sql server data warehouse: We have a couple data marts and a handful of users accessing the data =
mostly through Cubes. We work for a govt agency that doesn't have much =
money - so our analysis tools are limited to using Reporting Services; =
Report Model to query the cubes. =20

We are now having to fight to continue to develop more data marts. Some =
of the challenges were how long it took to build a data mart and ROI. =
It took about 4 to 6 months to get our last business section up and out =
onto production - but most of our team were using new tools and learning =
how to use SSIS, SSAS and follow the Kimball methodology. Any future =
cycles will go much faster as we have much needed experience now.

Another challenge is lack of money to provide our users with better =
tools such as Proclarity. 95% of all programming is done with Java =
whereas our DB Group is mostly Microsoft so there is always that touchy =
area and sometimes explosive topics when we talk about using VSS which =
integrates nicely with BIDS verses using CVS or other mostly free tools.

I am tasked with putting together some suggestive direction topics for =
management to decide if we even continue any BI efforts and was hoping =
someone might have some articles they could point me to to help keep us =
in a visionary direction - I firmly believe this BI stuff is where we =

I find most of the BI vision stuff to be garbage.

http://www.bptrends.com/deliver_file.cfm?fileType=publication&fileName=02-07-COL-BPMMWhatWhyHow-CurtisAlden-Final.pdf

This is a paper on Business Process Maturity. It's a low level
introduction to the concept. The idea is that an organization is
merely a collection of processes that support a common mission. All
staff and resources in that organization exist to design and perform
those processes. Business Process Maturity is a way to gauge how well
your organization is supported by existing processes, and how well
your organization can change its processes to adapt to change in the
external environment.

So how does BI fit in? I find that most BI vendors are information-
centric; not surprising, given that this is what they're selling. But
information doesn't mean anything outside of your organization's
processes. All processes need information: processing an application
requires information from the applicant and external data. Maybe an
operational process needs feedback to self-correct. Even the process
of setting high-level strategy requires information. BI is only
useful in the context of supporting those processes.

Along those lines, BI maturity is closely tied to BPM.

In terms of your organization, my advice would be to identify the
current processes in your organization that have the most pain points
(resource intensive, high operating risk, etc). Then figure out which
information inputs you can provide into the process to improve its
quality. In that way, you advocate both BI and BP. It also
translates nicely into projects that you can pitch; change the
conversation from "do we?" to "which one do we do?"

Remember that BI and data warehousing are not business strategies;
they're IT strategies. BP is a business strategy. You become most
useful to your organization by intimately knowing its processes and
coming up with innovative ways to improve them.

As far as platforms go, I'm an open source guy, so I always ask, "Why
MSFT?" Check out Pentaho; Mondrian is a ROLAP cube in that suite.
Also check out OpenI, which is a Java servlet based web tier for MDX.
Finally, there's BIRT on Eclipse for SSRS like functions.

As for source code control, all of the BIDS stuff is project-based
XML, which integrates perfectly with CVS and SVN. The only drawback
is the DE schema, which only VST has change control for. I would
recommend having Microsofties on your team learn to use Linux and bash
scripting with CVS and SVN. They'll be better for it.

[quoted text, click to view]