Is your leadership embedded with the right future visions for your organization to anticipate risks and opportunities? Most of the time probably not, despite most leaders’ faith in the opposite. Plausible future visions emanate from a perfect mix of correct data collection/analysis and the right intuition. Making anticipatory decisions based on future visions is risky, but can be extremely rewarding.
Forecasting is in itself a very peculiar exercise fraught with dangers, uncertainties and unforeseeable occurrences. But forecasting is exciting and may quickly place you ahead of the competition if it plays out well. But are you sure your data scientists provide you with the best possible information for forecasting? Furthermore, are you engaged in forecasting all by yourself as an autocratic leader, based solely on your intuition, or are you relying to a greater or lesser degree on your team members to collaborate with you?
Which most valuable trait do the most famous CEO’s such as Steve Jobs, Elon Musk and Jack Welsh have in common? They are visionary leaders. Visionary leaders have the ability to craft a mental picture of a state that does not yet exist (e.g. MLK). They confidently build an attractive image of the future and they convince their peers that they are able to take them there.
Ranadivé & Maney (2011) say, “The secret of being a tech CEO is having an efficient, agile mental model that can quickly predict what’s going to happen and be right most of the time.” Ben Horowitz, cofounder of Andreessen Horowitz and one of SiliconValley's most respected and experienced entrepreneurs says, “There are two types of C-Suite people: the ones and the twos.” In his view, the “ones” are those with the ability to predict. The “twos” are the others. He concludes that only the “ones” make great CEO’s.
Are your data scientists providing you with the best possible available information? Big Data mining and analysis are complex tasks fraught with risks. Jules Berman (“Principles of Big Data; Preparing, Sharing and Analysing Complex Information”, 2013) reminded us that “Big Data is a new arrival to the information world; with rare exceptions, database managers are not trained to deal with the layers of complexity found in Big Data resources”. This was two years ago and despite the amazing evolution of this field, Big Data still is in its infancy.
There are still many rough edges to smooth out in Big Data collection and crunching: data blind spots, bad resource inventories, correlation faults, modelling limitations, data systemic biases and hazy side effects of data scrubbing (especially when deleting incomplete or inconsistent records or running various algorithms to reduce data sets). All these add up to the arduous and complex task of extracting the correct information from Big Data.
Besides, current software/platforms for Big Data such as Hadoop, Python, Java, R and others still need to be further developed, notably to improve integration between systems. Most data scientists cope on a daily basis with the complexities and limitations of data crunching software to digest the exploding influx of data.
And who are these data scientists? Are you sure you can rightly call them all data scientists? It is interesting to note that a recent report published by RJMetrics on data science accounts for only around 11,400 data scientists worldwide. Jules Berman’s view that database managers are not trained for Big Data is still resounding. More training is needed in data science.
A data scientist is to some extent an evolution of the database manager. Data scientists need to be able to program, if possible in different programming languages. They should have an understanding of Hadoop, Hive and other platforms. They should be familiar with natural language processing (the interactions between computers and humans), machine learning (to improve and develop algorithms), statistical analysis (notably to figure out how to sort possible limitations in models) and conceptual modelling (to conceive models).
They also need to engage in predictive modelling (most Big Data goals relate to being able to predict future outcomes) and hypothesis testing. Altogether, it is a highly qualified job in short supply in the labour market at this juncture.
This has to do with your type of leadership. Are you an autocratic leader that seldom accepts inputs from your staff? If so, you basically make fast-paced decisions which adds up to the agility of your management. But if you’re a consultative or group-based leader, then you collect and digest the inputs of your team members and this usually slows down the decision making process.
But nowadays things have changed in this regard. A substantial number of Big Data and predictive analytics’ mobile applications have been launched in the market. They provide full mobile access to data to everyone in the loop. Consequently you can exchange insights with peers, advisors and customers to provide valuable context for collaborative decisions. This means that your collaborative decision making process is permanently online and can be accomplished extremely quickly.
This is a critical factor. Visualization tools such as digital dashboards, scorecards and charts are evolving quickly. The quality and the innovation factor of graphical dynamic imagery of data are of the utmost importance to provide you with the broadest picture of the environment and the right references and vectors to boost your imagination and intuition to look into the future. This is a field that is being currently cherished by many software households because the right visualization of what is going on and what may happen are paramount to the decision making process.
By accessing and analyzing data, your team members may also identify trends that may disclose new markets, channels and innovations. Predictive analytics empower people to look ahead. Most likely, some of your team members are also highly intuitive and good at building future visions. So you’re probably not the only highly intuitive individual in the organization. You should harness the potential of other intuitions, even if at the end your intuition should ultimately prevail.
Prescriptive analytics, as the final frontier of business analytics, go beyond forecasting by also recommending actions in order to gain from predictions. Prescriptive analytics may also give you insights into the implications of each possible decision. The main aim is to take advantage of future opportunities or avoid/mitigate future risks.
Prescriptive analytics software combines predictive models, simulations, business rules, scoring and deployment options of data mining and optimization techniques to produce computational recommendations for decision making processes. For instance, you can fully automate complex decisions and trade-offs to improve the management of limited resources.
You can also scale up threats and fraud prevention and proactively update recommendations on many other subjects based on changing events. Ultimately, prescriptive analytics software will show you a wide range of options to optimize your business processes to meet your operational goals. But in the end it is your intuition that selects the right recommendations and the course of action.