Customer Profitability

Customer Profitability

Hi How can I use MS SQL server 2005 analysis services for measuring the Customer Profitability Index

Pl suggest if you have any Data model and any special algorithm that can be used to measure the Customer Profitability

Thanks In Advance

Rajesh Ladda

Analysis Services is a product that serves as a platform for your analytical applications. Using Analysis Services you are able to model complex business rules of your organization.

Your question is a bit vague. If you are looking for examples of how to build multidimensional model and then use it to analyze your data, I suggest you first start from going through tutorial that comes with Analysis Services installation. Then take a look at the sample project: AdventureWorks and see how it implements calculations.

Edward.
--
This posting is provided "AS IS" with no warranties, and confers no rights.

Custer an index or not?

I see this as potential problem. I have Indexed columns for a Clustered
Index as follows:
Merchant_Account_ID
DateOfReport
Standard_Entry_Class_Code
This would repage my index on the daily insert of say 300+ returns where
they are going in by Merchant + Data+ Code.
I am finding a few of these, not as many as I thought :)
I want to remove the clustered aspect of the index, and still leave it as
unique. This would increase my insert performance wouldn't it?
TIA> I want to remove the clustered aspect of the index, and still leave it as unique. This would
> increase my insert performance wouldn't it?
Likely. But insert performance will be even better if you can cluster on something which it
monotonically increasing (instead of finding free space using PFS and IAM pages, SQL Server just
navigates the index, find the "end of the table" and there the row goes). Perhaps DateOFReport is,
so you can specify this as the first column in the index?
--
Tibor Karaszi, SQL Server MVP
http://www.karaszi.com/sqlserver/default.asp
http://www.solidqualitylearning.com/
Blog: http://solidqualitylearning.com/blogs/tibor/
"_Stephen" <srussell@.electracash.com> wrote in message news:uaNuXscTGHA.5856@.TK2MSFTNGP11.phx.gbl...
>I see this as potential problem. I have Indexed columns for a Clustered Index as follows:
> Merchant_Account_ID
> DateOfReport
> Standard_Entry_Class_Code
>
> This would repage my index on the daily insert of say 300+ returns where they are going in by
> Merchant + Data+ Code.
> I am finding a few of these, not as many as I thought :)
> I want to remove the clustered aspect of the index, and still leave it as unique. This would
> increase my insert performance wouldn't it?
> TIA
>|||yes, removing clustered indexes always increases speed for inserts, and
usually speeds up updates for the key fields.

Custer an index or not?

I see this as potential problem. I have Indexed columns for a Clustered
Index as follows:
Merchant_Account_ID
DateOfReport
Standard_Entry_Class_Code
This would repage my index on the daily insert of say 300+ returns where
they are going in by Merchant + Data+ Code.
I am finding a few of these, not as many as I thought
I want to remove the clustered aspect of the index, and still leave it as
unique. This would increase my insert performance wouldn't it?
TIA
> I want to remove the clustered aspect of the index, and still leave it as unique. This would
> increase my insert performance wouldn't it?
Likely. But insert performance will be even better if you can cluster on something which it
monotonically increasing (instead of finding free space using PFS and IAM pages, SQL Server just
navigates the index, find the "end of the table" and there the row goes). Perhaps DateOFReport is,
so you can specify this as the first column in the index?
Tibor Karaszi, SQL Server MVP
http://www.karaszi.com/sqlserver/default.asp
http://www.solidqualitylearning.com/
Blog: http://solidqualitylearning.com/blogs/tibor/
"_Stephen" <srussell@.electracash.com> wrote in message news:uaNuXscTGHA.5856@.TK2MSFTNGP11.phx.gbl...
>I see this as potential problem. I have Indexed columns for a Clustered Index as follows:
> Merchant_Account_ID
> DateOfReport
> Standard_Entry_Class_Code
>
> This would repage my index on the daily insert of say 300+ returns where they are going in by
> Merchant + Data+ Code.
> I am finding a few of these, not as many as I thought
> I want to remove the clustered aspect of the index, and still leave it as unique. This would
> increase my insert performance wouldn't it?
> TIA
>
|||yes, removing clustered indexes always increases speed for inserts, and
usually speeds up updates for the key fields.

Custer an index or not?

I see this as potential problem. I have Indexed columns for a Clustered
Index as follows:
Merchant_Account_ID
DateOfReport
Standard_Entry_Class_Code
This would repage my index on the daily insert of say 300+ returns where
they are going in by Merchant + Data+ Code.
I am finding a few of these, not as many as I thought
I want to remove the clustered aspect of the index, and still leave it as
unique. This would increase my insert performance wouldn't it?
TIA> I want to remove the clustered aspect of the index, and still leave it as unique. This wo
uld
> increase my insert performance wouldn't it?
Likely. But insert performance will be even better if you can cluster on som
ething which it
monotonically increasing (instead of finding free space using PFS and IAM pa
ges, SQL Server just
navigates the index, find the "end of the table" and there the row goes). Pe
rhaps DateOFReport is,
so you can specify this as the first column in the index?
Tibor Karaszi, SQL Server MVP
http://www.karaszi.com/sqlserver/default.asp
http://www.solidqualitylearning.com/
Blog: http://solidqualitylearning.com/blogs/tibor/
"_Stephen" <srussell@.electracash.com> wrote in message news:uaNuXscTGHA.5856@.TK2MSFTNGP11.ph
x.gbl...
>I see this as potential problem. I have Indexed columns for a Clustered In
dex as follows:
> Merchant_Account_ID
> DateOfReport
> Standard_Entry_Class_Code
>
> This would repage my index on the daily insert of say 300+ returns where t
hey are going in by
> Merchant + Data+ Code.
> I am finding a few of these, not as many as I thought
> I want to remove the clustered aspect of the index, and still leave it as
unique. This would
> increase my insert performance wouldn't it?
> TIA
>|||yes, removing clustered indexes always increases speed for inserts, and
usually speeds up updates for the key fields.