Its 1987

The dawn of the personal computing era has arrived, and you’re at home seated in front of your brand-new PC you picked up from the local electronics store. Suddenly, out of nowhere, you’re greeted by a time traveller who claims to be from the year 2021. He sees the bulky, beige appliance sitting on your desk and can’t wait to tell you about the future. “Where I come from, computers will be processing teraflops for data without breaking a resistor.” Amazing!” you say to yourself. “Not only that, people will be able to broadcast live events to the world without the need any of the Big Three Networks”, he continues. “The future is limitless, haha!” you think in your head. “Lastly, you’ll be able to make phone calls with a device the size of your hand but has the power of several mainframe computers”, he exclaims, building to a crescendo. Yes! Naturally! It makes total sense! Filled with awe and excitement, you ask this stranger, “Tell me, traveller, what will be the most popular and innovate computer program in this, the most amazing of futures?”

He pauses awkwardly and replies, “Spreadsheets.”

“Spreadsheets?” you spit out incredulously.

“Yes, uh…but these spreadsheet programs will be amazing. It will be written in 64-bit code, and it will have a ribbon — ” He has no time to finish as you grab the obvious charlatan by the neck and throw him out of your house.

‍

Hopefully, this story serves to preface the fact that spreadsheet applications have, in fact, become ubiquitous in our world today. What was originally meant for the accounting departments of America soon spread far beyond the office. Spreadsheets are now used at home to plan meals, at school to create detailed forms, at work to create mock-ups of websites, and in labs around the world to track COVID-19 data. And why not? Spreadsheets are a reliable, mature format that, while not as hip and glamorous as other apps on the market, have definitely had a long-lasting impact in our world.

A whole spreadsheet culture has materialized; there are websites, message boards, subreddits, and interest groups dedicated to sharing spreadsheet knowledge and experience, not to mention articles, guides, videos, and podcasts about spreadsheets, and even games that run on excel macros…

While one can appreciate how people have taken this versatile tool, originally designed for bookkeeping and applied it to a plethora of professional and social contexts, there’s a significant and growing risk that results from over-reliance on such a simple system. Familiarity causes many to overlook how insufficient a digital table actually is for what they’re trying to track, leading to expensive disasters. Sadly, the history of spreadsheets is filled with examples of over-reliance.

Before we dive into some of those, let’s take a brief look at that history: what are spreadsheets, and why did they come to be? The concept can be traced as far back as the 14th century in Italy (concurrent with the appearance of the modern bank in Italy around the same time). Double-entry bookkeeping was used to track and eliminate errors in accounting. If the balance sheets didn’t balance, you knew there was an error. It wasn’t until 1979 that computer programmers came up with the first electronic version, called VisiCalc.

Today, spreadsheets are an essential tool, used in some form by every organization. Even within a single company, spreadsheets are used in so many nooks and crannies: storing and tracking information such as client lists, staff charts, inventory, assets, orders, costs, birthdays, and countless other things. Spreadsheets are flexible enough to be used anywhere — and this has resulted in a widespread belief that spreadsheets are the go-to tool for anything that needs tracking. These tools are then copied and shared online, changed and distributed once more.

Sharing and recycling spreadsheets has become a mainstream practice. And why wouldn’t it be? It saves us the hassle of having to come up with a brand-new workbook, and if you find one online that doesn’t quite fit your project parameters, it’s easy enough to come up with some way to make it work — or start over in another tab. Spreadsheets are flexible enough that one rarely, if ever, has to come up with a new data visualization system. As if we don’t have more important things to do, right?

‍

The problem arises when we stop using them for simple tracking and start using spreadsheets as an entire database.

Excel, however flexible it might be, was designed to facilitate financial bookkeeping. Using a table — rows and columns — to try tracking or analyzing complex and evolving information is like using a pen and paper to paint a colourful picture. You can, technically, make shades of grey or replicate different colours by changing your strokes, but you will always be fundamentally missing information.

Because spreadsheets are so familiar and so flexible, it’s tempting to use them to track everything. But not everything fits in a spreadsheet. And for any use other than double-entry bookkeeping (from your Christmas card list to contamination exceedances in municipal groundwater), there are no built-in fail-safes to catch errors. Furthermore, when an error is found, no one can trace who made it, when was it made, or even where the data entered came from in the first place.

Conflicting naming conventions and version histories add to the confusion, and more complexity elevates the risk of an expensive catastrophe. See how quickly spreadsheets can court disaster.

People who use spreadsheets at work in no way escape these risks. In fact, high pressure, time-sensitive jobs (sound familiar?) exacerbate the very vulnerabilities spreadsheets are most at risk for: human error. And because everyone has their own unique workflow, creating the right spreadsheet system is, by necessity, a make-it-up-as-you-go process. That means everything from naming conventions to how data sets are organized will be decided by the person who controls the sheet. So anyone else — friends, co-workers, management — will need time and guidance to make sense of these documents. Lacking that, they may have to guess or even start from scratch.

Furthermore, spreadsheets applications aren’t designed to contextualize data, especially with more than one data set. Their complexity is limited to two axes. So what do people do? They create a brand-new master spreadsheet, then manually copy and paste each time-sensitive piece of information they need into it. This is just one example of where you can start to create more confusion and more opportunity for error when pushing the envelope of a table’s usefulness. Someone will try to remember in their heads how the sets are related to each other, or the notes will be in another document from another application. This all-too-common scenario may work for a time, but over time untraceable errors will corrupt the data until it becomes unreliable. And nobody will be able to tell when that happens or if it already has.

Spreadsheet applications are a fantastic invention. I, as an individual, am much better off for having them available, and the fact that people have been able to find uses for these ubiquitous tables outside their original purpose is a testament to the flexibility of the format.

But everything has its limits