Hacker News new | past | comments | ask | show | jobs | submit login

> There are simply no viable alternatives to something like Excel, for a lot of organizations at least.

Anyone want to educate me on this? Anything I want to do in a spreadsheet, I can do in Gnumeric, Libre Office, or even Numbers; anything more complicated I'd rather do in a proper programming language. What makes Excel really so indispensable?




> anything more complicated I'd rather do in a proper programming language

First you need to understand that this is a minority position. Despite the last decade or so of "software eating the world", most line-of-business type people are not programmers, and couldn't write a python script if their life depended on it. I don't say this to criticize; I couldn't put together a corporate financial statement if my life depended on it either. Not everyone has the professional time (or desire) to learn to program.

But if you have this tool, Excel, that is critical to your job in many non-programming-y ways, but can manipulate data in ways that programmers would usually use code to accomplish, well... that's great, you use it, and are able to do your job better and more efficiently. And it flows naturally from the skills that you already have.


This is true, but doesn't answer the question of why not use another spreadsheet program? It's a minor learning curve to move to googles offering, and while it may not be as useful, "less useful alternatives exist" is not a viable monopoly case, is it?


There are entire multi-step multi-document workflows all over the corporate world, which are built on excel. You can’t just drop Gnumeric in and keep going. The investment in it is incredibly large and the cost of switching to another system far too high to be feasible—and if any one company does it they’ll still need excel because it’s how they interface with the rest of the world.


"It's hard to use the alternative" still doesn't seem anti-competitive


Google Sheets is a toy. When I've tried to do complex analysis on large data sets it just completely chokes and hangs part way through recalculating. Excel is far faster and more robust.

Google Sheets is also just missing some more advanced data analysis and charting features. Casual users might never notice but power users consider it unacceptable.


It is great to note lists and small monthly expenses accounting, and that is about it.

Nothing that compares to the math functions charting and table manipulations, extensibility via COM and .NET, PowerQuery, Lambda,...


Microsoft Office has been installed on every computer I've used professionally.


>why not use another spreadsheet program?

This is HN so everyone will focus on the functions/features. While that is important its not the reason for not switching. The friction is much deeper.

At one point the business world went all in on MS Excel. Its not just that they use it, the entire management organization and feedback loops are based on Excel not the other way around. Moving away from excel is not a function choice anymore. Its a Executive leadership company structure decision. I'm guessing not many fortune 500 CEOS post on HN to confirm this.

No one can "beat" excel because to do so requires simply copying it pixel by pixel and function by function. If you did that MS would sue you out of existence.


You could totally put together a corporate financial statement if your life depended on it.


How quickly can you connect to just about any arbitrary data source, pull a whole pile of data. Do some analytics on it and form it into a few glossy 8.5x11 for the next quarterly review. You're looking at about 30-60 minutes for an experienced Excel user. How about making it update in near realtime and adding some interaction for manipulating the data?

More importantly, can you teach an analyst of some sort, who while being a power user is not a programmer to do it just as efficiently?

Doing math on a 2D grid of cells is the technology that Excel perfected in the 80s, the power of Excel is in the connectors, services, and interoperability that surround it. It's no small feat to have an application that can guide a beginner through grabbing 10k rows out of an Oracle database and putting it on a graph, while also having the power to allow experienced users the ability join arbitrary sources and construct models around the results then present it in a logical fashion.

I am very much not a fan of the current state of affairs, but unfortunately nothing does Office like Microsoft Office.


Ecosystem, Developers, and entrenched education/skills win the day.


This doesn't seem like anything that Microsoft is doing in a monopolistic way, though. I'm sure LibreOffice could do this if they wanted to. I'm sort of surprised they don't, given LibreOffice presumably has access to JDBC.


> This doesn't seem like anything that Microsoft is doing in a monopolistic way, though.

That's the thing that I was wondering about here. I don't really know this space very well, and still have bad feelings toward Microsoft for their behavior in the 80s and 90s, but is MS actually abusing their monopoly position here? I guess the article hinted at a few things; e.g. if you are an Office365 customer you have to be an Azure customer, and can't run it on AWS or GCP. But I didn't see a compelling case for how MS is using its Excel (or Office as a whole) dominance to actually harm consumers or competitors. Excel's features aren't magic and AFAIK don't require backroom deals to enable. Anyone can implement them, given a lot of time and hard work.

