-
-
Marketplace
-
Channel Resources
Articles from this Site
-
One brand, one Web site! DM Review is now the home of all the content you're used to at BIReview.com and much more. If you are registered at BIReview.com, you're already registered at DM Review. If not, take just a moment to sign up for all the free services we have for you at the new DMReview.com.
Building the Analytic Organization
Over the past 20-plus years the concepts of business intelligence, performance management and analytics have evolved with different approaches on how to leverage information to drive results. This has come in the form of technology to address user needs; architecture to centralize an organization's data; competency centers to grow the internal analytic capabilities; and development performance correlations through analytic experts.The first two major stages in this evolution addressed two core competency needs - Data Warehousing and Business Intelligence - that an organization must have to leverage information more effectively. The first-stage focus was on technical and architectural components needed to support and leverage corporate data assets, with emphasis on historical or lagging indicators of the business that did not address how to use the data to drive results. In the next stage, (today's status quo), organizations are evaluating analytics as a core competency. This is happening in business intelligence competency centers and within small analytic groups developing complex algorithmic business models. The gap that remains is to truly drive analytics across the business, and so the next evolutionary stage will be to think of analytics as a function - like finance, sales or marketing - and not just as a competency. In practice, the organizational analytic function would be run by a Chief Analytic Officer (CAO), supported by a dedicated group of functional and technical experts.
THE ANALYTIC ORGANIZATION
The ability to extend value creation through analytics requires a distinct organization responsible for all the analytic needs of the business. The CAO would oversee this organization and provide business and technical expertise. The CAO would also provide analysis capabilities to other functional areas such as finance, sales and marketing, and direct data collection and information distribution through the creation of a robust analytic technology architecture.
To make this change and develop the formal analytics function, the organization needs to apply enterprise analytics across all functions, make analytic sharing part of the culture and create a robust set of IT systems. Along with these key organizational changes, a set of four core actions is required to establish the function:
ACTION #1: FORM THE ANALYTIC FUNCTION
The first step is to make an organizational commitment to analytics as an enabler of the management process. This involves the actual creation of the leadership role and assigning a group of individuals responsible for analyzing the business to drive results. This group would consist of both cross-functional experts and technical resources from the IT organization.
ACTION #2: DEVELOP ANALYTICS AS A COMPETENCY
The core competency of the analytic function is a strong analytical capability tied to the enterprise business model. This includes an understanding of the strategic objectives and drivers behind each of the objectives. The ability to translate business concepts into analytic decision models increases the organization's analytic quotient to make fact-based decisions. This goes beyond the traditional query and reporting capabilities that focus on historical performance and starts to address future performance of the organization. This includes areas of predictive analytics, business modeling and scenario planning, data mining and root cause analysis. In his article "Politics and Provisioning"1 Tom Davenport offers the concept of a BI oligarchy with a small but elite group of analytic experts that develops robust analytical models to better understand and predict the performance of the business.
In addition to this core of expertise, analytic concepts need to be applied across the entire organization. The ability to leverage information in the decision-making process needs to be ingrained into the entire enterprise. A prime duty of the CAO and analytic function would be to educate and enable the entire organization on the concepts of analytic decision-making such that it becomes part of the culture. This is analogous to how organizations such as G.E. use Six Sigma in the operation of the business.
ACTION #3: GET DATA IN ORDER AND IN LINE WITH THE ANALYTIC NEEDS
Since the amount of data being produced internally and externally is growing at a faster rate than the organization's ability to process and consume it, it is more valuable to look at what data is critical to support the analytic needs. In Eric Beinhocker's book "The Origin of Wealth"2 the author discusses the concept of wealth as being based on information fitness. Ordered information creates knowledge, while information on its own is considered worthless. The same can be true when thinking of corporate data assets. Thus, information that is not placed into the context of the business can be considered worthless.
What is considered context? Quite simply it is the language of how the business would be described. This includes business definitions, attributes and calculations that facilitate a consistent view of the enterprise. As simple as that sounds, context is nearly impossible for many organizations to define and so they end up focused on the "how" solution to resolve issues. In addition, when analytics are applied to better understand business performance, you start to create knowledge and apply meaning to the data.
ACTION #4: EVALUATE HOW ANALYTICS WILL BE DISTRIBUTED AND CONSUMED BY THE ENTERPRISE
The final core action is to "release" the analytic capability created to the rest of the organization. The challenge is to understand that there is no "one size fits all" approach to delivering analytics across an entire organization. First, there are different types of analytics (static reports, dashboards, scorecards, guided analysis, alerting, and ad-hoc analysis) that can be delivered in different formats (paper, PC, smart phone). Second, the vehicle for distribution will differ depending on the consumer audience (executives, line-of-business owners, analysts, power users, etc.).
Both of these issues must be addressed by creating a matrix that maps needs to the entire audience. Once this is complete, a full analysis of the tools and technologies that can properly service each of these needs can be established. The best outcome may not be in choosing a single vendor platform, but in a toolbox available to the analytic function to select the proper technology based on the audience and need. The change is to separate the business functions from selection of the analytic tools; this puts the onus on the analytic function to have a set of tools to deliver enterprise analytics.
This shift could also mean removing responsibility from the IT function or breaking down the information silos still being created within functional areas of the business. Companies that make and adhere to these commitments have a much better opportunity to fully enable their management processes with analytics.
References
1. Davenport, Thomas, "Politics and Provisioning", BI Review Magazine, March 2006.
2. Beinhocker, Eric, "The Origin of Wealth", Harvard Business School Press, 2006
For more information on related topics, visit the following channels:
