Analytics has evolved. So have the technologies enabling data analytics. Of all the sensational displays across the landscape of data analytics technologies, a glaringly unmistakable scene is the mushrooming of new entrants in this soon-to-be-crowded space (if not already so). Such proliferation is attributed in part to the greatly expanding promises of analytics in optimizing operational processes and sharpening strategic lenses ultimately allowing businesses to carve out and sustain greater competitive edges when catering to ever-changing consumer preferences and expectations.
What are the key parameters to be considered in optimally-configuring a technological setting for enterprise-wide data analytics? Before making an attempt to tackle this, it’s helpful to briefly visit what the business side of an organization typically demands when it comes to data analytics. Such demand for analytics basically consists of three folds: reporting, which is a largely data reduction and descriptive analytics exercise–straightforward yet sometimes still a challenging feat to get it right; predictive analytics, in essence a propensity-focused probabilistic modeling endeavor which is quite useful but at times fraught with inadvertent misuses or overstated efficacies that can hardly be validated; and prescriptive analytics, a pinnacle of analytics (vastly promising) when analytical insights are interwoven into the business to impart fresh knowledge, new perspectives and novel concepts that ultimately inform business strategies and their tactical executions with regard to new products, new services, new markets, and the like.
For data analytics to be successful, it is essential to see its technologies configured properly within a given organization. To achieve this end, several parametric factors need to be thoroughly vetted. It is plainly clear that one has abundantly more choices of data analytics technologies and many different ways of configuration in recent years than a decade or more ago. Such configuration may then intrinsically vary from one business setting to another. However, a crude yet (hopefully) helpful yardstick may be a cognitive assessment and determination of the optimal cross-section between the complexity of the business and the complexity (and robustness) of the technological configuration for data analytics, as can be roughly illustrated in the diagram below.
Data is the foundational pillar for analytics. Quite a while ago, data warehousing with the central feature of ETL-powered data standardization and integration was the norm for enterprise wide data repository facilitating reporting and analytics (including productionized analytics-driven business programs). Today, this may still be the viable option for complex organizations. However, for many organizations on the lesser side of the complexity spectrum, leaving the data at the source (such as data repository attached to business system) while deploying the analytical tools to reach to the sources represent a more optimal approach thanks to the bounty technological advances in recent years. The data environment that comes with a business system these days is a day and night difference from those a decade or two ago. Equally noteworthy is the robustness of analytics technologies nowadays in terms of both the vast processing power and their abilities to fetch and integrate data from a variety of sources.
An analytics-conducive data environment is not complete until it is situated and integrated with the analytics technologies rightly fitting the organization. When evaluating analytical tools, one might find it beneficial to heed the following considerations. First and foremost, for a viable analytical environment, it’s seldom true that one tool may do everything in a versatile and efficient way. Turnkey solution for data analytics at a given business organization is likely more of a myth than a practically achievable reality. Instead, one usually finds the need for a set of tools working in a complementary fashion to meet the three-fold demand for data analytics across the organization.
Secondly, overinvestment is equally deplorable as underinvestment. While inadequate budget is frequently cited as a constraint in growing data analytics capabilities within an organization, it is not uncommon to observe such technological investment go significantly underutilized. Determining the data analytics technologies that are the right fit can prove to be a daunting task, and this is still a minefield especially at a time when the promise of automated analytics has become such an irresistible lure. Experience, expertise, and wisdom gleaned from learnings industry-wide and idiosyncratically are essential to guiding the development and recalibration of the technological environment for data analytics. Such an environment would facilitate data analytics to be dynamically immersed with the operational processes and management decision-making of the growing business.
Thirdly, the delivery of the data analytics technologies is expeditiously changing, unequivocally for the better. While the traditional approach of deploying data analytics tools on premise still makes a lot sense for large and complex organizations, the availability of such technological tools via the cloud is becoming more and more attractive, especially when data security related concerns (being addressed with the rapid advancing of technologies) are going to be no longer a prohibitive barrier for organizations with a tall responsibility for stewarding sensitive customer information. Analytics as service, which also allows an organization to more effectively cope with the scalability challenge that is inherent in every growing business, is hence poised to grow and is probably going to do so in a much more significant pace than what we may anticipate in the upcoming years.
Lastly, technologies for analytics are still tools albeit immensely useful, powerful, and transformative. Users of such tools, or analytics professionals (often referred to under fancier titles), will in the end have to be on the driver seat, not the tools, when business organizations strive for optimizing the configuration of their data analytics technologies. Nurturing a well-rounded team of analytics professionals with a collective skill set (ranging from technical strengths across the technological tools to innovative thinking, solid domain expertise and strong business acumen) is the crucial ingredient of the formula for a thriving data analytics environment. At the center of this maze-navigating journey is the leadership of analytics, intrinsically embedded in the business, which ultimately helps ensure a dynamic equilibrium of the supply of and demand for data analytics within a given enterprise organization and guides the organization away from the potential pitfalls and detours that are not only painful but also costly.