In the world of data warehousing the concept of “one version of the truth” has been around for a long time. And it continues to be central to the architectural philosophy of any enterprise data warehousing system.
In fact, in a recent InformationWeek commentary, the editor, Chris Murphy, says, “Employees need one data source they can trust.” Murphy acknowledges that the concept of one data source is difficult to establish.
And he quotes a number of executives who suggest that an enterprise should create one version of the truth by simply declaring it as such. The declaration will be a motivating factor to get the data right and “more accurate,” according to the executives quoted.
But what if the declaration doesn’t work? And what if the organization’s decision makers require a slow, managed transition process from multiple versions of the truth to a single version?
As a first step – the data management office, under the direction of the chief data officer, should put together a solid business case supporting the need for the single version of the truth. If the proposal is not adopted, then management should be prepared to manage multiple business intelligence (BI) installs.
In her article, Managing Multiple BI Tools Doesn’t Have to be So Hard, Nancy Williams offers some advice for managing multiple BI applications. She says, “A multi-tool business intelligence environment may be the norm, but it’s not the ideal.” In other words, multiple BI tools can be difficult to manage and will certainly increase costs.
Williams lists the following five suggestions to simplify the management and control the cost of multiple BI installs:
- Start by putting in place an effective BI governance policy. The policy should dictate how and under what conditions a new BI tool should be introduced.
- Aim to reduce the number of BI tools used within the organization by migrating common purpose reports and BI functionality to the application(s) that can handle the task(s) best.
- Remove any outdated reports and keep the reporting processes up to date.
- Ensure that any BI application in use is the latest version. Update to the latest version (after it has rigorously been tested and soaked, allowing for the bugs to fall out) to ensure that your organization can benefit from any improved functionality and to decrease complexity.
- Limit the number of servers and data stores by consolidating the infrastructure. Point to as few data sources as requirements dictate. Optimize the selected platforms so that there are no complaints about performance.
The author further suggests that by using the BI Pathway Model (a project management methodology practiced by DecisionPath Consulting), an enterprise will have at its disposal all the foundational components needed to manage a multiple BI environment.
- Subscribe to our blog to stay up to date on the latest insights and trends in business intelligence.
- Check out our complimentary “5-Minute Guide to Business Analytics” to find out how user-driven “analytic” or “data discovery” technologies help business and technology users more quickly uncover insights and speed action.
Dennis Earl Hardy
Spotfire Blogging Team