As an accountant, you do most accounting jobs using accounting software, and a few using Excel. It's usually the smaller ones where you'd use Excel; maybe a sole trader who doesn't need to track debtors and creditors and doesn't have a dedicated business bank account, say.
Sometimes a reluctant business owner might be resistant to using accounting software, and the classic objection is "why do I have to pay for a glorified spreadsheet?". It's not a glorified spreadsheet, that's why. It's a database, and a database has built-in checks on the integrity of the data (in the case of accounting software, thanks to the beauty of double entry bookkeeping).
If you're not careful, then those little Excel jobs that seem the most straightforward can actually be the riskiest ones, where you can make expensive or embarrassing blunders. When we do accounts in Excel, we always build in some kind of proof that all the transactions feeding into the accounts have been captured once, and only once, and that any adjustments are complete and transparent. If you don't do that, all you've got is some numbers that may or may not be right. Generally, when staff join us from another firm, they've not come across this kind of check before. But the extra minute or two spent proving that the output is consistent with the inputs are vital if you want the output to be reliable and credible.
Whenever we read about a big error involving the numbers in the accounts of some listed company - say an enormous error in a stock valuation - we always speculate that it will be a problem that's arisen in Excel, outside of the core accounting software. A SUM formula that hasn't been stretched to pick up some extra rows, for example. We're *always* right. And often the company will try to explain it away as a problem with Excel. It's not, of course. It's a problem with Excel being used for an inappropriate purpose, or Excel being used for an appropriate purpose but without some kind of integrity check in place.
So, when you're using Excel for accounting, it's important to think "we need to have some sort of proof in place that the data is complete". You also need that if you're using it to track a deadly infectious disease.
Comments