M.E.S.S. Revisited

From my earlier post, you know that Manually Entered SpreadSheets ("M.E.S.S.") are my nemesis. Not only are they time consuming to create and maintain, but they are also difficult to audit. As anyone who has been through a Sarbanes-Oxley (Bill 198) review knows, spreadsheets represent a significant control risk because of how easily errors can be introduced without detection. For example:

At this point you should ensure that there are the appropriate backup procedures for the spreadsheet. Ideally you will have documentation to support the spreadsheet as well. Then on an annual or biannual basis spreadsheets should be revisited, comparisons run on the tested versions and the audit trail reviewed to ensure that it is still functioning as required. Needless to say, any significant changes made to the model should be carefully considered, documented and re-tested at the time the changes are made.

In short, spreadsheets need to be considered as a separate financial computer system and all the controls associated with a financial system need to be applied to the extent the spreadsheet is to be relied upon for financial reporting. All of a sudden your "simple" spreadsheet is expensive to audit!

Currently I am working on a project to eliminate financial spreadsheets for an international client. Significant areas include:

• Multilevel consolidations,
  
• Canadian & US GAAP adjustments, and
  
• Tax valuations.

Luckily, this project is part of a larger implementation so I was able to participate in the re-design of the chart of accounts. We have come up with a way to isolate the international subsidiaries' raw data from the different GAAP and tax adjustments, all within the same Microsoft Dynamics GP (Great Plains) general ledger. When the client wants a US GAAP statement, they just add the US GAAP adjustments to the raw data using a pre-coded report in the FRx report writer. No M.E.S.S.!

Eliminate M.E.S.S.

M.E.S.S. stands for Manually Entered SpreadSheets. Don't get me wrong. I was raised on spreadsheets: first Visicalc, then Lotus 123, finally Excel. They are a second language to me. But when I see people pecking away entering raw data in them, I have to wonder why they can't get it from their accounting system. Ideally, their time should be spent analyzing their data, not keying it in.

Why does M.E.S.S. happen?

• A lack of trust in the accounting system, particularly when someone was not in favour of the change to begin with. Early in my career, a mining client converted their accounting system. The Accounts Payable manager, close to retirement, didn't trust the new computer, so she kept track of all of her vendors using a spreadsheet. About a month after going live, the new system crashed. They were forced to recreate the information manually. Everyone that is, except Accounts Payable. Guess who felt completely vindicated?

• The accounting system does not allow for all requirements. A government agency I worked with had a large number of "sidecar systems", meaning M.E.S.S. The reason was that their accounting system didn't track the information they needed. The spreadsheets were so ingrained in the system that when we converted to new software many people didn't think to include them in their requirements. The implementation team had to actually root them out and convince people that they were no longer necessary.
  

• Insufficient detail in the General Ledger. At one client it was someone's job to take all of the entries in the Employee Advances account and separate them by employee in a spreadsheet, so that the individual advances could be tracked. The General Ledger needed to have an account (or subaccount) for each employee so that time wouldn't be wasted making the entries twice.
  

Is M.E.S.S. ever justified? Of course if you are doing something that has never been done before, like due diligence on the purchase of a company, then you have no choice. The issue for me is when the same spreadsheet is prepared month after month. Then, it's a M.E.S.S.!