What do YOU think Excel is for?

Printing graph paper.

Excel is a data scratch pad. The cells are very versatile, being able to contain multiple types of data (numerical, date, time, text) without having to first be defined as such, or formulas instead, as well as labels for the data in adjoining cells, headers for the cells that follow below, or links to external data. And it can sort and do other cute tricks.

I rarely keep a project in Excel once it gets both complex and ongoing/permanent in nature — if it heads that direction I’m switching to FileMaker — but Excel is marvelous for getting started, when you have some data you want to type in and some of it is the sum of or product of some other info, and you want some running or grand totals or something.

I have an Office 365 subscription for $99 per year that allows five people to use it, and each person can install it on five devices. I used to hate subscriptions but that is a lot cheaper than what I used to pay under the old model even if I waited six years between upgrades.

If you have multiple people and devices, yeah, way cheaper. That’s fair.

I use Excel for pretty much everything except actual writing. I make financial spreadsheets with it, I keep lists on it, I make timelines with it, I create fictional characters with it, I keep linguistics data on it and make conlangs with it, I’ve tried to make an RPG with it but I keep dumping it and starting over.

ETA: Oh, and I once made a map of our company’s floorplan, using colors and patterns to (pretty accurately) represent the different areas’ floor tiles.

I’ve embedded Excel charts into PowerPoint.

That always goes over well in meetings. Management loves their charts.

Office 365 is free to educators and students. I registered using my “dot edu dot cn” E-mail address with no problem.

I’ve actually had engineers try to do real data analysis in Excel, including one who tried to do an FFT using cell macros (despite the fact that she had Matlab and the Signals Processing Toolbox available to her). I point out that it isn’t even all the great for data visualization of 2D data, and is absolutely terrible when doing anything but trivial operations because of how difficult it is to debug errors, but for many people it is the only data handling tool that they know and they use it for everything instead of spending some time to learn Matlab or Python. I even sent one engineer to a Matlab training class only to still have him cutting and pasting data and manually making each plot instead of writing a single script that could import, process, plot, and save data for a bunch of channels in one batch.

That is basically the perfect way to use Excel; you can import, inspect, and manually play with the data, then when you are ready to do more complex operations you can import it into Pandas (if you are a Python user) or otherwise build tools and scripts to do complex or involved manipulations. The problem comes when you interact with other people who insist on only using Excel. I’ve spent so much time rebuilding tools that I originally scripted in Python or Matlab to make something comparable in Excel, or trying to summarize data into a compact form suitable for a spreadsheet, and it is all just basically extra overhead to doing real work. At this point, it something requires filtering or manipulations beyond what I can do with a PivotTable or some creative conditional formatting, I kill Excel and move directly to Python/NumPy/SciPy.


Even at that, if they’re formatted in a standardized way, it’s super trivial to just parse out that number data into its own column and deal with it as such.

Excel, and spreadsheets in general ought to be used for some combination of calculation and presentation, and very small scale databasey type stuff.

Problem is, things have a tendency to grow, and stuff that someone developed in Excel with macros and all sorts of jack-legged workarounds is devilishly hard to reverse engineer and translate into anything else. So keep that in mind- if you think that your procedure might end up getting too big for a spreadsheet, don’t do it in a spreadsheet in the first place, regardless of how fast you might need it, or how easy it might be in the moment. It’ll save you trouble in the long haul, believe me.

Dude, you need to get out more. If $70 for 3 people with drinks is “nice”, I’d hate to see “normal”.

Generally speaking, nice starts at about $70 per person with drinks and can go up from there.

Welp, I’ve read about a third of the thread, and I’m ready to answer the title question.

I think Excel is for other people.

All the usual financial stuff, job revenue modeling , building models of product / service cost vs revenue, capex needs , overhead scaling and all that fun stuff.
Project economics and NPV calculations and all the PowerPoint charts that are needed.
It gets used as a field invoice, field revenue and cost tracking per job, all manner of job and equipment performance report templates.
Scheduling templates and resource allocation.
Simple databases for job history , operational performance , issue tracking etc
Bill of materials , equipment build tracking , inventory, POs and goods recipt docs.(ok we got a quickbooks compatible ERP pretty quickly when we were a startup)

Some of that stuff gets imported into a system.of some sort (I’ll not start an SAP rant, we have some specific product mangement process data entry templates that make me a tad irate) but for the raw templates which need some sort of user input that’s not from a sensor measurement, they are still excel templates. The user base is not huge and they are trained company staff so it works. Anything I have control over has the requirement, you only enter a specific piece of information once, no matter where else it is needed and you better think about the input format.
This company has vast amounts of reports of useful past data we could mine, but it’s all pretty useless due to the way it was entered. That said I have. a couple of projects on the go for some people (not me for sure) to look at NLP to try and get something out of it all.

