The BI Year
-
Maybe I'm wrong.
It wouldn't be the first time and certainly not the last. I've been skeptical of the whole BI thing, but this report on SAP earnings shows triple digit growth in their BI space. Add to that Microsoft making a charge in this area and maybe there's something here.
I've written on this before and I've been skeptical on the extent to which BI will become a mainstream technology. I backtracked a little yesterday since I think that BI might be poised to explode. I'm not sure about 2007, but I'm thinking 2008 is the year. A new release of SQL Server is likely, we'll have had 2 years of this paradigm of SQL Server development, Reporting Services will be widely deployed, and many of the new tools from Microsoft and others in the BI arena will have been available for a year or so at a reasonable price.
In the past I've thought there were two things preventing much of BI from becoming widely deployed. First was the price. Most BI sales were 6 figure deals, which was great for the Cognos and Business Objects of the world, but it meant a limited number of companies and people would get to use them.
The other thing was the whole cube/MDX model. It was cumbersome to build and work with, required offline builds, and was very complicated. I think most DBAs couldn't understand it, so the scope of projects was limited and not many people wanted to deploy the technology.
I see lots of that changing as more and more people start to write articles, the tools become easier to use and the cost goes down.
Maybe it will even get to the point where an old DBA like me can understand it 🙂
-
I really doubt that there would be an explosion of any kind. Based on my days as a part time DBA/System Support, most clients are just not able to understand what cubes and DW is all about.
They will just come around and say, hey, I want this and that but I do not want to pay the amount that's asked for by MS/ORACLE etc.
We normally get around it by giving them the data in MS-EXCEL and ask them to design their own pivot tables so it will definitely take sometime before the general marketplace accepts BI as a component of everyday business computing.
-
I think we might be failing by asking them to get "it". They get the Excel metaphor and I say use that until they realize they've outgrown it, then they'll be ready for something more complex. I think the biggest driver of BI is performance issues on reports and dashboards. Cubes and other techniques make reports run a lot faster, sometimes enough to make running them possible vs not. Dashboards for all the gee whiz buzzword feel have real value to business leaders and again, BI approaches under the hood seem to help those (not that you cant get there on an OLTP system).
And as for "Maybe it will even get to the point where an old DBA like me can understand it
"...so many smart comments to make, none worthy of typing! -
We are witnessing the commoditization of BI. The "explosion" is not an "if," but a "when" - and that's not including the heavy napalm that's rolling in as we converse ...
Check out my articles for hands-on introductions in how to put this stuff to work in the real business environment ...
Thanks for providing an excellent support and learning site!
Bill
William E. Pearson III
CPA, CMA, CIA, MCSE, MCDBA
Island Technologies Inc.
931 Monroe Drive
Suite 102-321
Atlanta, GA 30308
404.872.5972 Office
404.873.3128 FAX
404.277.3104 Mobile
http://www.islandtechnologies.com
----- ----------- ----------- --------------- ------------
Publisher Sites:
http://www.databasejournal.com/article.php/1459531
http://www.sql-server-performance.com/bill_pearson.asp
http://www.informit.com/authors/bio.asp?a=862acd62-4662-49ae-879d-541c8b4d656f
-
The available tools are certainly mature enough now for BI to explode. The problem is that the amount of resources required to carry out a BI project properly has not changed and is probably beyond the capability of most SMBs. The amount of requirements gathering, design, development, implementation, and training effort required to complete a BI project is substantial. Our medium sized business has been trying for 2 years to no avail as changes to our operational systems have taken priority and resources away from our BI project. We have also, in the process, discovered that our OLTP system currently does not capture information which managers want for their key business metrics and that our operational systems will have to be modified to capture this data before a meaningful BI project can proceed.
-
I'm noticing lots of vendors now throwing in a BI solution to their package. So I agree it is on the rise. But I am not seeing BI used as initially thought. I rarely see managers, directors ... using the cubes. Instead we hire financial analysts that prep reports for the managers and directors. I just don't think most people understand data, let alone how to use cubes. It all leaves me wondering if the cube concept is really worthwhile. It feels more like we prep and clean data and put it somewhere so people can query it easier. Rarely do the cubes contain all the facets that are needed, so the analysts have to build something new anyway. I still believe in the principles of DW, but the intended audience is not there and building pretty analysis tools appears worthless.
-
Yes I agreed the BI tools are getting earier to use and more robust. I am a data warehouse developer, I got calls from vendor everyday. Most of them can tell you that it takes less than a week to implement the system. Yes it takes less than a week to install the tool, but who is going to put the data in?
The tool is the car, no matter how good it is, the driver is the important person. No matter which database you use, SQL Server, Oracle, Essbase...., and which tool you use, SAS, Hyperion, Cognos, Business Object. You still need DBA, data architect, database developers to build the data warehouse.
my 2 cents.
-
In my experience, the limiting factor is the target audience. If they are not ready for the technology it will be a waste. The company I worked for had gross revenues over $3 billion and only one accountant made use of the cube we built. More use was made of a subsequently developed operational data store. The benefit of the ODS was bringing data from data silos into one database so that all business units' data could be compared at a corporate level. In the end it's those pesky business types that will determine if BI survives or thrives.
-
BI isn't something you just do for a year, it's a whole lifestyle change...
-
Steve, you going to comment on the lifestyle change?
-
I think I'll pass
-
You can take it how you want. Which is exactly how I meant it
Some companies do adopt Business Intelligence as a lifestyle... My point is that you can't just haphazardly undertake anything in the whole field. It takes a good year or more to get a good BI project off the ground, with careful planning, and the company itself has to be signed-on. The one project I worked on years ago would have been really good, but the only people who cared about it were the programmers. In order to be effective, we needed sign-on by everyone from the CEO to the mailroom. Without that, our efforts to measure processes, warehouse the data, and so on, would have failed miserably.
I think companies are realising new methods of managing the situation, and the technology is maturing, so it seems reasonable for more companies to put BI projects into the works.
-
"technology is maturing"
You have no idea! Everyone seems to have forgotten (or didn't know) that there is language out there that has been using multidimensional arrays for over 20 years that I know of and probably longer.
I worked as a COBOL programmer for a large insurance company in the early 80s and we developed and reported on "cubes" regularly
MG
"There are two ways of constructing a software design. One way is to make it so simple that there are obviously no deficiencies. And the other way is to make it so complicated that there are no obvious deficiencies."
Tony Hoare"If you think it's expensive to hire a professional to do the job, wait until you hire an amateur." Red Adair.
-
But what makes you think the explosion isn't already happening? Actually, I think I can answer that question by picking up on another thing you said, which was that "I think most DBAs couldn't understand it". The tools in the Microsoft BI stack, certainly Analysis Services, are as complicated as any other server product and asking someone who's already an expert on the SQL Server relational engine to be an expert on Analysis Services, Integration Services and Reporting Services as well is an insurmountable challenge: you have to specialise in one or the other. Anyone who wasn't a DBA could say that relational databases were "very complicated". So my point is that in my experience it's not the DBAs who are implementing BI solutions, it's specialised BI professionals, and that it's a distinct job rather than something a DBA can do in his or her spare time.
This probably explains why my experience of BI seems to be so different from almost everyone else who has commented here: perhaps because DBAs aren't much involved in BI projects they assume they're not happening! I've been working with Analysis Services and before that OLAP Services since the latter was in beta, and since the release of SQL2005 I've seen an immense upsurge in the amount of BI projects going on - certainly enough to merit the description of BI as a "mainstream technology". And as you say, the financial results of companies like SAP and the presence of Microsoft in the market seem to back this up.
-
I agree with Loner because those tools have allowed all types of Advanced developers and Business Analyst to do BI, if one is lucky to work in places with very complex solutions it is not that complicated to add the needed skills. I remember telling a developer about UDM in SQL Server 2005 in beta 3 and he told me he does UDM with Hyperion it is a C# developer who moved to C# from C++, he just uses SQL nothing more. The odd skill in the middle is Reporting when it is both on the Algebra and Calculus end then a hybrid developer is needed or some of the Reports will have performance issues. It is not something a DBA should do part time but I have seen teams that combine both ends in the Data teams in big companies.
And a side note people on the Algebra end of the Relational model can choose to do BI because Joe Celko have proved you can think in Algebra and later move on to the Calculus end with his recent OLAP book remember E. F. Codd defined both ends of the relational model. The Calculus end was later implemented by Bill Inmon and Ralph Kimball. The rest is minor details. Now I am not saying it will be easy to be Expert of both ends but one can be expert on one end and use the other as needed.
Now to Multidimensional Arrays as cubes not in the Microsoft platform because those cannot be sorted in the FCL(framework class library), Microsoft in packaged software that includes BI like Project Server runs Cubes as .sql files.
Kind regards,
Gift Peddie
Viewing 15 posts - 1 through 15 (of 15 total)
You must be logged in to reply to this topic. Login to reply