The BI Update
-
If you've read these ramblings for any period of time, you realize that I'm not much of a BI supporter. It's not that I don't think there's value in building BI solutions or that no companies are implementing them. It's that I think many of the Analysis Services components are very difficult for most DBAs to implement. Or even understand. I'm not convinced that even end users always get the data they're looking at, but that's another topic.
However I'm up in Seattle for the first Microsoft B Conference to see if maybe I'm reading things wrong. I know that I've been reading The Data Warehouse Lifecycle Toolkit to get myself prepared for the conference and I've learned a few things I didn't know.
Like the dimensional modeling seems less complex than ER modeling. Not to say it's easy, but it has less parts than an ER system, at least from what I've gleaned from Kimball's book, and it's a little more natural view of the world. Not that I think it's easy to implement, but at least the warehousing might not be as hard as the Analysis Services portion.
I've also come to start thinking of a warehouse as a completely separate thing from an Analysis Services implementation. I think they are both still complex and the ability to drive your organization to agree on single definitions of data elements and then get it cleaned is a large task. One that I still think that most companies have trouble spending the time and effort on implementing.
I think this area is growing, but not nearly at the speed of the hype. Too many companies still haven't even begun to implement anything and may not for a long time to come. That's a double edged sword for BI experts. Less jobs and choices on where to work, but also more money because the supply of you is still pretty thin.
I'm up in Seattle right now, probably listening to Jeff Raikes keynote as you read this. I'm here as more of a skeptical reporter than a geek trying to learn, but I aim to do both this week and try to get you some reports and thoughts on the conference and the state of BI in the Microsoft world. So keep tuned for updates throughout the week.
Follow me on Twitter: http://www.twitter.com/way0utwest
Forum Etiquette: How to post data/code on a forum to get the best help
My Blog: www.voiceofthedba.com -
My co-workers went to the conference too. They said they only allowed 1500 attendees and they just registered last week. I did not think it was very popular conference.
-
Hiya Steve,
I know you're not a fan of Analysis Services and that's absolutely fine - in a lot of ways I'm right there with you. What irks me is that you seem to talk about BI and Analysis Services in the same breath interchangably as if one is the same as the other. They're not. Analysis Services is only a very VERY small subset of what BI actually is.
At the very high level BI is about getting value from the data that you possess - simple as that. It can incorporate many things including a data warehouse, a multidimensional data mart server, an ETL tool, a data mining tool, web services, data services, a kimball data model, an Inmon data model, a Tom Dick and Harry data model. Or...it might include none of those. Its important to make a distinction between BI as a discipline and the tools that one can use to accomplish it. I have to be honest - I don't think you do that.
Digressing slightly....Having implemented many Analysis Services systems in my time I would absolutely agree with your assertion that "I'm not convinced that even end users always get the data they're looking at". In my experience that's (generally speaking) the fundamental problem with Analysis Services - end users don't get it. The true benefit of Analysis Services isn't the wonders that you can achieve with MDX - its the aggregation engine which gives the huge speed increases. Too often we (and I firmly include myself in that) forget the KISS principle when building Analysis Services solutions - Keep It Simple Stupid.
Say hello to Mr Knight for me. I'm sure he's there!!!
-Jamie
Jamie Thomson
http://sqlblog.com/blogs/jamie_thomson -
I started doing Business Intelligence work before it was called "Business Intelligence" - back when years started with "1" and we had to carve our own ICs out of wood. Ok I'm exagerrating a little (but not much).
Back then this was called Decision Support Systems (DSSs). In manufacturing, we called them Manufacturing Execution Systems (MESs). Even before that they were called Data Acquisition and Reporting Systems (ok I'm done with the TLAs).
There's always been a market for looking at data in some format other than columns and rows. In my opinion that's what Business Intelligence truly is - another way to view data.
I agree with Steve - the hype gets out of hand at times - but doesn't it always whenever there is money to be made and a few non- to moderately-technical sales people involved?
Just my $0.02...
:{> Andy
Andy Leonard, Chief Data Engineer, Enterprise Data & Analytics
-
I do dev work at a data analysis company which reports on sales data for manufacturers. What always surprises me is not that there are a staggering number of permutations in the ways in which you can slice the data to discover non-obvious relationships and interesting BI, but more that the presentations to the clients usually end up having to be so dumbed-down.
I know that there are a lot of smart guys in business, and I know that we in the rarified air (*smirk*) of IT are often viewing information with several years of specialised experience and training, but it is frustrating to see an insightful and potentially valuable piece of information get ignored because there isn't a simple enough way to spin it to the clients without losing them in the technical details required to get there.
As I see it, that is the greatest challenge for BI as a commercially viable area of IT - not finding the intelligence, but finding ways to to present it which neither trivialise it (think gimmicky displays) nor result in eyes glazing over...
-----------------
C8H10N4O2
-
I'm not into BI, but then again after reading these responses, maybe I am a little:-)
Anyway, thanks to all for your insight, and to Steve for bringing the subject to the forum.
-
i certainly would not describe business intelligence (either part of the 2 pieces you decided to break it into) as harder to understand than 90% of the stuff that hits this site.
at it's most basic it is straight forward and common sensical.
-
Thanks for the comments and it's definitely a large, complex area. I think Jamie brings up a great point that there's more to BI than SSAS. Digging through my Kimball book on the plane, I'll back off on the warehouseing side. A data warehouse and all the ETL work that can go into it, building a place where the corporate data can live and be reported on doesn't seem like nearly the complex project that an SSAS one is.
I definitely think the media hype on BI, the marketing by companies, has sort of blurred the line of what BI is. But that's some of why I'm here this week. To learn more about it, see what projects have worked and not worked, see what people have done and try to report back on some of it.
Definitely look for some articles coming ni the near future.
Follow me on Twitter: http://www.twitter.com/way0utwest
Forum Etiquette: How to post data/code on a forum to get the best help
My Blog: www.voiceofthedba.com -
Steve
I want to know how's the break up sessions of the conference. Do they provide adequate information? When I looked at the agenda, 2 hours of break up session talking about SSIS and BI can't even touch the surface of the subject.
-
Agreed. I'd like to see lots of comments from conference attendees coming out this week.
-Jamie
Jamie Thomson
http://sqlblog.com/blogs/jamie_thomson -
Making notes and I should have some writeups tomorrow and over the next week. Trying to hit lots of sessions
Follow me on Twitter: http://www.twitter.com/way0utwest
Forum Etiquette: How to post data/code on a forum to get the best help
My Blog: www.voiceofthedba.com
Viewing 11 posts - 1 through 10 (of 10 total)
You must be logged in to reply to this topic. Login to reply