I did some Monte carlo simulation modeling many years back, that was pretty clunky until I found the crystal ball plugin software.
A lot of the other part of my day job is dealing with how to understand and make a decision from downhole sensor data . I often use excel as a scratch pad to work out the algorithms on ideal data , but then I punt it to matab for when I need more than a few test data points, and for sure when we get to real world data , that’s in matlab for sure. We also keep ‘sanity check’ spreadsheet so when new algorithms are implemented into the operational software we can hand check the outputs during software testing and actual real field trials.

It would be a strange day when I dont have multiple excel work books open for different needs.
Could I use project for scheduling , some data base for records and MS Forms or adobe or some such for forms, sure , but for ideation, quick work , flexibility etc you can’t beat excel. When you need to scale the tool , for sure move to something more suitable.

Geez, I’ll say! As they noted later, clearly the poster quoted about a “nice” restaurant meant reasonably upscale in the cost sense. And of course I completely agree. I’d put “nice” in that upscale sense as meaning that it starts at $70-$100 per person before tip, with perhaps an inexpensive glass of wine each, and rapidly goes up from there. No, I’m not rich, and I rarely indulge in such pleasures, but at least I know what a “nice” restaurant is. The highest tab that I’ve personally paid for was around $300 per person, and that was most mostly food, with no expensive drinks.

I’ve used Excel in small ways myself, just because it’s part of the Office package I already have. And it’s a fantastic tool. I also know someone who’s been running a successful small consulting business from home who uses Excel to maintain customer databases. Like all tools, it has a range of uses, and like all tools, it can be pushed beyond the limits it was designed for. That’s not a criticism of the tool itself.

I’m happy to bash Microsoft whenever it needs bashing, but many aspects of Office were ultimately pretty well designed and integrated from the user standpoint, even if some of the UI sucks. The integration of Excel as a de facto ODBC database for query functions like Word mail/merge, for example, is extremely useful.

This reinforces my stand. For me, the computer is primarily a toy.

I reserve most of my Microsoft praise for Excel.

Unlike a lot of other Microsoft offerings,

a) It isn’t bloated and buggy and standards-noncompliant. Even with the damn ribbon, it isn’t horribly off-putting to work with. (PS: we Mac users still get an overhead menu, heh heh…)

b) They didn’t acquire it by buying some other company to raid their merchandise and rebrand it as Microsoft. As far as I can tell, Excel was truly developed in-house.

c) They didn’t push aside some significantly better competing software to enshrine it. I’m sure Lotus 123 and Quattro Pro have their fans, but it’s difficult to argue that Excel is inherently inferior and didn’t deserve it’s hegemony in the spreadsheet market.

d) It plays nice with other software, not just other Microsoft software. Exporting to CSV, tab delim text, SYLK, DBF, DIF, space delimited text, and earlier iterations of the Excel format, thank you very much; being able to query other data sources (and again, not just Microsoft’s in-house implementations, but ODBC as used by non-MS environments).

I’m sure you liked the flight simulator!


I have used excel automation as part of software that generates graphs of industrial data.

Our customers were not programmers, but they were domain experts who wanted graphics on an intranet display that showed the results of various complex calculations they were doing. They wanted the ability to create new ones, edit old ones, change formatting whenever they wanted, etc.

So… We gave them an excel template for what the graph name and columns of data labels had to be, then at runtime we’d inject data into the spreadsheet, generate the graph, save it, and link to it on the intranet page. A few performance optimizations such as only allowing the graph to update once per defined imterval was all that was needed.

This saved is from having to build an entire editing suite for data and graph creation, and the customer from having to be trained on a new graph builder, and it worked perfectly.

Excel is used by everybody for everything because even with the most restrictive IT policy you can count on excel being installed.

Forget other reasons.

It is the only program that consistently is there.
Idiot IT types have trained generations to do everything with excel and now they are complaining they do everything with excel.

“Why are you trying to fillet a fish with a flathead screwdriver?”
“Because it works better than the Phillips, duh!”


From “Triumph of the Nerds”

Now we’ve set this up, OK. Then we type a new value in, then I’m going to take that one hundred, I’m going to change it right and here, it recalculated! Woa! That saved me so much time. People who saw it and went and got it like an accountant, I remember showing it to one around here and he started shaking and said that’s what I do all week, I could do it in an hour you know, you know, they would take their credit cards and shove them in your face. I meet these people now they come up to me and say I gotta tell you you know…
BOB: You changed my life.
DAN: You changed my life. You made accounting fun and…

Oh, so Mr. Anchovy finally managed to became a lion tamer thanks to Excel? That is indeed a great use for it.