Spreadsheets, we have a problem!
In previous posts, we’ve talked a lot about how spreadsheets empower domain experts to solve many of their own IT needs:
2. Excel — a Domain Specific Language for Finance?
3. How can I get things done NOW (without talking to IT)?
However, you don’t have to spend a lot of time around spreadsheets — or in business in general — to realize some of spreadsheets’ obvious shortcomings.
GRID is built to allow spreadsheet users to make even more of their spreadsheet skills and assets by overcoming some of these weaknesses.
But what are they?
Governance and administration
The main reasons IT people, auditors and to some extent corporate management are often not too keen on spreadsheet-based solutions — and what may have coined terms such as “Excel hell” and “Frankensheets” — probably have to do with governance and administration:
Access control is difficult, and very binary: Either you give someone the entire sheet and they can do pretty much anything with it, or you give them a very limited — often static — version of the sheet and its value is greatly reduced. There are ways to be more granular about this, but very few people use them.
There are no audit trails. Who changed what, when and why?
Version control is mostly in form of naming conventions including dates, employees’ initials and status of the document. Ever seen a file with a name like: Budget2019-rev2-HGcomments.xlsx?
Important spreadsheets may not be a part of the company’s official document store, nor even backed up as they live on employees local computers or private cloud accounts (the latter most often on Google Sheets because they needed the collaboration features, but GSuite is not officially approved by the CIO).
In short, the organization has very little insight into or control over the “spreadsheet fabric” that plays an incredibly important role in almost every organization. Often far bigger than companies realize themselves.
Usability and user interface
Some of the governance and administration problems described above may be an annoyance to business users, but mainly they are a thorn in the eye of IT and management. However, there is another set of issues that end users feel more directly on their skin, and they all have to do with the fact that what’s made in a spreadsheet, stays in a spreadsheet.
This has several implications:
The way to distribute data, analysis or a model pulled together in a spreadsheet is sharing the spreadsheet itself. If it’s made in Excel, the recipient needs to open it in Excel (still mostly local, but usage of Excel Online is gradually on the rise), if it’s made in Google Sheets the recipient needs to open it in Google Sheets.
Mobile spreadsheet experience is very limited, to say the least. If you receive a spreadsheet file attached to an email you may be able to look at approximately 10 rows by 5 columns on your mobile phone, but you get no oversight and cannot interact with it. Even opening a cloud based spreadsheet in one of the native mobile spreadsheet applications is painful and far from the user experience of desktop spreadsheets, let alone the everyday mobile applications we’ve all gotten used to. The spreadsheet “grid” metaphor simply doesn’t lend itself well for mobile.
Spreadsheets are very limited in terms presentation and interaction. No matter how good you are at changing background colors, placing images and configuring your fonts, the layout never looks good. And building interaction beyond basic typing into cells is hard and may require macros or other tricks that will cause issues when distributing the spreadsheet.
Errors, risk and quality assurance
Despite what has been detailed here, the issues that perhaps get the most attention when it comes to spreadsheets have to do with errors, quality assurance and the associated risk.
These are certainly worthy of attention, but personally I think they are in many ways a by-product of how important much of the work that happens in spreadsheets is. It’s not like there is strict quality assurance around our emails, slide decks or text documents either! It’s more that those errors are (usually) not as consequential. It turns out that there’s a qualitative difference between errors in “hard” data and logic on the one hand, and the “soft” written word on the other.
Don’t get me wrong, these are important issues, and spreadsheet software can certainly be improved to help avoid them, but the problem has more to do with people (training, etc) and processes than the spreadsheet as a tool.
In other words: Spreadsheets are where they are because they are very powerful tools. When people get their hands on powerful tools they become more efficient and productive, but don’t expect perfect outcome without proper guidance and safety procedures.
This article has also been published on Medium.