The business intelligence (BI) process at its best consists of envisioning a response that creates the possibility of business value through working smarter for an enterprise and tracing that through the labyrinth of information sources, system interfaces and transformations of meaning undergone by the data, resulting in delivery of that value. Oftentimes, the business value envisioned requires surfacing, formulating or aggregating answers to questions of the form, "What customer is buying or using what product or service? When and where are they doing so?" Process is a coordination concept; the BI process coordinates the key terms in this question. Of course, this can occur in 1,001 different variations and extends to all imaginable corporate entities, such as employees, suppliers, documents (e.g., invoices), events (e.g., deliveries) and whatever drives your business. In every case, the critical path to BI lies through building a consistent unified representation of customers, products and other essential corporate data entities.
In short, business intelligence is now an essential part of every business process and deserves to be acknowledged as such. In order to understand how and why BI is now a priority process, having a high-level schema for organizing business processes will be useful. Fortunately, one of the most concise and powerful schemas is available in the work of Peter G. W. Keen.1 This framework distinguishes the following four fundamental processes:
An identity process is what defines a firm's fundamental way of doing business. It is what differentiates a firm for its customers, employees and investors. For example, McDonald's is identified by its fast service with consistent (though not gourmet) quality; FedEx is guaranteed, on-time delivery; Nike is athleticwear for doers, not procrastinators on the sideline.
A priority process is a key enabler of corporate results. It determines how a corporation stands relative to the competition and how the identity process gets implemented. For example, for McDonald's, food supply, logistics and distribution are priority processes. Aircraft and motor-pool operations are a priority process for FedEx. Customers generally do not have visibility to such processes unless there is a breakdown and the package doesn't arrive on time. Even then an exception process kicks in, and you get a phone call or e-mail.
A background process is an administrative or overhead function required to support day-to-day operations. For example, payroll is a key background process. Document management, accounting and benefits administration are other examples. Unless such processes are badly broken, reengineering them rarely provides business value. Such processes can become a liability, not an asset, if pursued without the necessary competence. If you are not in the business of running payroll systems - as is ADP, where payroll is actually an identity process - this and related background processes are strong candidates for outsourcing.
A mandated process is (by definition) required by a legal or regulatory authority. The business case for regulatory reporting and filing tax returns is obviously the severe penalty for failure to comply. From the perspective of the individual enterprise, such processes are generally a liability and no one would do them except for the sanctions attached. You have a limiting case in old-line industrial firms where it is occasionally less expensive to pay the fine for polluting rather than replace and rebuild the factory from the ground up.
A final kind of process is a folklore process - one that no longer moves any part of business operations, an idle wheel. An example would be a report for a single executive who left the company shortly after its implementation but which continues to be produced.
Business intelligence as a process serves to enhance the value of a firm's other identity and priority processes. BI is a process value-builder when rightly deployed in relation to other business in what might be described as a meta-process. This can turn processes which are struggling, stumbling and thrashing from liabilities to assets. Alternatively, it can reveal that such processes should be abandoned or outsourced. BI does this by:
- Managing by the numbers. A single fact is worth a thousand words. BI applications are required to tell firms which 20% of customers are responsible for 40% of revenue; which 20% of products are responsible for 60% of costs; and which customer segments or brands are the most profitable. This is just the tip of the proverbial iceberg, with many other particular combinations being near and dear to a particular enterprise or executive. Knowing the lifetime value of a customer implies aggregating a lifetime of transactions, which is what a data warehouse does best.
- Providing a single, canonical state of the business (version of the truth). Too often, companies put data into a data warehouse and also store it in a plethora of other data stores. If a data warehouse has to be match-merged with dependent data marts to provide needed information, then it is a potentially useful data store, but it is not integrated. Admittedly, this can be an organizational ("political") issue; but those firms with managerial courage, communication and leadership are the ones that profit as their competitors incur the coordination costs of business design by politics.
- Substituting information for effort. By coordinating commitments and exposing emerging trends, capturing and storing time-series data about sales, distribution or shipments, firms are able to track market trends, make inferences about market hot spots and forecast future demand. Once a demand plan is available, product inventories can be adjusted downward to correspond to demand. This is pure cost savings.
- Integrated BI represents a closed-loop process. Closing the loop is a key differentiator of integrated BI from mere active data warehousing. In particular, the integrated BI is used to optimize processing in the upstream operational or transactional system. The operational systems feed the data warehouse, which, in turn, feeds back to the operational system to optimize the relevant transactional processing. The interfaces go in both directions. Business intelligence points to and reveals operational efficiencies and, as integrated, can be described as driving operational processing.
- Keen, Peter G. W. The Process Edge: Creating Value Where It Counts. Boston: Harvard Business School Press, 2001.
Register or login for access to this item and much more
All Information Management content is archived after seven days.
Community members receive:
- All recent and archived articles
- Conference offers and updates
- A full menu of enewsletter options
- Web seminars, white papers, ebooks
Already have an account? Log In
Don't have an account? Register for Free Unlimited Access