I know a manufacturing plant that used an Excel spreadsheet to do all its production planning. There was only one person who understood the spreadsheet and could modify it, a consultant who made more than the plant manager.
"Understandable and fixable" depends more on the complexity of the application rather than the fact it's in Excel.
Agree. IT has forgotten that computing should enable more people to be producers instead of mere consumers. IT management cares about control, audit, permissions and expense - no focus on achieving productivity in the workplace and in many cases are anti-user.
If you try running a business where several workers get involved with fixing and extending information systems (e.g. spreadsheets), you'll soon understand why successful IT management cares about controls, audits and permissions.
I would argue that excel being "fixable" by regular office workers is half the reason why these projects fail in the first place. I've worked on migrating people's reporting to BI platforms before and what looks like a simple spreadsheet produced monthly is often really 12 different sets of formulas, special cases, kluges, hard-coded data and long-gone sources etc. etc. Because instead of correcting the source of data used for the report, it's all "done in post" in the excel sheet itself by a regular office worker.
Yes, important point. I wonder if a system could be made that made things properly fixable. As in, a regular office worker spots an error. Can fix it in post. But in doing so they will also create a change suggestion for the source data that can be approved and committed by master data admins.
The fact that every office worker understands excel, does not mean that every office worker understands every excel sheet.
Most of the projects we did in consultancy dev, was turning that one critical excel sheet nobody but 'the excel guy' understands into a simple to use web application, so that everybody could use it and the business won't explode when mr. excel would leave the shop.
Most of those office workers were not capable of fixing anything on the first day they used Excel. Many didn't understand it at all. The main difference isn't that Excel is super accessible and easy to use for non-technical people; it's its ubiquity, and especially that of training on its usage.
Ubiquity is important but it's not the only important factor. An excel sheet can typically be downloaded and experimented with. You can't download an ERP system and try stuff with it.
This is so huge in finance. Lots of smaller shops will hire data scientists or even SWEs hoping to up productivity and replace slow Excel sheets, and end up disilusioned when the team just glues together some Python scripts with no UI and no way for stakeholders to tinker without talking to someone else first.
It's a bit surprising that we don't have that feature as a requirement for most IT infrastructure. It would make it so much more usable.