Are errors rife in your reporting system?

Filter By Topic

All

furious businessman throws a punch into the computer

It happened to me with a client the other day. I was sweating! I was comparing data in new reports with those from the "old” system: FRx. I was reviewing previous years data, and I couldn't reconcile them with the new way of doing things.

It was my last test, and then... BOOM ! ERRORS !

Why? The income statement I was using from FRx had a couple of hard-coded line items for special adjustments, yet I could still run the report for previous years. This meant most of the data was correct except for the hard coded line items. Oops!  Any time this report was run for previous years were always going to be wrong, and I didn't know about it!

So how do you fix this?

Dump the reporting tool.  It’s just too lightweight and requires too much time and effort to maintain.  If your reporting tool requires hard coding structures (account ranges for instance) into the report template itself, then you might as well be using Excel (people often do). Especially if you’re going to hard-code structures, consolidation points, calculations, etc. in the report.

The best way to report

Use BI software that centralize hierarchies and data to manage accounts, divisions, companies, etc. with a tool/database that does this efficiently and without hard-coded reports.

Centralizing your data and structures means having a single database that contains all the information about your business that your reports need.  It’s the old “one version of the truth” that I am sure you've heard before. Centralized databases contain numbers, account structures, divisions, cost centers, people, etc. in one place for easy access and trustworthy results.  All reports get their structural data and numeric data from here.  Everyone looking at reports knows they are looking at the right data all the time because the templates just display data from your central database and nothing is customized in the report.

I prefer cubes for this purpose but in-memory tables that sort of mimic cubes are good too. But the current offerings from vendors are less flexible and powerful than cubes.  Relational tables will work too, but performance is much slower, harder to manage and harder to maintain. Also, you will often have to tune your reports, so they don’t run slow and don’t affect everyone else’s reports too. Cubes don’t have these issues.

The good news

We found the problem and since turned FRx off. The client now has much less to worry about and has much more time to complete all of those unfinished projects because they have cubes and CALUMO.

If you want more information about this, email me – wleitch@calumo.com

compare-button

Leave a Reply

Your email address will not be published.