• Do not launch the data warehouse unless and until your company is ready for it.
  • Find the best executive sponsor. Ensure continued, long-term, and committed support.
  • Emphasize the business aspects, not the technological ones, of the project. Choose a business-oriented project manager.
  • Take an enterprise-wide view for the requirements.
  • Have a pragmatic, staged implementation plan.
  • Communicate realistic expectations to the users; deliver on the promises.
  • Do not overreach to cost-justify and predict ROI.
  • Institute appropriate and effective communication methods.
  • Throughout the project life cycle, keep the project as a joint effort between IT and users.
  • Adopt proven technologies; avoid bleeding-edge technologies.
  • Recognize the paramount importance of data quality.
  • Do not ignore the potential of data from external sources.
  • Do not underestimate the time and effort for the data extraction, transformation, and loading (ETL) functions.
  • Select the architecture that is just right for your environment; data warehousing is not a one-size-fits-all proposition.
  • Architecture first, technology next, and only then, tools.
  • Determine a clear training strategy.
  • Be wary of “analysis paralysis.”
  • Begin deployment with a suitable and visible pilot to deliver early benefits.
  • Do not neglect the importance of scalability. Plan for growth and evolution.
  • Focus on queries, not transactions. The data warehouse is query-centric, not transaction-oriented.
  • Emphasize and distinguish ...

Get Data Warehousing Fundamentals for IT Professionals now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.