July 21, 2010 at 9:57 pm
Comments posted to this topic are about the item Self-Service vs. Managed BI
July 21, 2010 at 10:42 pm
I think your formula one car vs house plan analogy is spot on. Sadly, it seems that the only people who get this are the ones doing the development!
July 22, 2010 at 1:46 am
Oh TOTALLY agree to both of you!
Our company has recently been taken overy by another company, creating a huge company. We're now doing the BI for it, and it's giving us a chance for a bit of both - planning the "new" solution and building to plan, whilst using the old version for learnings.
That analogy, as Kenneth said, is spot on.
July 22, 2010 at 2:17 am
It's not like a house where everything is planned from the start, but more like an evolving Formula-1 car that is tweaked and tuned with the user inside of it.
I like that methaphore!
However as you metioned yourself: building (great) software is hard. Why put the reponsibility at the (Powerpivot) users desk? It's hard enough for users to analyze conformed data in a (traditional) Datawarehouse let alone build a powerpivot solution.
Dirk
July 22, 2010 at 2:21 am
From my experience Dirk, the users genuinely dont know what they want.
Giving them power pivot allows them to play with the data until they get what they want. We can then build what they want.
What tends to happen now is that they'll ask for everything.
We'll then have to say no and ask them what they actually need.
They'll have a guess (it will be wrong), we'll build what they ask for and then have to go through several itterations of that.
Frustrating for both parties.
July 22, 2010 at 3:05 am
I can see where you're coming from and if you put it this way it makes sense to develop a BI solution by that procedure. However I still think that most business users don't have enough knowledge of Excel, PowerPivot and Database Concepts in general to be able to build a prototype. And then again even if they could, if all users build a prototype using different assumptions and definitions wouldn't it be really hard to for the BI Project Team to build a solution with all the functionality in those prototypes?
July 22, 2010 at 3:18 am
Oh yea - really hard, but a lot easier than months and months of itterations.
I think we're in a fairly good position in that most of our users are pretty switched on when it comes to data, and if they're not they get a business analyst in to do the digging for them.
July 22, 2010 at 6:13 am
Sometimes simple is all that's required. We DBA and developer folks tend to over-develop. We want to give the suers the world when all they really need is one small corner of it. Maybe Powerpivot is that one small corner.
July 22, 2010 at 6:20 am
There is definitely something to say for user managed reporting. I don't know how many 'great' projects I have created that are now gathering dust, because nobody is using it, whereas the data-based spreadsheet or quick-and-dirty report that fills a need is still in use.
The key seems to be to stay highly involved with your user, and begin with something basic. Grow and polish it as needed.
July 22, 2010 at 6:24 am
I think the selling point for the simple Excel and Access files is the ability to easily manipulate them.
People like to put their own mark on things.
July 22, 2010 at 6:49 am
Too often we don't build the best software the first time. It's certainly not all our fault since there are plenty of business people that "think" they know what they want, but they really don't until they see something.
Oh so true. And don't knock it .. heck it has kept me in a Developers job lo these many years.
July 22, 2010 at 6:55 am
However I still think that most business users don't have enough knowledge of Excel, PowerPivot and Database Concepts in general to be able to build a prototype.
++
"Data? You can't handle the Data!" 😛
July 22, 2010 at 7:00 am
It's not always just putting their mark on it. It's speed.
Sometimes as a result of an issue that just popped up they need immediate answers on, sometimes it's that speed of "development" means they can economically work through multiple iterations of views of the data before finally settling on one.
If someone comes to me with a problem, I don't jump into T-SQL and start writing stored procedures. I jump into Access, work up a model of what I think they want, then work with them on refiniing it. Once it's honed we can "formalize" it if needed.
My experience is that if they have access to the tools, about 5% of users will become experts with the tools, and other users will turn to these inhouse experts for a lot of the "quick and dirty" stuff. This can be a great way of freeing up IT resources for the larger projects.
Unfortunately IT (generically) is frequently viewed as an ivory tower unable or unwilling to address user needs in a timely manner, and likely to come up with a solution that is way overblown when they do.
Allowing users to model up their own solutions and working with them whenever possible helps overcome both problems.
IT should focus on helping the critical 5% with training and mentoring whenever possible, and encourage users to bring
the solutions they have come up with to IT for review and possible conversion to a "formal" solution for the organization.
Perhaps a quarterly or annual "show us what you've done" meeting with a simple ground rule that IT not comment on how it was developed or what it was developed in, just focus on what the organizational problem was and how this solution helped solve that problem.
Doing this would benefit IT, the users, and the greater organization as a whole.
July 22, 2010 at 7:02 am
Ben Moorhouse (7/22/2010)
I think the selling point for the simple Excel and Access files is the ability to easily manipulate them.People like to put their own mark on things.
True - but there can be hidden dangers in users manipulating data, besides the time involved.
We had removing data that was 'bad' to create a cycle time measurement.
They left, and one day the person who took it over asked me 'how come I always have to remove this data points to create the cycle time measurement, can we fix it?'.
My first question was Work Order or Sales Order cycle time?
They found out that this 'creative' person had started with the SO Cycle Time detail, removing some data points, and presenting it as WO Cycle time. So for almost 2 years, they had been presentling progress towards a KPI using the wrong data. They found out that our rules and the data for WO Cycle Time met the requirements and the data was cleaning measuring what they wanted to see.
A mix of self service and managed BI is a good compromise - users can get creative and explore, allowing them to prototype. But when views expand beyond a few people, it's time to have a review. Is it something that should be published for all levels and roles in the company? Is it based of an existing report that maybe should be modified? Was this the result of them removing some data or creating a new dimension?
I totally agree with Steve's statement - they usually don't really know what they need until the see it.
Greg E
July 22, 2010 at 7:03 am
I think the selling point for the simple Excel and Access files is the ability to easily manipulate them.
yeah, but how can those reports be trusted?! you don't want (business) users manipulating gross profit figures for your company...
Viewing 15 posts - 1 through 15 (of 38 total)
You must be logged in to reply to this topic. Login to reply