Most discussions on analytics and its evolution tend to get diluted by someone’s “differing perspective” on what analytics really is? While the term Big Data is becoming ubiquitous, the definition of how big data have to be to be called Big Data gets into a debate all the time. So, to start the discussion, let me briefly cover the four different views that I have encountered on a day to day basis.
The view of analytics as an industry – If, of an equilateral triangle, you could map the interaction between technology, analysis and corporate strategy, any point inside the triangle is a different version of what’s currently being discussed as analytics. Now, if you add another dimension to it, which is called “visualization/presentation” it makes the view almost complete.
For instance –
- So, high analysis and strategy is the current focus of advanced analytics – simulations, neural networks, forecasting, etc.
- High on analysis and low on strategy and low on technology is ad-hoc legacy organization type reporting.
- Low-medium analysis and High technology is the product based dashboards such as BO/ Microstrategy, etc.
- High analysis and high strategy is the analytic based differentiator where analytics sits in the corporate strategy organization and supports CEO’s office.
There are very few examples of organizations that are maximizing across these four dimensions simultaneously even today, despite the discussion around big data. In my experience, every time I reviewed a business development opportunity, I would fit the opportunity in this quad and accordingly think about the solution that could be offered. There are engagements where a client is only looking for visualization that aids executive thinking at one extreme. And there are engagements where the client does not care about visualization or strategy as much because the focus is on tactical decision making.
Now, the Big Data discussion has thrown the buzzword in every corporate boardroom and people want to know what’s being done in their organization with respect to BDA. However, the sensibilities around extracting the value from Big Data are still heavily driven by which of the axis is driving the adoption. Somewhere, what the entire analytics industry is hoping and preparing for is the exact center of gravity of this quad technology, analysis, strategy and visualization come together in a balanced way. Technology (scalable, agile and business integrated) is the key to adapting to the evolving marketplace. Analysis (speed, accuracy, robustness and presentation) holds the key to managing for growth. Strategy (problem solving, business value and resource optimization) will hold the key to dancing with the times. Visualization makes all of this consumable.
From an academic point of view, analytics is a combination of four core subjects – mathematics, statistics (and while mathematics is a precursor, I’d like to keep statistics out as a separate subject), economics and strategy. And in a separate post, I will probably share some of my thoughts around this. Like all business, there is a lot more that you need to be an ace analyst, by the way.
The third view is that of an analytics organization. I have had a chance to work through the inception and expansion phases of third party analytics teams, but as part of a large consulting firm, I have also had a chance to see several analytics organizations embedded within Fortune 500 firms. The differences in quality, quantity, application and perception are stark across these organizations. Take for instance, a Capital One, which is renowned as an Analytical Competitor. So is P&G. Take HUL on the other hand. Or Citibank. Heavy users of analytics as a support function which drives several important decisions, but the firm does not use analytics as a primary skill. Take a Coca Cola, Pepsi, Kraft Foods, Nestle and so on – the variety of analytics and the application is also significantly different across these organizations. Take all the payments companies such as Visa, MasterCard, PayPal, etc. and despite heavy and decentralized analytics team that support all aspects of business, analytics comes in as a secondary in more cases and primary in a few. Take manufacturing companies which might have adopted several inventory management principles and techniques eons back, but when you delve deeper, you realize that there is something amiss in the quality and application of analytics. And finally, when you take third party organizations, and reflect the first industry view on top of the organization, while the business of analytics is at the core of these organizations, their own perceptions of where analytics fits in their clients’ organizations is a major factor in the way they hire, train, and develop their people.
Which brings me to the fourth view – the people view. The industry is full of quants (math, econ, statistics, etc.), architects/technology specialists (data architects, modelers, DBAs, etc.), analysts/ engineers/programmers, visualizers (dashboard specialists, designers, etc.). However, the industry view at large defines the kind of people you need in your world view of analytics being brought to reality.
I have often felt that this difference exists because the organizations differ in where their key decision makers are placed on the quad. Likewise, our ability to build the type of team we want to is a function of where we want to place our team, and this vision being aligned with the vision of the organization.
Try deep learning using MATLAB