The ultimate goal of the Corporate Information Factory (CIF) is to get the right information into the right people's hands when they need it. To date, much of what has been written about the CIF architecture covers the design and implementation of the strategic components (data warehouses and data marts) used to accomplish this goal. This article covers the decision support capabilities available today for use in solving a business user's difficult tactical problems. Within the CIF, these capabilities are supported in structures called operational data stores (ODSs) and a new component, the oper mart, for tactical analysis.

An increasingly important need today exists for tactical analyses to be performed on integrated operational data. Let's look at two examples of this need.

Suppose a corporation launches a new, highly targeted campaign through its Internet site. The campaign is aimed at specific customers for specific products through a limited sales channel. The marketer wants to monitor the campaign closely and tweak it if he discovers that it is not tracking where he wants it. To perform this task, the marketer needs current data containing a subset of customers, their interactions with a single sales channel and the results of those interactions compared to an expected outcome scenario.

Another example of this need could be an insurance company that becomes aware of a hurricane heading for southern Florida. The company would like to know its exposure in terms of policies for property and personal belongings in that area of the country. They need a current view of the policies, policyholders and outstanding claims for that area.

Daunting, you say? Impossible perhaps?

If you have constructed an operational data store containing the current, integrated data on your customers, their customer touchpoints and current set of products, you have a start at satisfying this type of analysis. However, you must go further and build the oper mart to achieve the performance and ease of use required for business analysis.

What is an Oper Mart?

Over the years since the publication of Building the Operational Data Store1, we have learned a great deal about ODS utilization and construction and its importance as a CIF component. Furthermore, we have discovered the need for a new structure, heretofore unnamed, that is used for tactical analysis. This structure is called an oper mart. It is similar to the data mart in that it is a smaller subset of data available in the ODS; however, it differs significantly in its currency, usage and business community members using it. The characteristics of an oper mart are:

  • Small subset of operational data store data used in tactical analysis.
  • Stored usually in a multidimensional manner (star schema or hypercube).
  • Updated by transactions occurring in the ODS (no history stored here!).
  • Created in a temporary manner and dismantled when no longer needed.
  • Sourced from the ODS.

Oper marts are created when there is a need to analyze current operational data in a multidimensional fashion. Our marketer or insurance analyst is a prime user of just such a temporary structure. He or she may need to analyze this data in several different ways to get an understanding of the current situation. These people need immediate access to current data in an easy fashion, yet they cannot impact the functioning of the operational systems. They need an oper mart that is quickly constructed, easily understood and constantly rebuilt as the day's business progresses.

A Temporary Component

We are starting to see oper marts being built from ERP systems, ODS implementations and even some legacy systems. These small cubes of data are used to answer the very real need for operational data in a multidimensional array. The most common implementation of an oper mart is created from an ODS because the ODS offers the best source of integrated and current data. However, there are several warnings that should be ringing in your head at this time.

Like all things, these components should be used conservatively; and you must have a rational plan in place for their construction, maintenance and eventual destruction. You should not begin spitting out thousands of these cubes of data, or you will eventually suffer from oper mart chaos much like the data mart chaos plaguing many organizations. Control the construction, and monitor the usage carefully.

Oper marts are not all-purpose tools and should never be used to replace strategic data marts or the data warehouse. Their purpose is not historic in nature. Rather, it is for only the most up-to-date information of a scenario. They contain only current data and should be used in that capacity. Do not circumvent the purpose of the data warehouse, its data marts and strategic analyses.

Oper marts are temporary in nature and should be destroyed when no longer needed. These are not structures that generally serve a permanent purpose. Therefore, you need a mechanism to track their construction and usage so that you can determine when they are no longer needed and can be deconstructed to aviod oper mart chaos.

The operational data store serves an important and useful purpose, especially in your CRM implementations. It is an ideal source for enterprise integrated, current data. However, the ODS may not be able to handle the quick, multidimensional, current analyses needed to understand a critical, immediate situation.

That is where the oper mart is effective. This cousin to the data mart allows business users to have the same luxury of multidimensional analysis in a current, operational world – something unheard of a few years ago. The key to the construction and maintenance of these new components is to understand their temporary nature, the tactical analysis purpose they serve and the ability to control their existence and eventual destruction.

References

1. Inmon, William H., Claudia Imhoff and Greg Battas. Building the Operational Data Store. New York: John Wiley & Sons, 1999.

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

Don't have an account? Register for Free Unlimited Access