Hello All

I am using Analysis services to build some cubes for
planning/budgetting and for what-if analysis.

I am bit at cross roads and do not know how to model the cubes for
planning/budgetting & what-if analysis I was told by somebody to use a
dimension for version, & scenario to model these kind of cubes.

All I want to do is sales planning I would greatly appreciate if any
of you can kindly share with me ideas to model cubes for
planning/budgetting and for what-if analysis. Please let me know the
dimensions I need for these kind of applications to enable me to
writeback.

ALso, is there some web site where I can find some cube data models
for some of the common planning/budgetting applications like
- sales planning
- Balance Sheet Planning
- Investment Planning
- Profit planning

Please advice where I can find some good data models for the above
scenarios.

Thanks
Simon

Simon

Please feel free to contact us for a live demonstration on how to achieve
this and how to automate this process. We will also show you how to tie this
in to our general ledger applications.

Howard.Taylor @domain below
www.o2olap.com

"Simon Lenn" <simonlenn (AT) yahoo (DOT) com> wrote

Simon,

Do you already have the cubes for the actual sales figures?

Usually, the planning is done at a higher grain; i.e. not as much
detail as actual sales.

In such cases it is best to have two cubes - one for actuals & one for
planning.

The planning cube will have a version/scenario cube.

Using parent-child dimensions for the others will let you writeback at
a non-leaf level.

The planning fact table will initially be empty.
To process the cube for the first time you will have to 'con' Analysis
Manager into thinking there is 1 line of data in the fact table by
changing the fact count field in cube editor.

Remeber when you are writing back the numbers are not going to this
fact table but to a writeback table. You need to periodically move
these numbers to the fact table.

You then need a good front-end that can bring numbers from both the
planning & actual cubes together to do the variance analysis.
My favourite is the Excel add-in from XLCubed (also web version
available).
Others like Howard's O2OLAP can also do this.

If you are doing a serious amount of modelling & planning TM1 has the
advantage over AS.

What would be nice is to have a front-end that can bring the numbers
from AS & TM1 together seemlessly!

Regards,

John Keeley

www.johnkeeley.com




simonlenn (AT) yahoo (DOT) com (Simon Lenn) wrote in message news:<3641e2c2.0408270345.2e1d771f (AT) posting (DOT) google.com>...