Introduction: Why Data Analytics is Important

"In God we trust, all others bring data."

W. Edwards Deming (Ratcliffe, 2018)

Everyone remembers the troublesome process of raising a request with IT or the vendor for a small data analysis task and waiting days or even weeks before getting the result. More often than not, the result was not presented in the most useful way or just raised a follow-up question which required new data to answer, new requests for IT or the vendor. For decades, managers have relied on this kind of process because they had no choice. This process has a fundamental flaw: if you want to make timely decisions, lagging and outdated information cannot be used. Hence, timely decisions had to be done without having the foundation of real-time data and sometimes based on gut feeling.

Times have changed. The quantity of available data in all functions of any organisation is growing daily. The access to this data gets easier and easier. Nearly everyone can acquire the data necessary to run their own analyses. And almost everyone has a formidable computer with powerful analytics tools immediately available. The question now is, how to turn the data into business-relevant information for making the right decisions when needed (Data Never Sleeps, 2020).

Hence, it is time to ensure that the right data are collected in an appropriate way, screened, transformed, and analysed using valid methods in a manner that delivers business-relevant information that is turned into intelligence for making appropriate decisions for business success.

Data analytics is the process of collecting, processing, and analysing data with the objective of discovering useful information, suggesting conclusions, and supporting problem solving as well as decision making (Wikipedia, n.d.; Payne, 1976).

"Data Analytics is a business practice every Manager should be familiar with".

Data analytics encompasses the main components: descriptive analytics (post-mortem analysis), predictive analytics, and prescriptive analytics.

Big Data is characterised by its three Vs, Volume, Velocity, Variety (Russom, 2011). It is too voluminous and complex for conventional hard- and software to handle. At the beginning of the 2000s, the volume of available data went up exponentially and handling such data was reserved to a few companies and organisations who were relying on the analysis of data to stay in business.

Nowadays however, computers with huge data storage and handling capacity are available to nearly any organisation, be it by installing hardware and software inhouse or by renting external capacity. Two trends seem to be the result of this change in the IT environment. Firstly, more and more organisations have the means and see the need to collect data about their operational environment. Secondly, these organisations are widening the scope of their data analytics activities to include all functions.

Not only is there a move from Big Data analytics to analytics of any kind of data, there is also a healthy trend towards involving all levels of management and even junior staff into this, not so new, field of information management. Progressive managers are familiar with the types of data available and with the trends, shifts, or other patterns in their data and can use them for decision making.

The former speciality data analytics is gaining popularity amongst all managers of an organisation. Hence, it is time to ensure that the right data are collected in an appropriate way, screened, transformed, and analysed using valid methods in a manner that delivers business-relevant information that is turned into intelligence that prepares appropriate decisions for business success.

"The ability to take data - to be able to understand it, to process it, to extract value from it, to visualize it, to communicate it - that is going to be a hugely important skill in the next decades."

Google's Chief Economist Dr Hal R. Varian (2009)

Why This Book Has Been Written

To be very honest, I did not like statistics very much. After I covered statistics during my engineering study, I made sure I passed all the tests and exams and phew . never again.

After joining General Electric (GE) in the 1990s, I had to relearn statistics. At GE, acquiring knowledge in statistics was purpose driven, i.e., it was real business problems at hand I had to solve using numbers. Consequentially, I started to gain some degree of interest for maths and stats.

As the following examples testify, having the numbers is good but not enough. In addition, we need to ensure the data are properly collected, cleaned, and analysed before making any decision.

Some years ago, the director of a blood bank came back from a meeting with other blood bank heads. She was not happy because she had the chance to compare certain blood bank performance indicators with others and recognised that her own blood bank was obviously wasting more blood products than some other blood banks. She was talking about bags with platelets that were taken from blood donors, tested, and then disposed of because they did not meet the quality standards. By her criteria, this kind of situation was not acceptable.

A team was set up to investigate the root causes for the wastage of the most precious blood products. After data collection and some basic analysis, it became clear that the blood products were not of lower quality than in other countries. The root cause was in the process of evaluating the quality of blood bags - the data collection.

This case example is explained later in this book.

"Having wrong data is worse than having no data at all".

This case alone generated some life-long learning that we wish to turn into recommendations:

Firstly, do not trust numbers blindly. Even numbers that are produced by a computer can be wrong, biased, or otherwise made useless. Check how these numbers were generated and how they got into the computer in the first place.

Secondly, before you perform any data analytics, ensure that the data are collected following a proper procedure. Therefore, this book does not start with data analysis. It starts where the collection of the data is thought about and designed.

Thirdly, like the head of the blood bank with her very powerful business case, confirm that your data analytics serves a purpose, a need the people with whom you work know, understand, and share. Only with this need, business case, can your data analytics be more than playing with numbers.

The following chapters elaborate the use of data analytics to solve business problems, to make critical decisions, and to drive organisational strategy. We will identify some typical pitfalls and remedies in the process.

At the moment, there are numerous Data Analytics courses available. Interestingly, many of them have similar titles to Data Analytics for HR Professionals or for Customer Relationship Management. This book takes a wider scope and shows the application of data analytics in any organisational situation, where the proper use of data is critical. Therefore, we call it "Data Analytics for Organisational Development: Unleashing the Potential of Your Data".

This book is written with the intention to close a well-known gap mentioned by Amy Gallo (Gallo, 2018). Every manager should know four powerful analytics concepts in order to be informed about his organisation and to make data-based decisions. These concepts are in no way new. However, they gain more importance with the increased amount of data available and the apparent need - and the chance - to turn this data into business-relevant information. This is supported by the availability of a multitude of easy-to-handle tools for data analysis and data visualisation.

These tools can only be used by managers if these managers understand the basics of data analytics from data acquisition to data analysis. Therefore, as identified by Gallo (2018), managers need to know the basic concepts.

These concepts are randomised controlled experiments, hypothesis testing, regression analysis, and statistical significance.

Randomised controlled experiments include data collection techniques such as any kind of surveys, pilot studies, field experiments, and lab research. Instead of outsourcing such services to specialists and relying on them analysing the results and developing recommendations, it would be beneficial for managers to understand data analytics. This knowledge would certainly help draw customised conclusions for the organisation; conclusions an outsider cannot easily draw. Experiments also comprise testing new routines or products on their performance. Experimenting with processes is a powerful way of improving the output whilst observing and changing settings at the same time in a controlled way.

Hypothesis testing contains statistical tools that compare stratified business-relevant data and answer the question for the "better one" including the inherent risk of this decision being wrong. Hypothesis tests find their application in all units of any organisation. Analysing survey results uses hypothesis tests to answer questions like "Is there a difference between last year's and this year's rating?" or "Did Department A perform better than Department B?" The result of a hypothesis test can be much more than just...