Prior to predictive analytics becoming the rage, CEOs more often than not relied on their own interpretation of historical data as the basis for making business decisions
In today’s day and age, one doesn't need to be a data scientist to understand the role data is playing in transforming businesses across industries. It has certainly become very critical when it comes to business decision-making. Never an easy ask to begin with, decision-making has become even more challenging for business executives, who often find themselves swamped with humongous amounts of data, all of which may seem relevant, but most often, only clouds their thinking and leads them astray.
Prior to predictive analytics becoming the rage, CEOs more often than not relied on their own interpretation of historical data as the basis for making business decisions. When predictive analytics tools first appeared on the scene, they helped overcome the limitations of the human mind by analysing vast amounts of data. The inferences they drew were often met with a good deal of skepticism, especially when they went against the intuitive conclusions arrived at by the executives responsible for making the decision.
This scenario hasn’t changed much in the years since. Even today, almost as many as 81 percent of CEOs in India are less confident about the accuracy of predictive analytics than historical data, according to KPMG’s India CEO Outlook 2018. Even more interesting to note is that as many as 62 percent of CEOs in the country have, over the past three years, overlooked insights from data analysis models because they were contrary to their own experience or intuition.
That said, there are times when predictive analysis can misfire and go wrong. Human behaviour is so unpredictable that even the best of algorithms, computational models and analytics tools can sometimes fail to take all the unknown and unexpected variables into consideration, thus resulting in incorrect predictions. The outcome of the 2016 US Presidential Elections, for instance, proved to be a complete antithesis of what many from the data science community had predicted. Politics, it became apparent, is one of those areas where predictive analysis will, more often than not, be ineffective. The human resource function in companies is another, for similar reasons.
However, unlike politics or HR, when it comes to business or sports, historical data can more often than not help in spotting trends and patterns that could influence the future course of actions. This is why predictive analytics is finding application and acceptance in an increasing number of industries and business functions and serving a wide range of purposes, which includes helping with churn prevention, estimating customer lifetime value, defining market segments, streamlining maintenance costs, identifying the best marketing channels, ensuring quality control, identifying potential areas of risk, performing sentiment analysis, suggesting product upselling and cross-selling, and more.
What's more, predictive analytics is becoming increasingly affordable for all. Until a few years ago, only large corporations with large marketing budgets could afford it. Now, however, almost any organisation can avail of cloud-based predictive analytics for a monthly subscription fee.
So, does all of this mean that intuition doesn't, or shouldn’t, have a place in business decision-making anymore? Not at all. Relevant data may not always be available on every desired subject. Or sometimes, the issue in question may involve moral or ethical considerations. No machine, however intelligent, and no amount of data can substitute for human experience and judgement in such cases. A CEO, for example, may be faced with the decision of whether or not to sign on a new client for his company’s business. It may be that the deal is a great financial fit, but the client has a reputation for being nastily difficult to manage. It is one of the many possible situations, where intuition comes into play.
Though intuitive decisions tend to be successful only when the decision-maker is an expert on the subject in question, intuition is of little or no use if one delves into an area outside their expertise. Moreover, calling upon one’s intuition to make a good decision requires more than just domain knowledge; it calls for the ability to introspect and jettison any and all biases that may influence the decision-making process. It is important to note this because intuition can go wrong for many reasons, including one-off bad experiences, emotional biases, prejudices or vested interests.
Ultimately, what we require to do is balance both intuitive and data-driven decision-making and also take care to correctly identify the situations to which each must be applied. For instance use their intuition to narrow down a wide choice of options to a manageable few, and then analyse them in a logical, data-driven manner. Contrarily, performing a detailed analysis may yield some options that all seem equally good, and then one may use their judgement to decide on the best one.
What can be inferred from this is Indian CEOs, today still favour their experience and intuition over insights from data analytics. But they do acknowledge that decision-making is likely to be increasingly aided by data science in the foreseeable future. This can be attributed to the fact that 41 percent of them are, in fact, planning to increase the use of data analytics in decision-making over the next three years. It’s all for the good. In time, it will become evident that the decision-making ability that CEOs pride themselves on isn’t confounded, but complemented, by machines. It’s simply a case of the yin and the yang in a business setting, and it is waiting to be recognised as such.
(Shalini Pillay is Office Managing Partner, Bangalore, KPMG in India. Views expressed are personal.)