Save Billions the Easy Way
Not that long ago, Oregon State University released a press release stating that its computer scientists have created a new, much simpler approach to fixing errors in spreadsheets, a system that is easy to use and might help businesses around the world reduce mistakes and save billions of dollars.
According to the article in the United States it has been estimated that 11 million people create about 100 million spreadsheets a year, which in turn might be managed by up to 60 million users ... but they are notoriously prone to errors. And that most users of spreadsheets are overconfident, they believe that the data is correct but that it has been observed that up to 90 percent of the spreadsheets being used have non-trivial errors in them. In fact, one auditor has said he never inspected a single spreadsheet during his entire career that was completely accurate.
Thus, the new approach to fixing errors, which allows a non-specialist to identify and fix a problem by selecting a fix from a short list of change suggestions, is expected to be quite beneficial. The system, which attempts to try and identify the ways that humans commonly make mistakes and then suggest what the correct approach might have been, can suggest several programming mistakes that might have created the error when an error is found and allow the user to sort through them and identify the root cause. The system is currently performing rather well. In 80% of the cases, the fix is in the top five suggestions and in 72% of the cases, the fix is among the first two suggestions. The GoalDebug System (short for "Goal Directed Debugging of Spreadsheets") gives end users the chance to explore, apply refine, and reject suggested changes, a systematic approach that allows people with comparatively little training in programming and spreadsheet development to identify and repair errors.
However, as far as I'm concerned, this is not good news. All I see is tens of thousands of man hours and millions of dollars of research funding wasted on a patch solution that doesn't address the real, underlying problem - spreadsheets are bad!
As far as I'm concerned, the answer is the development of appropriate financial applications that contain the calculations and report creation abilities end users need and negate the need for them to develop these huge complicated spreadsheets. Then there wouldn't be errors to track down in the first place.
Fortunately, for your procurement department, there is an answer - and the answer is e-Sourcing complete with spend analysis and decision optimization. That's the easy way to save billions.
You criticize spreadsheets. I agree completely, even though I distribute one in my seminars. I'd love to see a combined freight optimizer, landed cost comparer and risk analyzer combined in a commercial application aimed at source selection.
Anybody know of such a tool? I'd be happy to donate my spreadsheet to a company who wants to develop one.
Among the issues that would overtax a spreadsheet-based tool (and I haven't tried to include) are:
-multiple source-multiple destination total landed cost and decision optimizer
-currency risk analysis on an assembled product that contains parts from multiple countries/currencies
-integration of the currency risk analysis module with history-based estimates of the probability of an exchange rate change that would make a supplier selection suboptimal
I do think the probability of errors is only slightly higher in the early days of a spreadsheet than in the early days of a commercial application. It's often the fear of errors and lack of connectivity of a spreadsheet that limits its value. A spreadsheet is a good way to develop the algorithms needed to do a full blown commercial application.
Dick,
You are welcome to reach out to me and I will walk you through an online demonstration of a tool, I believe, meets your outlined 'need'.
You can reach me on the numbers provided at http://www.combinenet.com/site/contact_us/
Forgive me for not publishing my contact information in blog comments.
Paul Martyn
The Doc waxes somewhat too enthusiastically about getting rid of spreadsheets. The good news about Excel is that there are lots of competent users out there, and, as Dick points out, there are few better ways to try out an analysis.
The Doc suggested a while back that one should use a transactional analysis tool to do some initial exploration of the data. That exploration then makes optimization model choices more obvious, saving tons of time. And Excel is definitely a very useful part of that process. Here's the link:
http://blog.sourcinginnovation.com/2007/03/27/analytics-vs-optimization.aspx
I guess I was not clear. I am recommending we get rid of spreadsheets as an in-depth analysis and reporting tool, as it is simply too error prone for such tasks. It is a great user interface for quickly accessing and populating large amounts of data, particularly numerical, as a grid is simple to work with, but it is not a great tool for sophisticated analysis or reporting.
Michael,
I think you were quite clear, I just don't agree with you entirely. Much of what folks try to do with Excel (such as optimization) shouldn't be done, admittedly; but Excel is actually a fantastic reporting tool for many purposes, especially if models can be populated with complex data and iterated from an OLAP store.
Perhaps you don't think of such models as "sophisticated," but I have seen some remarkable efforts.
Eric:
Despite the fact that Excel 2007 apparently cannot even redo a simple sum on spreadsheet update?
Anyway, I have no problem using Excel for reporting, since that's a UI-use ... but I would do all the heavy lifting in another tool.