Time to extend the information supply chain metaphor.

By Jeffrey Cornett

How many of you would say you are proud to work in a “data-driven” organization? If the decision-makers in your organization say this, perhaps your organization truly is data-driven. Unfortunately, this is not a compliment!

Wouldn’t you rather be information-driven, or better yet, insight-driven?

INFORMS defines analytics as “the scientific process of transforming data into insight for making better decisions.” If we truly understand the value of our profession, we should not allow our organizations to label themselves as data-driven. Decisions should be based on analytical insights, and not simply based on the raw numbers stored in our data warehouses.

The problem is in the overuse of the term “data” to include anything and everything that involves numbers. This is one reason why the science of data warehousing has been over-generalized to be the solution to all of our decision-support problems – seeking that one-truth solution through self-service reporting.

Information Supply Chain

It is time to extend the information supply chain metaphor to go well beyond the warehouse where we store raw materials (data), and also well beyond the reporting and analytical tools we use to manufacture information products (reports, graphs and analytical models).

It is time to explore the science of “information retailing” or the technologies we use to store information products so that the insights they generate are effectively learned, shared and not forgotten. The field of retailing is consumer-focused, not raw-material-driven. Yes, information retailing does depend on an effective data supply chain. Nevertheless, information retailing should be how we get our decision-making customers to actually shop for and buy the information products they need and then learn the insights they contain.

We need to be developing the technologies to stock our retail store shelves with the valuable information products we produce, and to do so in ways that make the information shopping experience enjoyable, efficient and effective. This requires going beyond broadcasting as our means of information delivery. We must create knowledge banks where all of our best information (or at least the information worth remembering) is stored.

What we store in our retail knowledge banks are the information products – not the data raw materials. These information products should be packaged and delivered in ways that maximize the generation of insights. As a profession, let us focus our optimization skills on how to collectively optimize the learning and sharing of insights, and not just optimizing the solutions to individual decision-support problems.

Some of the principles and benefits of information retailing are the same as data warehousing including self-service – anytime, anywhere. However, what we stock on our retail shelves are finished goods, not the raw materials from which finished goods can be produced. The mining of “big data” is a rapidly emerging field, but the retailing of “big information” is largely unexplored.

Information Learning Organization
Information Broadcasting

Many organizations are drowning in an ocean of data, but dying from thirst for useful and usable information. It is not because information is not being produced. The problem is that the information we produce is not well managed. Too often, it is forgotten because it was not effectively delivered and then stored as a valuable information resource worth remembering.

Typically, our valuable analytical products are e-mailed out to a selective audience. One-way broadcasting in this manner reaches a selective audience who may or may not be interested in what you sent them at that moment. Even if the analytical subject interests them, your mailing could be forgotten and lost the very next day.

Those not on your distribution list may never see this information. New decision-makers yet to be hired into those roles do not know what information they are missing, and so they have little hope of learning from those insights. Data warehousing and self-service reporting is a partial solution to catch up on what they have missed, but much better is to allow customers to browse and shop a retail outlet whose shelves are stocked with the best information products displayed in the most enticing and effective ways.

It is not the role of your customers to organize and remember the information you send them. This is especially true if you have many information products (big information) worth sharing with many customers. Rather than expect hundreds of users to organize and remember the many products you send them, you need to build a knowledge bank for them – a knowledge bank that completes the information retailing supply chain.

Role of Knowledge Bank

A comprehensive knowledge bank serves as the big box retail outlet where your best products get stored, and better yet, delivered in ways that entice your information customers to come back again and again when they seek information. This store should be designed to maximize information usability and learning effectiveness.

The role of a knowledge bank in an information learning organization can be diagrammed as in Figure 1.

The knowledge bank does not compete with data warehousing. Every organization can benefit from a world-class data warehouse solution. A knowledge bank complements an organization’s data warehousing and information manufacturing tools. A knowledge bank allows the best of your information products (as well as valuable information produced elsewhere) to be remembered, recycled and shared for use by anyone at any time in the future. Ideally, your information supply chain should contain optimal solutions for all three: data warehousing, information manufacturing and information retailing.

The most common approach to a knowledge bank is to simply dump your reports onto your website, sometimes aided by the selective querying of an online data mart. Extranet and cloud solutions can be employed where the information is confidential. Website-style publishing counts as a knowledge bank, but may be well suited when only a limited amount of data or information is to be shared. However, when you have big information to share, with complex insights to be explored, you need a better solution than dumping 500 reports (perhaps alphabetically sorted) onto your website retail solution. This is where the field of knowledge management comes into play.

Science of Knowledge Management

Knowledge management (KM) is taught in universities, but most commonly as a component of library science rather than an extension to OR/MS analytics. Perhaps every OR/MS curriculum should include a lecture at some point in a student’s education that introduces the science of how to prevent your best work from being forgotten the day after it is e-mailed out.

Many schools offer degrees in knowledge management. According to the University of North Carolina School of Information and Library Science, “KM focuses on processes such as acquiring, creating and sharing knowledge and the cultures and technical foundations that support them.” The school’s website provides an “Introduction to Knowledge Management” [1] and lists nine knowledge management supporting technologies including decision support systems, extranets, intranets and data warehousing (perhaps someday information retailing?).

The UNC site also summarizes the results of a survey to assess the present and future state of KM in an organization. This survey introduces a six-point scale for how well KM is practiced in an organization: 1. not considering, 2. evaluating, 3. planning, 4. pilot projects, 5. implementing and 6. currently practicing. Perhaps a scale such as this could be included in the “Analytics Profession Maturity Model” initiative now being developed by INFORMS [2].

Coming Soon: An Example

Let me close this essay with a quote from “Choruses from the Rock” by T. S. Eliot as uncovered by Mayank Trivedi [3].
“Where is the wisdom we have lost in knowledge?
“Where is the knowledge we have lost in information?”
Information sciences can help us learn what we need to remember, and remember what we have already learned. As an extension to analytics, let’s label this field “information retailing.”

Look forward to an article in an upcoming issue that describes an interactive knowledge bank we have developed at Ivy Tech Community College to summarize our region’s best institutional research over the past two years. Because it was so obviously needed, we jumped right to the “implementing” and “currently practicing” stages of KM evolution. We are now trying to back up and evaluate what we did, and why it has been so well received. ORMS

Jeffrey Cornett ( is director of Institutional Research at Ivy Tech Community College-Central Indiana.

1.     Liu, Sunny, and Parmelee, Mary, “Introduction to Knowledge Management” (2002), University of North Carolina School of Information and Library Science; accessed March 16, 2014 (
2.     Robinson, Steve, “The state of INFORMS and the profession,” OR/MS Today, December 2013.
3.     Trivedi, Mayank, “Knowledge management in Health Science libraries,” Journal of Academic and Special Librarianship, Vol. 8, No. 2 (Summer 2007).