Thats a hell of a title...most often EDW development end up in tragedy. Examples are many.
There are however a few examples of organisation that actually have done a very good job, proofing the case of an Enterprise Data Warehouse and it's value for the business.
I wanna make this a short article....Technically there is no excuses - it's all there. My CSF's are simple:
- Highest management commitment....no, middle management won't do
- Entrepeneurship in management, but especially in the project management. This aint 'just a project'...
- Inter-disciplinairy approach in the development of EDW increments. Business and IT need to go hand in hand here. A succcesfull EDW WILL ALWAYS lead to a compentence center where expertise is centralised, budgets are centralised etc...Don't fight it...accomodate it. After all, Information is something that extends marketing, logistics or business units...it's a generic function.
- When u start..gather expertise in the market, do not try to re-invent the mistakes...oops sorry..the wheel. There is a lot of expertise out there. Try to find people with a passion for this trade and with an overview of the whole process, not just one part. Find specialists with a broad scope.
- Make a plan and stick to it....
- Make your first delivery something that can be celebrated....don't make the mistake of starting a first increment that is complex, uncertain or not well defined. DON't do it...
- What a respected EDW/BI manager of a large company told me last week; marketing marketing marketing....sell it in the company..don't go underground. Such a wise lesson.
- Oh and one point i wanna make...A good design is the difference between succes and failure, dont be tempted to go for fuzzy temporarily workarounds.