Friday, 20 June 2008
Will the credit crunch hurt?
Will the international credit crisis affect the business intelligence market in a big way?
Sunday, 25 May 2008
New branding live!
At last, after many discussions and trials, we have launched our new colours, logo and website. Thanks to my team and input from our graphics partners for the hard work. I hope that the new branding goes down well with our prospects!
Tuesday, 8 April 2008
Rebranding
Hi, I noticed that I hadn't been on the blog for a while. Just to let you know that we are going through a rebranding exercise at the moment, so look out for the new colours soon!
Monday, 20 August 2007
KPI Industry Specific
Increasingly we are being asked to help out and suggest KPI's for specific industries. We are building up a short list of what we think people should be measuring at the very least in various industries and they will be available on our website shortly http://www.consolidata.co.uk
Saturday, 4 August 2007
OLAP KPIs
Is SQL Server 2005's Analysis Services a good tool to use for reporting and KPI reporting in particular? In short - yes, but it does require a fair amount of skills to achieve this. I have found that you really need to understand how the cubes work properly before attempting to add in the more exotic KPI measurements.
Where the cubes really come into their own (as opposed to stored procedures running of relational tables) is in the MDX scripted calculations, in particular I have found the 'periods to date' functions extremely useful. Previously we would have had to precalculate these measures, or in some cases do it on the run, through a SP with mountains of T-SQL code. Now we simply have the short MDX expression on the cube - even better is that it allows any period calculation ( Week to date, Month to date, quatrer to date etc) to be implemented instantaneously, and even better vs the same period last year.
The KPIs added to the cube are just one step further (another tab in the Visual Studio Designer). So now implementing and reporting on these measures is so much quicker and much more accurate than it ever was before. The big challenge is always (and should always be), exactly what does the business need to measure?
Where the cubes really come into their own (as opposed to stored procedures running of relational tables) is in the MDX scripted calculations, in particular I have found the 'periods to date' functions extremely useful. Previously we would have had to precalculate these measures, or in some cases do it on the run, through a SP with mountains of T-SQL code. Now we simply have the short MDX expression on the cube - even better is that it allows any period calculation ( Week to date, Month to date, quatrer to date etc) to be implemented instantaneously, and even better vs the same period last year.
The KPIs added to the cube are just one step further (another tab in the Visual Studio Designer). So now implementing and reporting on these measures is so much quicker and much more accurate than it ever was before. The big challenge is always (and should always be), exactly what does the business need to measure?
Wednesday, 18 July 2007
Current Technologies
There are a number of database providers out there but the one I will concentrate on in this blog is Microsoft SQL Server 2005. Why? You may ask. No we are not MS Gold partners but in this product we have a full end to end Business Intelligence (BI) solution - specifically aimed at the middle market.
We have the ETL technology - SSIS (SQL Server Integration Services), which is a massive leap forward from the old DTS. We have the OLAP analysis layer - SSAS (SQL Server Analysis Services), which provides the multidimensionality required in modern analysis and we have SSRS (SQL Server Reporting Services) which provides the report delivery mechanism. In my opinion the last is the one that needs updating the most. Of course, on top of this we have the actual SQL Server database engine itself which rivals the likes of Oracle and the whole package together becomes massively cost effective as full BI product.
How does this fit in with KPI discussions? Well, to be able to deliver accurate measurements you need a trustworthy datasource in the form of a datawarehouse. Both SSIS and SSAS provide the mechanism to get us there and in the next article I will discuss SSAS in a bit more detail, specifically referring to how we use it for KPI measurements.
We have the ETL technology - SSIS (SQL Server Integration Services), which is a massive leap forward from the old DTS. We have the OLAP analysis layer - SSAS (SQL Server Analysis Services), which provides the multidimensionality required in modern analysis and we have SSRS (SQL Server Reporting Services) which provides the report delivery mechanism. In my opinion the last is the one that needs updating the most. Of course, on top of this we have the actual SQL Server database engine itself which rivals the likes of Oracle and the whole package together becomes massively cost effective as full BI product.
How does this fit in with KPI discussions? Well, to be able to deliver accurate measurements you need a trustworthy datasource in the form of a datawarehouse. Both SSIS and SSAS provide the mechanism to get us there and in the next article I will discuss SSAS in a bit more detail, specifically referring to how we use it for KPI measurements.
Friday, 13 July 2007
Can't See the Wood for the Trees
Many companies we have dealt with decide to implement a full MIS solution after many years of neglect, however, where on earth do they start? After all there are loads of technologies out there with snazzy front-ends, dashboards, scorecards - you name it.
To me the critical thing is designing the data warehouse correctly, in a well designed star or snowflake schema. So very often these projects are split onto phases whereby the datawarehouse is designed and built first. Some power users may even be given read access to this database to implement their own queries - this needs to be controlled carefully.
The next phase would take into consideration how the data should be presented, what reports should be deliverded and how, and also what KPIs will be displayed and to whom. My next blog will discuss current technologies and best practice for presenting an MIS solution.
To me the critical thing is designing the data warehouse correctly, in a well designed star or snowflake schema. So very often these projects are split onto phases whereby the datawarehouse is designed and built first. Some power users may even be given read access to this database to implement their own queries - this needs to be controlled carefully.
The next phase would take into consideration how the data should be presented, what reports should be deliverded and how, and also what KPIs will be displayed and to whom. My next blog will discuss current technologies and best practice for presenting an MIS solution.
Subscribe to:
Posts (Atom)
