When you build a cube in analysis services it is very easy to expose every attribute in every dimension to the user. You can also add-in every measure and if you don't feel that's enough then you can create your own calculated members to add to the users' fun.
Once you have given the users every thing they asked for you can now sit back and relax just like on the IT crowd.
I would say that at this stage the work is less than half done. A raw cube is like your digital music collection before you've tagged it. It's all in there but you can't find anything. What your users need even, if they don't tell you, is some direction and guidance to make sense of the monster you have created.
There are several things you can do in the cube to help with this:
The other really big thing you can do to help your users is document what you have created. This needs ot be done in two ways:
To conclude business intelligence is there to make sense of the data in an organisation and so writing coool MDX and buying a stonking 16 way server is the easy bit. The hard bit is encouraging them to use it and to get the best out of it.
Yes! Please share ideas for end-user documentation. We are strugging as one of our cubes has hundreds of attributes and metrics and it can be overwhelming. We use perspectives and folders to organize data, but that we need good old-fashioned documentation too. Definitely interested in seeing other ideas. Thanks!
It will be my pleasure but probably not unilt next week. I will dig out the stuff and try and get something working. AF
Does your tool allow cube data sets to be combined?
Not sure what you mean?
Cubes can be built from multiple data source (though keys need to match).
You are building one cube with all your data in (in separate measure groups as described above), so I don't need to do any combining.
Finally Micrsoft front end tools e.g. Excel and Reporting Services would allow you to combine data on one report or sheet as needed.
I fyou need more please e-mail me (firstname.lastname@example.org)