CIO here

Can confirm that I will take Excel from the cold dead hands of our CFO (and the rest of his dept). Hell even macOS Excel is too much weirdness for them. And being user of both platforms I agree with them.

The rest of the company are very happy though on Google Worksheets, making lists and replying to RFP requests in .xlsx, on their Macbook.

We do a survey every year; 95% of staff is happy with the resources out at their disposition. Including those working on a platform that is not their native platform.

End users are more flexible than we think. Especially non technical ones.

Excel is powerful? Does it do forecasting, statistics or machine learning? Only in a rudimentary sense. Spreadsheets like Excel are low-code tools. Experts in those domains use a real programming language.

People use Excel because it is familiar, because it has distribution. They don't (care to) know any alternatives. If they'd never been familiar with Excel they wouldn't ask for it.

As a former ML engineer, Excel is a surprisingly practical tool even for ML folks. It’s a very tactile tool to touch and feel your data and do quick what ifs without doing a lot of coding. Want a graph? Plot it instead of fudging with matplotlib.

Forecasting? Yes. Dynamic formulas? Yes — much better than writing code (you’d have to write a reactive widget in Jupyter).

Most people don’t need to do ML but they need to do complicated formulas and get immediate feedback. Excel is really good at that. Libre office can do 90% of it but Excel has some advanced features that it doesn’t come close to.

Excel is for playing with data that is already aggregated. You don't see any data scientist in a tech company reaching for spreadsheets. Business analysts use them, maybe, because they work with the raw data rolled up by the engineers.

How are you even going to ingest the data you want to visualize from the database it's in otherwise? It won't fit on your computer as is. You need something like Tableau to query and inspect it.

edit: Sure, let the finance people have their spreadsheets, but don't let that stuff anywhere near engineering and data science. It's a major anti-pattern.

Yes, yes, Excel is the Blub tool. Plenty of other options which are better typed, performant, intuitive, whatever.

Every F500 company has accountants, FPNA, auditors, likely even the CFO who are doing their analysis in Excel. They live and breathe it. It is their everything app which they run 40 hours a week. They know Excel, they are the 10x rockstars of their field.

It is a common refrain here that you should absolutely invest in SAAS tools at $/month if it makes your workers more productive. Even if they are not slinging code, these people already have the best tool for their job. Making them all switch to save a few bucks to a less feature complete option is hamstringing yourself.

Could many people switch without issue? Sure, but it is foolhardy to imagine a 100% cut over without enormous pain to the elite staff.

Good points. It is egregious that HN 'technologists' will tell people that the tool that they use for 40+ hours a week to _get stuff done_ is completely wrong and that they must use python and a text editor instead.

Excel is not powerful. People use it because they are too lazy to learn something new.

There are countless millions of people that are experts in certain domains that do not have the aptitude, confidence, or training to use high-code tools. It is insulting to imply that those people are not experts because they don't use a 'real programming language'. Besides, while imperative programming languages are good at solving a great many problems, they are far from being the best tool for every person, every problem, and every environment.

> There is nothing on the planet that competes with Excel

Debatable. Even then, why that matters? I would think 90% of the world population never used Excel anyway.

> You can just use notepad, right?

No, you are missing the point. Intentionally.

Imagine a world where if experts didn’t spend hundreds of hours learning the non transferable skills of VBA macros and Power-Noun integrations, but rather they spent the much less time developing a basic understanding of scripting and prompting and managed tabular data in ways vastly more powerful than excel in a way that is transparent and debugable. Would we not be much better off? Excel is not an engine of progress. It is an anchor.

this is exactly the point I was trying to make.