I think the main compelling part was that MS doesn't have an incentive to focus on security as much as they should, because people will keep using Office365 regardless, as there are no viable alternatives. But that doesn't seem like an anti-trust issue to me. That's fixable through legally-mandated fines for security incidents, fines that actually hurt MS significantly, not just token fines that are shrugged off as the cost of doing business. Make it significantly cheaper for MS to develop a better security posture, and they probably will do just that.


> I don't really know this space very well, and still have bad feelings toward Microsoft for their behavior in the 80s and 90s, but is MS actually abusing their monopoly position here?

It's not illegal to have a monopoly, but one thing that's illegal (theoretically anyway) us using your monopoly in one area to gain market share in another area.

But it is kind of counter-intuitive to say that the linchpin of the entire Microsoft juggernaut fundamentally rests on their amazingly functional spreadsheet (which it sort of sounds like constitutes an actually useable "no-code" platform). Supposing that were true, what would be the strategy for people trying to bring about "The Year of the Linux Desktop"? Try to get LibreOffice or Gnumeric up to parity? Or try to get documentation / education for Jupyter / Python up to parity?


If I had to point a finger at something I'd say it's probably how they bundle everything, market everything, have cultured decades of cult-like loyalty in business IT employees, and so on.

That is, aside from the bundling I'm not sure MS has done anything particularly illegal or immoral? Just incredibly good business sense.

And I say that as someone with a stubborn disdain for microsoft lol.


I have the same issue. I also don't like MS, mostly for Azure reasons, but companies are sleepwalking into lock-in, for the reason you state: IT people are mostly Windows/AD/Exchange-centric, and Azure looks a little more like those tools. They're not being forced there.

My worry is nebulous: Microsoft is somehow buying its way into everything - AI, OSS development, gaming, etc, and it feels insidious.

However, it doesn't feel illegal at all. Governments might fancy some cash, and it's their right to write some rules to get that cash, but I don't see how in practice Microsoft is doing much obviously wrong.


> the power of Excel is in the connectors, services, and interoperability that surround it.

Hell, Excel still has the absolute best text/csv import of any spreadsheet I’ve used.

Fixed versus delimited columns. Arbitrary delimiters. And best of all, split existing column on delimiter to create multiple columns.

None of these are that fancy of functions. And there’s no reason why every other spreadsheet couldn’t implement them. But they don’t.

Maybe because it’s not sexy. Maybe because of bias against spreadsheets. I don’t know. I just wish Excel competitors would add them.


The csv importer of Libre Office is fine. Excel gets the staying power because it's already in the contracts for all things Microsoft. I think Teams is the new powerhouse that makes the office suite indispensable in large offices.

At home environment Google Sheets works just fine.


Its amazing how many products have little moats like this. Things that are days to at most weeks away from adding to any product, but that no other product seems to care enough about to build.


> Doing math on a 2D grid of cells is the technology that Excel perfected in the 80s

Lotus 1-2-3 gets that specific credit, I think. But Excel is so much more than that (for good or ill).


Visicalc would have been the first. In the early years of the PC, Visicorp was bigger than Microsoft -- mostly from their sales on the Apple II. They never had much success on the PC platform, which is where Lotus ate their lunch.


Yeah, "first on micros" definitely goes to VisiCalc. But "perfected in the 80s" I definitely think goes to Lotus. And while Microsoft was a contemporary developer of spreadsheet software, Excel didn't become dominant until the 90s.


"Microsoft released the first version of Excel for the Macintosh on September 30, 1985, and the first Windows version was 2.05 (to synchronize with the Macintosh version 2.2) on November 19, 1987"


I’m sorry, did Microsoft really design Excel for the Mac first?


It is indeed true. Windows had only just been released at this point, and was still limited by running MS-DOS in something of a pseudo-multitasking mode. The Macintosh platform, however, was slightly more mature, having already been publicly available for a little over a year.

The majority of Microsoft's software was being written to target non-Microsoft platforms at this point, which started to change with their increasingly anti-competitive marketing techniques (such as the so-called 'AARD code'[1] in 1991).

[1]: https://en.wikipedia.org/wiki/AARD_code



As a side note to the sibling comments, Microsoft's spreadsheet product for DOS was named Multiplan: https://en.wikipedia.org/wiki/Multiplan .


Excel is ahead of LibreOffice and others by laps. Just take a *.csv file and try importing it into LibreOffice and into Excel (with Power Query) and see how many additional processing options you are offered to format, interpret and transform the data. Next to Excel, CSV importing in LibreOffice is very barebones and not much better than a primitive example found in a programming tutorial.

