeing services county of Pirkanmaa, Finland

Emil Ackerman

Chief Data Officer

I lead the development and implementation of knowledge management and data utilization strategy for the biggest wellbeing services county in Finland, with approximately 20,000 employees.

News & insights

Look into our latest...

Statement on the WHA77 Resolution on Climate Change and Health

The IHF and its members share a commitment to address sustainability challenges working with, and for, our communities. In November...

Leading digital transformation in healthcare

On 23 May 2024, the Association Leaders Circle hosted the virtual meeting “Unlocking the future: AI technology and digital transformation...

Telehealth outside the hospital setting

On 16 May 2024, the Telehealth Special Interest Group, led by SIG sponsor Apollo Hospitals Group, hosted its latest webinar...

Graphic from Pia article

Do you manage your organization with reliable data?

In healthcare, numerous decisions are made every day, with significant effects on patient care. Decisions guide how to use human...