Skip to main content

Don't Trust Your Data to Automation

Mon, 06/09/2014 - 11:13am | by jennaleesmith1

I recently switched my financial system from a series of individually-managed spreadsheets to the popular financial-tracking service Mint. Lifehacker rated Mint as the "best mobile personal finance tool," and I was eager to offload the responsibility of sorting expenses and income by category and location to an automated service. 

At first, it all went well. I still kept filing business-related receipts, but I no longer had to keep personal grocery receipts or individually track every iTunes transaction. I didn't have to spend over an hour every week figuring out where my money was going; I could simply log on to Mint and click on one of its handy graphs. Starting every morning with a quick Mint check-in became one of the highlights of my day. In fact, using Mint seemed to earn me more money; according to Mint, even though I was only halfway through September, I was close to meeting my monthly freelance income goal. 

As October neared, and Mint's morning check-in told me I had surpassed my monthly income estimation by nearly $3,000 dollars, I realized the online service was actually showing incorrect data. I had to dig down deep into Mint's interface to realize that it was actually recording all of my freelance income twice: once when my clients paid me through PayPal, and once when I transferred those PayPal funds to my brick-and-mortar bank account. Every increase in a bank balance, whether from a client payment or from a transfer, was recorded as new income, which is why the data was so wildly off the mark. 

The truth is that, even with advances in automation, we still have to perform regular checks to ensure that our systems aren't providing incorrect information. This is the case both with financial systems like Mint as well as CRM systems, mailing lists, Facebook business pages, and any other site that offers to do the work for us. As the saying goes: trust, but verify.

Or, in more business-like terms: be the intelligence behind your business intelligence software. To quote Gabe Fenton Boulder Co area business and financial consultant:

Business intelligence refers to a vast collection of data that plays an integral role in the development and ongoing success of a company. While some aspects of business intelligence are theoretical and others are strictly analytical, it is essential that all forward-moving professionals understand how to interpret this data in order to determine a progressive strategy for a venture.

What does that mean for those of us running small businesses? For starters, if your financial software shows a sudden uptick in income, don't automatically assume the information is correct -- and certainly don't make any advance purchases on what may be non-existent funds! 

If you haven't looked into your contact management system recently, it's probably time to do a serious scan of the names and addresses, and to update people who have moved, changed jobs, or otherwise no longer fit with the provided data. If your website shows an increase in traffic from a particular source, visit that source, figure out why you were linked, and see how that traffic affects your website conversion rate.

Don't take your business intelligence software at face value. Often, the numbers are not what they seem. Other times, the data represents trends that are only visible if you take the time to interpret the data. Building a business strategy on a single software-generated graph or pie chart only leads to trouble down the road; like it or not, you need to spend as much time with your business intelligence software as you would have done sorting the information by hand.

That's what I do, anyway. I spend 30 to 60 minutes every Friday making sure all of my Mint transactions were allocated correctly. It's the only way I can guarantee that the automated results will be accurate. Computer programs may do a lot for us, but they still can't think -- and that means we still have to do the thinking for them, in order to fully trust the results.

No votes yet
Bookmark and Share