Screenshots tell the difference rather nicely:

https://ift.wiki.uib.no/images/7/71/Csv_import_libreoffice.p... (LibreOffice)

https://learn.microsoft.com/en-us/power-query/media/power-qu... (Excel)

And even there, Excel tucks away a ton of functionality behind tabs and submenus: https://learn.microsoft.com/en-us/power-query/power-query-ui

Most people who say that Google Docs or other alternatives are good enough, or that they can program what they need in Python, barely scratch the surface of what Excel offers out of the box with little effort once you've mastered its concepts and workflows. It's like doing version management with "final_report_draft_v2_final (copy 2).txt". Might work for most people, but git offers so much more for those who know how to use it. Excel is the git of the business world.


It has been a few years, but the last time I tried importing CSVs into both programs, LibreOffice had a much lower chance of mangling the data. Excel loves to be "helpful" which often resulted in turning lots of things that aren't dates into dates.


Same experience. If your main language has a decimal comma instead of decimal point, and you CSVs must use alternative delimiters, in my experience LibreOffice CSV importer is better.


Opening csv and importing data from csv are two completely different things in Excel


The filtering operations shown on the Excel screenshot seem like they might equally well be done after the import step, as in LibreOffice. That way, the user only has to maintain familiarity with navigation of a single GUI for filtering tasks.


> anything more complicated I'd rather do in a proper programming language.

Yes you can. Most people cannot or don't want to. Programming is just a skill used to reach an end and if your job description doesn't require it, you skip it.

Excel democratizes the data analysis better than any open source alternatives and programming languages. It is easier to use and relatively less buggy than all of the open source and proprietary alternatives. When one really needs programming, VBA is there and it provides a much shallower learning curve for the curious.

From a corporate point of view MS Office has unmatched integration with Windows, Active Directory, Sharepoint, SQLServer and many other programs. A huge amount of financial, management and engineering software tightly integrates with Microsoft software to provide functionality like automatic BOM dumps to Excel and then integrating that with manufacturing, currency conversion. The developers of such software are pretty content with it, especially due to long-term backwards compatibility MS provides for their APIs for all their products.


While I do think the likes of Jupyter use will become more common in time for a lot of what Excel does, I do agree. It's far ahead of alternatives... it isn't any one feature, it's the sum of all the one feature(s) used across a given org.

A lot of it is entropy and entrenched knowledge.


> What makes Excel really so indispensable?

Not having to train all your accountants on something which is similar, but different enough they are less effective until they know it as well.

The integration is another piece, as the article mentioned - the MS ecosystem makes making all documents cloud documents and sharable and collaborative within your enterprise a total snap.

I'm not aware of actual core functional pieces of excel which matter to most users that you can't get elsewhere.


> the MS ecosystem makes making all documents cloud documents and sharable and collaborative within your enterprise a total snap.

Same for Google Sheets, they're fairly interoperable with the MS formats too.


I know for a fact that trying to do some of the integrations Excel can is painful by comparison.

Also, Google Docs doesn't come close to Word. There are a lot of little features missing and each is essential to someone.


Excel has some nice integrations with PowerBI, SQL, Azure and a rich SDK.


Also, that it has nearly everything anyone could need in single product. Maybe outside some very specialised analysts. So one tool works for most people.

And then there is interoperability, everyone is using it and everyone accepts it as document format.


The most valuable feature of excel is that its available. Office is assumed to be installed in most large enterprises. In organizations that otherwise love long linear processes with many gates, excel is the closest thing many employees get to programming without having to constantly justify their use case.


Addons are one part of it. At my current job at a financial services company, there are plugins to access the firm's internal analytics library, the risk system, as well as to integrate with the front end trading systems and external vendors such as Bloomberg.


>> Anything I want to do in a spreadsheet, I can do in Gnumeric, Libre Office, or even Numbers; anything more complicated I'd rather do in a proper programming language. What makes Excel really so indispensable?

Are you a vim user. Great I want you do that in emacs, or an ide, or vs code or...

You are looking at the problem at the wrong level. Excel is an IDE with a built in programing language for array/set based processing (it a matrix but hard to work with in that frame). Even if it looks 90% the same that last 10 is a huge change for power users of the system. Those power users (10x accountants and analysts) are going to fight you. The organization is going to fire you when you kill their productivity.


Yeah, that mirrors my experience pretty neatly. I worked for a small company that at one point wanted to save the licensing costs and had me install OpenOffice on every machine for a 6 month trial run. No one issue that anyone had was particularly insurmountable * . But it was death by a thousand paper cuts. Every single person had a different part of their personal workflow disrupted, some requiring work arounds, others requiring learning new tools. But at the end of 6 months it was clear this wasn't going to work, it broke too many things in too many unpredictable ways, and made the little outside world compatibility they did need much more difficult.

