December 14, 2009 at 1:22 am
Comments posted to this topic are about the item PowerPivot
December 14, 2009 at 2:25 am
Like you I have seen a demo of PowerPivot and was impressed by what it can do, but I find it a little scary. Its very easy to design a bad data warehouse and because of the design incorrect figures get churned out as a result, how much easier is it now going to be to produce bad results.
I hope that common sense rules when rolling this product out, but I've been in IT to long and doubt that will happen, I can see moderately capable users using this product believing that they are now Data Warehouse specialist and lots of bad statistics resulting.
December 14, 2009 at 4:44 am
I'm hoping it will work quite well for us as the biggest consumers for this information is the accounts department and you know how accountants love excel...
As they are quite sceptical about reports on data that they haven't written themselves, they tend to be rigorous in their scrutiny of the report cross checking against their own systems. You can't really complain with that level of testing and as the poster above says it will be easy to make a bad data warehouse hopefully this will lower the risk somewhat.
Plus it stops them using enormouse spreadsheets with vlookups all over them that take forever to run and use up vast amounts of disk space...
December 14, 2009 at 4:55 am
I think Carolyn has a point. It is something that long term developers and DBAs know only too well. MS Access anyone?
There is a place for user tools (I am no advocate of the locked down business environment). There also needs to be an understanding that somewhere down the line instances of ad hoc analysis and reporting will become well known, required to be shared and reportable in a common and repeatable manner.
Just like with MS Access apps, I can see that some PivotPoint work will remain ad hoc (and so it should) whilst other work will go through a natural progression to being an IT department maintained BI utility for the whole business (or part thereof).
As an industry we should be embracing the self-service attitude of businesses and be ready and willing to rationalise, commoditise and standardise those parts that graduate to that level of requirement. That also means we need to understand that we need to keep our hands off the rest.
Gaz
-- Stop your grinnin' and drop your linen...they're everywhere!!!
December 14, 2009 at 5:39 am
After seeing a demo of PowerPivot, I was reminded once again that no matter what "BI-for-the-masses" tool selected, business users won't benefit that much w/o a strong understanding of their own data as well as basic data relationships (cardinality, granularity, etc.). The ability to "mash-up" data from disparate sources is challenging enough for IT people with expertise. To expect that the garden variety end-user will easily be able to do this is a bit optimistic.
December 14, 2009 at 5:49 am
I think that when users reach the limit of their understanding that these so called data "mash-ups" will provide an excellent way of communicating understanding of the business to IT practitioners.
I also think each organisation will have to understand what level of support they will offer their PivotPoint users just like they do with Excel now.
Gaz
-- Stop your grinnin' and drop your linen...they're everywhere!!!
December 14, 2009 at 7:30 am
I've maintained for years that pivot tables in Excel were as close to BI-nirvana as humanity is likely to get.
This looks like a good addition to that toolset.
- Gus "GSquared", RSVP, OODA, MAP, NMVP, FAQ, SAT, SQL, DNA, RNA, UOI, IOU, AM, PM, AD, BC, BCE, USA, UN, CF, ROFL, LOL, ETC
Property of The Thread
"Nobody knows the age of the human race, but everyone agrees it's old enough to know better." - Anon
December 14, 2009 at 8:53 am
The ideas I've heard, and what has been built into Powerpivot with Sharepoint 2010 integration is a way to watch and see which applications are heavily used and they can be rewritten as a true SSAS application or enhanced.
Great points raised above, and I don't think this eliminates DWs as there is still a need for those large, clean sets of data. Without them, it is a tool to allow someone to make mistakes much, much quicker and with more confidence.
December 14, 2009 at 8:54 am
GSquared (12/14/2009)
I've maintained for years that pivot tables in Excel were as close to BI-nirvana as humanity is likely to get.
Actually, Gus, I don't think that I could disagree more. Now if you were to move the words "in Excel" from the middle of that sentence to the end, then I would be in sad agreement (though with the sincere hope that PowerPivot will change that forever).
The thing is, even when Microsoft added Pivot Tables to Excel some 15 years ago, it was already obsolete BI-UI technology by at least 5 years. You see the height of End-User BI-nirvana was actually just starting then in a set of products going by the terms "Executive Information Systems" or "Decision Support Systems" or just DSS. This software tried to marry the easy end-user customization and interface of spreadsheets with the high-power analysis and data integration of that time's data analysis tools like RS/1, SAS and Minitab. Although it did have some success, it's hugely complex hardware, software, enterprise integration and the need for extensive customization to adapt it to each different corporation kept its cost in the stratosphere and thus the need to market and sell it only to executive teams. This also made it essentially a background player for years to come.
However, two very important things did come out of it. The first and most obvious is the whole field of Data Warehousing (DW) and Business Intelligence (BI). The second is that *some* companies at that time were able to successfully adapt the analytical techniques (though generally not the data integration) to reasonably priced end-user tools. These were called Multi-Dimensional Analysis tools (MDAs). Several of them even went the one additional step of adapting them too the most sucessful end-user analysis metaphor of all time, the spreadsheet.
The one that I am most familiar with is a wonderful product that was called Lotus Improv. Although it only ever got to version one (AFAIK), it is still my favorite spreadsheet product, ever. It's biggest shortcomings were that it was not integrated into Lotus 1-2-3 and that it did not have full spreadsheet capabilities. (Again, it also lacked real data integration) But what you colud do with the data that you got into was a dream for end-users, BI-nirvana, truly.
