Continue in 2 seconds

Data Warehouse Killers

Published
  • April 01 1999, 1:00am EST

When it comes to building a data warehouse, mastering the technology is the easy part. The hard part is handling the political and cultural issues. These "soft" issues can "kill" a data warehousing project faster than a turkey on Thanksgiving morning.

Political and cultural problems tend to fall in three primary categories:

Expectations Management. The inability to set realistic expectations for the data warehouse in terms of functionality, performance, scope and schedule.
Change Management. The failure to ensure that end users adapt to new ways of analyzing data and IT managers adapt to new approaches for designing databases and delivering systems projects.
Data Management. The inability to implement policies and procedures that ensure sufficient data quality, and the failure to manage users' perceptions of data quality.

If your data warehousing project runs aground on the shoals of political and cultural problems, it's likely to exhibit the following symptoms in each of the categories. So beware!

You know you have an "expectations" problem when...

  • End users think they'll get "fast" (i.e., sub-second) responses to all queries.
  • End users expect the warehouse to define and calculate data the same way their existing operational or analytical systems do.
  • Business managers threaten your career if the initial data warehouse doesn't support their group's information requirements.
  • Executives think the project is simple. ("Just move some data from one system to another.")
  • Executives think you can deliver the entire vision for the data warehouse on the first iteration.
  • IT professionals think they can deliver the entire data warehouse in the first iteration.
  • IT professionals believe a data warehouse project can be completed.
  • IT professionals don't think a data warehouse is a mission-critical system.

You know you have a change management problem when...

  • Six months after deployment, users are still using the "old" system or paper reports to get data they need.
  • Six months after deployment, users are still calling you to create custom reports.
  • Your users can't fathom "OLAP."
  • Your internal help desk is still flooded with calls six months after deployment.
  • Users continually issue runaway queries, bringing the data warehouse to its knees.
  • IT professionals don't have a strategy in place for prioritizing and responding to change requests in a clear, easy-to-understand fashion.
  • Query performance is worsening because IT professionals didn't anticipate end-user requirements.

You know you've violated a data management taboo when...

  • You've offended the "owners" of a source system by pointing out integrity problems with their data.
  • One department doesn't want to share its data with another department even though both would benefit from each other's data.
  • The data in your new reports is accurate, but users think the data is wrong.
  • Your users are overwhelmed by the amount of data in the warehouse and don't know where to start.
  • You've failed to incorporate certain business rules for interpreting or collating data.
  • You have no procedure in place to reconcile your data with source systems.
  • Users point out errors in your data, and they lose confidence in the warehouse.

What Can You Do?

If your project is afflicted by these symptoms, it may be too late to save it. The best strategy is to address political and cultural obstacles before you begin the project.

The key to success is to become both an astute politician and savvy marketer. Although most IT managers feel uncomfortable in these roles, no data warehouse can succeed without someone who applies political and marketing muscle to run interference for the project.

Specifically, data warehousing managers must find guidance and refuge in a business executive who champions the project. Second, data warehousing managers must use every communications channel available ­ newsletters, meetings, e-mail, external magazines, etc. ­ to drive home the goals and deliverables of the project. Third, the data warehousing manager must identify individuals who can potentially thwart the project and work to gain their trust and cooperation. This typically means getting these folks involved up front, asking for continuous input and feedback and keeping them abreast of all developments.

Keeping a data warehouse from capsizing is challenging work. But addressing political and cultural issues up front can maximize the chances that the data warehouse will succeed.

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