* There was one insurmountable problem. At the time OpenOffice still had a 65k row limit that Excel had long since moved past. There was a 5 year old bug ticket opened to fix it, but the developers in that thread were still having debates insisting the users were wrong and that 65k ought to be enough for anyone. You should use a real database if you need more data they insisted. Needless to say, the first time a vendor shipped us a 100k spreadsheet that we couldn't open, the writing was on the wall even if no one else had had other issues.


Just this morning I made a mental note of Power BIs 1 million row limit in certain circumstances. I don't have any immediate plans to pull a million rows right now but I could see it happening. 65k would be easy to pull off, a year of some poorly formatted data blew through that recently.

(I did push 10 million rows but that was by accident, some Power Query code got a little greedy)


I imagine its the same reason why cobol & fortran is still around, billions of lines of hard to understand business critical programs that work which nobody really wants to take a chance on replacing.


Poor vBA support in those alternatives, which is used more than you think.


MS is seeing their own issues shifting to JS/TS to support the online versions better.


> I can do in Gnumeric, Libre Office, or even Numbers

It's funny you list these when millions of SMEs are running on Google Spreadsheet.

I don't know about Gnumeric but I think the following guesstimate is not far from the truth: 99.9% of all Excel users are using not even 1% of Excel. Most Excel users can be "switched" to Google Spreadsheet and won't even notice any difference.


I also think this is only true because a generation of office workers have been trained to believe it is true. We don't have a better option because it is impossible for an alternative to be established. Excel begets Excel.

Further lock-in like VBA on Excel desktop, the crippled capability of the web version, and second class support on platforms other than Windows should not be taken as virtues of Excel. Forcing everything into a "document" paradigm sucks. Needing the editor to view the data is awful. There is no separation of concerns between content and presentation. Opening a file in the state the last person edited it in is horrible.

Conceded it is a powerful application, but what I see it used for mostly us visual grid layouts using cells, rather than true tabular data. I constantly see PMs using it to painstakingly make gantt charts, when MS Project is available. It's one step away from oil and canvas.


Weird but genuine question: would it surprise you if it was true that most people would agree with that statement? (that there are no viable alternatives to Excel)

It seems obvious to me that it is true, from my particular slice of the world, so I find it surprising that anyone would find it unsurprising. (The reason is that I am aware that whole businesses/parts of businesses are built on Excel in a way that is not replaceable by programming languages, and which everyone would refuse to replace with anything besides Excel because it's best-in-class and integrated into everything already.)


Although Google Sheets, LibreOffice Calc, and iWork Numbers may suffice for internal communication within individuals or small organizations, file compatibility becomes an issue when exchanging files with external parties. Excel files are not guaranteed to open seamlessly in other spreadsheet software, making it safer to use Excel as the de facto standard. This applies to Word and PowerPoint as well. Furthermore, Microsoft's current offering of these applications through Microsoft 365 via the cloud naturally leads users to become vendor-locked into Azure.


There is nothing comparable to Microsoft Office.

99% of white collar employees globally know/have to know MS Office. Every government uses MS office. Every sector uses MS office.

Who doesn’t use Office? Startups, creatives, scientists.

What makes Excel indispensable? Think of it as the equivalent of the English language in business IT. You can speak with your peers in your language, sure. You need some sort of common understanding? You use the tool that everyone knows.


It's not about you it's about every feature needed by 100% of the users in your organization. I've seen some impressive integrations with excel and Azure Devops or a number of databases used by non-programmers.


Have you never worked for a big company? Business people use Excel - that's it. Nobody cares what you (surely a developer of some sort) rather use. There's just no way around it. I once worked for a health insurance company whose claims processing system was fed by...Excel spreadsheets for all of their offered plans. To be fair, it was nuts - but it worked for a very long time (to the tune of processing billions of dollars of claims per year.)


Those other tools have gotten much better over the years, but for power users Excel probably still has functionality the others do not.


Who said "Hell is other people?"

Anyways, that's your answer. I'm guessing you, like me, pretty much never have personal problems with your own data and stuff that only you work on; you keep backups and such and know about cross-platform things and so on.

We're the extreme minority. Most folks rely on what was sold to them, idea-wise or other. Since I've been doing more independent real-life IT work along with my IT teaching, I've learned to be less judgey -- and even though I know the tech up-and-down, I've learned it's infinitely harder to get a significant number of people to see things the way people like you and me do.




Join us for AI Startup School this June 16-17 in San Francisco!

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: