Harini is a strategic HR Leader with proven expertise in organisational restructuring for M&A and Change Management. A design thinking practitioner, behavioral analyst, transactional analyst, and innovator, she is a strong advocate of the concept of Happy Workplaces. Harini has been leading the HR function in MNCs and is a guest faculty at prominent B schools and universities.
We, as humans and professionals, have been fascinated with data and the resulting patterns for centuries. Fast forward to today and data has taken a centre stage at the workplace and changed the lives of CXOs. While data analysis driven by Artificial Intelligence and Machine Learning is often seen as the best way forward, it is a good idea to validate the pros and cons of relying on this data. Harini Sreenivasan takes us on this data-bound journey to help us figure out where we stand in the digitally competitive world.
Mankind’s fascination for observing patterns of data and converting it into information suitable for consumption dates back to centuries ago. Psychologists would agree that this must certainly be a form of evolution by learning. Today we are talking about data and analytics as key decision-making tools. Let’s rewind and go back in time to see where it all began.
In the 18th to 20th century, the Industrial Revolution brought about a seismic change to the global economic landscape. Methods were put in place to do a scientific study of tasks carried out. Scientific management principles were put to use, collect, and study data. During the World Wars and subsequently, hiring decisions began to depend on available data and its trends. And now in the 21st century, the emergence of technology and its integral role in managing data became a game-changer. Data can be shared, analyzed and even retrieved easily. organisations have started to look at analytics as a critical tool for business growth since trends and patterns gave a lot of insights into what was beyond the obvious.
How Data Analytics Has Changed the Lives of CXOs
Being a CXO and that too in such rapidly changing times is a tough job. Ask me, for I am someone who has been through it! Irrespective of their core function, they are expected to understand the data trends in the business, engage with the next level of leaders, maintain healthy relationships with all stakeholders and also take firm and fast decisions that accelerate business growth. A 2019 Gartner survey revealed that by 2020, more than 50% of organisations would be totally driven by data analytics for their decision making. If the CXOs decide to take this lightly or go slow with the transformation, competition would outdo them in no time.
As one would expect, the competitiveness has made analytics take the centre stage to become a predictive tool from being a prescriptive supporting one.
…more than 50% of organisations would be totally driven by data analytics for their decision making. If the CXOs decide to take this lightly or go slow with the transformation, competition would outdo them in no time.
Are We Ready for Predictive Analytics – The New Age Jargon?
Wouldn’t it be good to be able to predict the future? If only we could know things in advance, especially the ones that are likely to impact our business. Analytics driven by Artificial Intelligence and Machine Learning is often seen as the best way forward. However, wouldn’t it be even better to first be aware of certain pitfalls before taking a plunge into data and its analytics?
Predictive analytics have played a pivotal role in areas such as business forecasting, especially in market research. Here, trends could be predicted based on research findings of the market, context etc. Such models typically forecast something that is beyond our current visibility or development that is likely to emerge and we need to quickly adapt our decisions to suit it.
On the other hand, when it comes to predictive models related to people and their data, there is a different perspective. We do not want to limit ourselves with the findings, we jump to look at the predictions so that we can intervene and bring in more of the human aspect. This is where Machine Learning comes to our rescue. It could serve a wide range of purposes, like recommending the most suitable courses, suggestions for making teams successful, and a lot, lot more.
Here comes the twist! Interfering with what you are predicting may complicate things. Actions that follow the predictions require some consideration of the current context and the future journey. In my experience, chances of success with predictive analytics are greatly increased when you keep the following four guidelines in mind:
1. Get an alignment on the definition of ‘predictive’ in your context
2. Perfection is a waste of time
3. Build on ‘what can be influenced’ and not on ‘what else can be done’.
4. Keep it simple and straight, connect with the context.
The trick is to get on the same page with stakeholders and agree on what needs to be predicted” in order to solve the business case at hand. The best way to get there is to start with the problem statement and work from there to see how a predictive model may support it.
Get an Alignment On the Definition of ‘Predictive’ in the Context of your Business
Being a new age jargon, predictive analytics tends to be sought as the best solution to any business problem. It is thereby very important to ask ourselves this question: does everyone involved mean the same thing when they say “predictive model”?
Theoretically speaking, predictive analytics may not be for short term results. For that matter, it may not be time-bound at all. For example, predictive analytics may be used for organisational redesign and development which could often be regardless of time since it is continuously evolving. The trick is to get on the same page with stakeholders and agree on what needs to be “predicted” in order to solve the business case at hand. The best way to get there is to start with the problem statement and work from there to see how a predictive model may support it.
Do not get hung up on the term ‘predictive’ and work with your stakeholders to decide what insight is needed, and how a model may contribute.
Perfection is a Waste of Time
There is no such thing as ‘perfection’ in prediction. More often than not, results of predictive analysis are in numbers. But if you take a closer look, they are outcome-focused and are beyond just the numerical values. For example, the predicted turnover rate could be a percentage but the business looks at it as a vital statistic that talks about the health of the organisation. More than the accuracy of ‘how many people are leaving’, it is more important to understand ‘why many people are leaving’. In fact, many times the more powerful and accurate the data, the harder it is for us humans to grasp how it transforms data to a particular output or prediction.
Build On ‘What Can Be Influenced’ and Not On ‘What All Can Be Done’
Even when everyone agreed on what should be predicted and your model is transparent, you may find that there is never a clear way to use insights gained from that model.
Any analytics that is complex to grasp and interpret has a very short shelf life. Providing useful results tends to lead to more specific questions. And these can lead to the development of more specialized solutions for the business.
To counter this, take into account what can be influenced or controlled. Absenteeism models may rely heavily on age or gender to provide an outcome, neither of which an organisation can do much about. Take care not to develop a model using mostly or only such features, which is a risk when using HR data exclusively.
Instead, consider adding data on working night shifts, short or long shifts, alone or in teams, and so on. These are factors that can be changed and are therefore more likely to yield actionable results. Gather information on relevant business processes, identify related data sources and use those, if you can. This way, your model is more likely to provide insight into how HR themes and business processes are related and what could be done to improve both.
Keep it Simple and Straight (KISS), Connect With the Context
And finally, it is extremely important to seal it with a K.I.S.S. Any analytics that is complex to grasp and interpret has a very short shelf life. It does take a lot of effort behind the scenes to make the collected data robust, develop tools to analyze the trends and provide slices and dices of useful and insightful information. Providing useful results tends to lead to more specific questions. And these can lead to the development of more specialized solutions for the business.
In Conclusion
When talking about machine learning and predictive analytics, we commonly hear that there needs to be a “digital solution” at the end. Whether digital or not, predictive analytics development is largely about research and that means that you cannot know in advance what you will find. Does that mean that these projects tend to fail? Not at all, provided the outcome leads to useful business insights.
