IT guy here, Excel is a data analytics tool, not a database, not a word processor, not a sales system, not a photo album, not a notepad, not a paint program.
If at anytime you are treating Excel as a database, you are doing it wrong, and you deserve me mocking you when asking for help recovering it when it breaks, I won't as I am not a dick, but if I did, you would deserve it.
If you want a database, build an SQL database, or have someone build it for you, not me.
Years ago, I've recommended KeePass to a girl from marketing who kept a long list of passwords on paper on her desk. She forgot the master pass after a week or so. That was the end of my trust in users' ability to maintain a safe environment.
I work for a Fortune 500 company and I can tell you the reason why excel (and Google sheets) are used inappropriately is because cyber data controls make creating and maintaining a database very hard. Not only that but the skills required to know how to make a table in a spreadsheet is nowhere near the skills required to deploy, maintain, and provision a database table.
Spreadsheets don't require a UI to be built. People don't have to learn a new app just to be able to see data.
I'm an IT guy too and I'm the first to tell you that spreadsheets suck. But when it takes an act of a board to create new tables in a database, I tell ya....might as well just use spreadsheets.
The problem is, people dig to deep into excel functions, some of them could easily build a database or do some programming (if/else), but they know nothing outside of their ms-office -ecosystem.
Just a hint for ms-office devs, why not a low-code-builder with SQL backend. Just call it squirrel or powersql or something.
It's more than just knowing things outside the ms office ecosystem. People use the tools they have. So when IT locks down the whole system and it takes an act of God to get anything else installed, you find ways to hammer that nail with whatever blunt object you have in hand.
Yes, there are the people who think there is genuinely no problem with this. Just like there are people who will never delete a line of code in favor of commenting everything and who refuse to write commit messages no matter how many times their co-workers beg them to.
But, generally, people know it is a horrible workflow and is prone to failure. But there is no time and resources available to revamp the entire system. Because that likely involves going "offline" for the migration as well as the subsequent retraining. Its no different than the technical debt we all laugh and cry about. We know that server is held together with chewing gum and shoe strings but we don't have time or authorization to tear it down and rebuild it from scratch. We are just hoping it doesn't fail at a bad time.
If you're lucky? You can periodically export the excel sheet to a database (sql or access, it doesn't matter). You are still doing things wrong but you at least have a recovery option at that point. But, if you can't, you are more or less fucked and know it.
As for another Lesson Learned. A database solution without high-ish availability and backups is actually worse than the god awful spreadsheet. Because people know when the spreadsheet fail and likely are self-important enough they will stop everything to recover it. People tend to ignore error messages when they try to submit a record or save something and you find out that the disk failed last week and you lost everything.
Shit, I'll mock them. I'm too jaded and depressed at this point in my career to give a fuck. I'll go full Nick Burns on their asses if one of my end users wants to use Excel as a database and expects me to make it work. The may even learn something in the process. It might be the fact that I'm a dick, but everyone figures that out pretty quickly.
It's not even a good analytics tool. If you submit an academic paper with excel plots in it, I'll reject that shit without reading it and type "lmaoooooooo..." To the review character limit.
My 12 year old child knows how to use matplotlib and he thinks Santa can fit down a chimney.
It is good enough for financial and marketing analytics, just because there are better tools for scientific applications doesn't make Excel a bad analytic tool for general use.