[delusional]

> So, in practice what usually happens is that, when people hit UX challenges they go to a consulting firm and commission a backoffice software to address the Excel limitations. Some business may pay for nocode but that is very rare. They go to backoffice software firm and they build a CRUD software that is now not replacing Excel but compliments it.

Heh, where were you 6 month ago when my team was pitching some of our traders on replacing a system they'd built for balancing their strategy. It was built in excel of course, but what surprised me was the resistance to changing it at all. They all hated the thing, it was slow and crashed often, and right in the middle of the difficult part of the work too. Working with it was a terrible experience, even for people used to medium business enterprise crap. Yet they refused to consider any alternative to building some CRUD around it that would extract the worst parts.

The issue we faced was that there was no way we could contribute to this excel monster while still following the risk tolerance of our department. It's not that we're opposed to building something that is probably a bad idea, if it helps build relationships that let us build something better later. It's more that as soon as IT touches it, we get to own every legal aspect of it too. So they ended up with an external consultant that built them what was essentially an external database they could query, which was then supplied with data by some program somewhere.

It ended exactly as you say. They got a back-office firm to develop some one-off hack that alleviated the immediate problem, without replacing excel.

[antasvara]

I think this highlights the main issue with replacing Excel, which is that most people have an undocumented, informal workflow with their tools in Excel that would be difficult to replicate fully in other software.

Relevant XKCD: https://xkcd.com/1172/

As someone that has had other software tools replaced with alternatives (think Slack replaced with Teams), the new solution almost always speeds up and fixes a lot of problems. It also invariably screws up at least one thing that I need the tool to do as a part of my job.

I love my IT colleagues, but (and I mean no offense to them or other IT folks) they don't understand that the way we use our internal tools is super flexible and doesn't adhere to an end to end process. It's how we ended up with an entire software platform that takes more manual steps, breaks in a way that we can't fix, and still requires the use of Excel to get the output into the formats that our stakeholders need. It also can't adapt quickly when we need to change things in our process.

It is a technically impressive tool that does the things they designed it to do super well. But we're like a month away from needing an Excel tool (or another piece of software) that can reliably do all the things we didn't realize relied on the "hot spacebar" from the XKCD.