Now what was obvious at the time was that Lotus intended to integrate these features and capabilities into Lotus 123, either as and add-on in the next edition of Improv or directly in the next release of 1-2-3. And IMHO that would have forced MS to add similar capabilities to Excel. Unfortunately this was when Lotus started its implosion, and Improv was lost to history
And instead of any real MDA tools in Excel, we got Pivot Tables, whose only relationship to a real MDA tool is it's appearance, it has virtually none of the real features of an MDA. Comparing Pivot Tables to an MDA would be like taking Excel, removing every feature, formula and function except the Aggreagates (Sum, Avg, etc..) and cell formatting and calling *that* a spreadsheet.
[font="Times New Roman"]-- RBarryYoung[/font], [font="Times New Roman"] (302)375-0451[/font] blog: MovingSQL.com, Twitter: @RBarryYoung[font="Arial Black"]
Proactive Performance Solutions, Inc. [/font][font="Verdana"] "Performance is our middle name."[/font]
December 14, 2009 at 9:42 am
I see this as exciting and very valuable. Any advancement into BI is great for business, Data Quality, and End Users. And any way to make it easier, faster, or better is excellent.
I could say that they are still out of date, but would rather celebrate that they are doing something about it.
i could also say that this will make the bad DW more available to the user and expose more problems with the data. But I choose to say that this will allow those building the DW and tending the data more rapid and better access to the data for better data mining and cleaning. Thus the better and easier the tool the easier it should be to raise the data quality. By the way we do not know we have bad data till we find it and if this new toolset helps us find it sooner, this is wonderful.
In the past I have built small BI solutions that do not see the light of day as far as the end user is concerned but have caused the quality of the data to rise considerably.
The new tools will make the job of the data professional better if they dig in and use them. Good step forward, and using Excel allows us to use a toolset that we know already and build on the success of a commonly known tool. It is a good idea.
Have a great day!
Not all gray hairs are Dinosaurs!
December 14, 2009 at 2:43 pm
Barry:
I've spent more years as a manager/exec than as a DBA (though that changes next year, when I will catch up with myself, as it were).
As such, I've seen the high end and the low end of data analysis, and a lot of the in-between.
Two observations I've had over the years, first as a manager, then as a DBA:
First, any sufficiently complex system will be useless in the hands of non-expert users.
Second, most managers specialize in managing, not in data analysis.
Having high-end, comprehensive, powerful data analysis tools available for people who are trained to not only use them, but to understand the importance of things like statistical deviation, the differences between average and mean, data granularity, the importance of accuracy and precision and the relationship between the two, is a powerful thing.
But going that far with the usual manager is just going to create chaos.
- Gus "GSquared", RSVP, OODA, MAP, NMVP, FAQ, SAT, SQL, DNA, RNA, UOI, IOU, AM, PM, AD, BC, BCE, USA, UN, CF, ROFL, LOL, ETC
Property of The Thread
"Nobody knows the age of the human race, but everyone agrees it's old enough to know better." - Anon
December 14, 2009 at 7:45 pm
I'm hoping it will work quite well for us as the biggest consumers for this information is the accounts department and you know how accountants love excel...
I am not sure PowerPivot is for accountants because it is not much different from pre Oracle Hyperion used to connect SQL Server 2000 relational database by bank MBAs to do complex analysis and aggregation.
In the interim Microsoft is telling all developers PowerPivot will make all including the T-SQL and everything data challenged BI experts. Microsoft joined others in this market place so the BI experts will be more productive and some MBA will be empowered.
😉
:Whistling:
Kind regards,
Gift Peddie
December 14, 2009 at 11:00 pm
I was excited enough to get myself signed up for the Office CTP so i could try it out...and i am less than impressed. We were hoping that Excel 2010 and Gemini would offer a more comprehensive range of options for querying our cube, but as far as i can tell, there are few improvements to the normal pivot table functionality without importing the data from SSAS into Power Pivot and querying it from there...and i can just see all the Power Pivot workbooks springing up all over the place that aren't kept updated and people reporting on old data...and then it coming back to the BI team to explain over and over what the problem is!
December 15, 2009 at 1:08 pm
PowerPivot and its cohort Excel are generally a 'bad idea.' I have worked for three companies deploying data warehousing and every time I see someone doing analysis in Excel I cringe...why?
1. Bad data - I can almost guarantee that a power DW + Excel user has their own 'private' copy of data. Where did that come from? How do we know it is aligned with the corporate data? If the end user finds that the data is bad they typically correct it in their spreadsheet but seldom does that get propagated back to the source.
2. Private rules and processes - I am constantly battling this one, if something has been done in Excel then what happens when that person goes on vacation? Who else can run those quirky processes? Who has backed it up.
3. My Private Island - Excel (in many forms) leads to more and more islands of information. This is the opposite of good BI. Good BI makes information more widely accessible.
Maybe the capability of PowerPivot is a great metaphor but implementing it in Excel just guarantees more problems down the road.
do it right, or do it over and over, it's up to you
December 15, 2009 at 1:28 pm
Richard,
Excellent points, and some of that is trying to be addressed with Sharepoint integration. I'd agree there are issues, but that's not a reason to ignore Excel, especially when a "real" application can take a long time to build. Instead I'd agree with you and like to see a way to lock excel down to using data from a query and not saving off private copies.
Viewing 15 posts - 1 through 15 (of 21 total)
You must be logged in to reply to this topic. Login to reply