Becoming a BI-(or Analytics-) Competent Organization
I have to preface that the terms of Business Intelligence (BI) and Analytics are used interchangeably in this article. I feel like somewhat excused to do so, as the unit I manage has been called Business Intelligence since more than a decade ago albeit its changing organizational affiliations and varying job titles of mine.
Analytics benefits from not only a good reporting know-how but also the very same data infrastructure underlying the reporting function
BI, in essence, is for IB, (making) intelligent business. BI is a critical function for an enterprise in a rapidly changing society whereby consumer needs for products and preferences for service interactions are constantly being shaped by technological advances. BI enables an organization to run routine productions, optimize its processes, discern signals from noises, and craft innovative strategies and tactics along with its products, services, and customers. When done right, it ultimately confers competitive edge allowing a company to reap bottom-line benefits, ranging from reduced operational costs to new revenue streams. Business Intelligence has been around for quite a while. It has certainly progressed in very significant ways, especially in the past decade or so. BI technologies have advanced, the adoption of BI across industries has been more widespread, and the BI function within a given enterprise is nowadays more entrenched. Nevertheless, challenges remain. Over-investment and under-delivery (of a tall promise) are not rare to observe. Lukewarm acceptance and bottlenecked growth are not uncommon to encounter. Organizations thriving on their ever-evolving and constantly-optimizing BI competency are probably still in the minority camp across the spectrum.
A BI-competent organization starts with a good reporting environment. Categorically, reporting consists of two types: operational reporting and metrics reporting. The former is largely a data reduction (via various summarizations) process, whereas the latter is a distilling endeavor to transform or synthesize various operational stats into relevant metrics underlying the business performance. The former serves a foundation for the latter, and the latter leverages the former and further helps the former from drowning decision makers. Getting this two-fold reporting right embodies not only a solid data environment, the fulfillment of the prerequisite for BI, but also the proper configuration of the BI tools within a given organizational setting. Operational reporting is indispensable for frontline managers to perform their production-specific functional tasks, and metrics reporting, often in the form of KPI’s (key performance indictors) or KPM’s (key performance metrics), is a necessity for upper management keeping a pulse on the business while navigating through the competitive maze.
Analytics, or predictive analytics, is a logical extension of the reporting prong of BI when cognitive explorations such as why, what, and when naturally become next level of business inquiries beyond reporting. BI can survive on the reporting prong but prospers on both the reporting and analytics prongs. Analytics benefits from not only a good reporting know-how but also the very same data infrastructure underlying the reporting function. When analytics becomes an integral part of an organizational being, it helps impart business insights that inform strategies and tactics with respect to products, services, customers, and processes. In a global economy where competitive advantages are often short-lived, analytics is an enabling agent for enterprises to constantly push the innovation frontier when it comes to product development, pricing maneuvers, market positioning, strategy recalibration, and the like. Organizational competency on analytics is an enterprise asset in every sense.
Thanks to the Big Data trend characterized by velocity, volume, and variety, the machine learning branch of traditional analytics has somehow morphed into its own being: artificial intelligence (AI). AI is by and large automated analytics augmented by guided data-fueled machine learning via iterative optimization. AI is robustly versatile and prescriptive oriented. It surely qualifies for the third prong of an enterprise BI competency. AI did not just emerge recently. It is at least as old as the early BI and analytics around the turn of the century when data mining powered by neural nets was a fad. Spurred by the Big-Data-led paradigm shift in analytics and related technological advances in recent years, AI is gaining momentum in its growth. AI is still in its infancy stage but is immensely promising, with a growing range of enterprise applications in the areas such as customer service, fraud prevention, anti-money laundering, pricing optimization, product recommendation, and financial advising.
A BI-competent enterprise is an organization where analytical endeavors are not an after-thought or a mere exercise of validation but intrinsically embedded in the corporate decision making process. In such an environment, analytics is central to the very design thinking in conceptualizing, developing, testing, implementing, and recalibrating business processes and programs tackling production problems and growth challenges. A BI-competent organization often boasts a solid data environment, a set of complimentary technological tools, and a team of knowledge workers. Needless to say, data is the first and foremost pillar in the BI competency pyramid. A good data environment is frequently characterized by the integration of various disparate sources, a rich reservoir of information (about customers, products, programs and processes) both longitudinally (history) and horizontally (dimensions), and a high level of secured accessibility that is tool-agnostic. To achieve this end, a delineation of the two-fold data ownership can be helpful: while IT has the technical aspect (staging, storage, backup, access, etc.), the business must own the content concerning intricacies and nuisances such as data integrity, content validity, anomalies, extreme values, missing value patterns, imputation, and so forth. Evidently, an organically integrated data environment, as opposed to departmental silos, is already half of the success in an enterprise’s journey to BI competency. Second, implementing a set of technological tools that are complementary is important. Turnkey BI solution is more likely a myth than a practically-achievable reality. Organizational fitness is the key. Technological choices are abundant. Selecting the right set of tools starts with a proper assessment of the business complexity of the organization as well as its maturity level in the spheres of data and analytics. Thirdly, developing a team of knowledge workers comprising database developers, business analysts, statisticians, and analytics managers is essential. Healthy knowledge spillovers not only stimulate productivity but also build up valuable institutional learning repository.
The hallmark of a BI-competent environment is the dynamic equilibrium of the supply of and demand for enterprise BI whereby the fine supply of metrics and insights, informing and empowering business decision makers who ultimately become advocates for data-driven culture, helps secure the steady demand and fresh business challenges, in turn, provide fresh impetus to spur the supply side for continuous analytical innovations. It is rewarding to be organizationally competent on BI and analytics. However, BI competency cannot be imposed or grafted upon to an organization overnight. Business analytics has to be rooted in the organization and immersed with corporate decision making processes, both operational and strategic. It takes time to grow and evolve. While following the herd may be instinctive, carving out a pathway tailored to the organization, which requires expertise, wisdom, and measured discipline, can spell much higher chance of success. There are many things that could go wrong along the journey leading to overinvestment and misalignment. Ultimately, the business side has to be on the driver seat and the right BI leadership has to be in place if one aims to avoid the painful pitfalls and costly detours.