Hacker News new | past | comments | ask | show | jobs | submit login
Companies fret as costs soar for software subscriptions (2019) (ibj.com)
346 points by roxanneonhacker 24 days ago | hide | past | web | favorite | 445 comments



This article is painfully stupid. I don't mean wrong; it's actually intellectually insulting to the reader. "And we still don’t know where this is going." Spare us.

First of all, being able to build an entire company faster and cheaper than ever before by taking advantage of a massive ecosystem of SaaS tools isn't just worth millions, it's the only reason many companies can get off the ground.

Second, SaaS tools have displaced entire categories of information workers that you no longer need to have in your org chart. Remember DBAs? There's an entire generation of entrepreneurs that do not.

Third, do you know how much it costs to deploy SAP into a company and keep it running for five years? Roughly $10M. https://blog.cfbs-us.com/blog-1/how-much-is-erp That's a fuck of a lot of Twilio.

Fourth, and the reason this "journalism" is making me #ragequit, is that business types have been making literally billions of dollars on the backs of open source software for decades, and they pearl clutch when someone takes that free code, attaches customer service to it and tries to build a sustainable business.

The problem is not the supposedly rising costs of SaaS, it's the notion that hedge funds can use free software to legally steal without even pretending to contribute to the projects and infrastructures that make their operations possible.

If you ever encounter someone complaining that SaaS is cutting into their profit margins, please give them an earful for me.


Not sure I understand you.

From your post, it sounds like you have beef with people who complain about SaaS costs, when many SaaS products are arguably cheap, cover valuable functions like DBAs used to, and allows many companies to get off the ground in the first place? All while proponents of SaaS are fundamentally no different than companies like SAP or Hedge funds that are making billions via rent-seeking off open source software?

Yeah well, I think that's the same reason people complain about SaaS costs. I know some SaaS software my company uses that wouldn't cost us anywhere near what we pay for it, if we opted to build our own via open source tools and maintained it ourselves in house...

SaaS is effectively the same as outsourcing - you're paying a third-party service provider for software/IT functions. Companies complaining that SaaS is cutting into their profit margins, are no different than ones complaining that their IT department is cutting into their profit margins.

The matter is really about which is cheaper (to make in house vs buy). In some cases it's cheaper to outsource and in other cases it's not.


> Companies complaining that SaaS is cutting into their profit margins, are no different than ones complaining that their IT department is cutting into their profit margins.

You’re betraying a fundamental misunderstanding of how software deployments work for most businesses.

Theoretically, the choice is between an expensive SaaS product and its equivalent OSS product which would need to be deployed in house. What this scenario misses is that most businesses have really ineffective IT departments which absolutely do not have the ability to:

- deploy the OSS versions

- guarantee an SLO for the deployment

- provide tooling around the OSS deployment

Remember that here I’m talking about most businesses; specifically those outside the SV bubble.

So SaaS is fundamentally different from outsourcing. The promise of outsourcing is heavily skewed towards reducing costs and SLOs hardly figure into consideration. SaaS tools might be expensive but they guarantee some things. Predictability is hard to get in the software industry; SaaS businesses promise that. Their success reveals a deep desire in the market for predictability around software products. Unless you have a highly qualified SRE team I doubt you can promise the SLOs promised by SaaS vendors.


Totally agree regarding IT effectiveness.

Outside of the tech sector, many businesses choose a saas product because they want IT to have nothing to do with it.

Outsourced IT can be difficult to work with (under statement), nearly as if they’re not on the same team...


It's like this:

SaaS companies generally add extraordinary value to the software that they commercialize, both for the customer and for the software itself.

You are simply mistaken regarding the cost difference between SaaS cost and having to hire an expert in an open source package (I'm simplifying for the conversation). It's not just a question of instance sizing and data transfer, or even how much it costs to pay those humans to provide support. The real cost is in terms of liability and service level. Most corporations quite simply won't take on the risk of using "raw" open source because if something goes wrong, they need someone to blame. The alternative in a real company would be to use the expensive, generic proprietary solution. They are happy to provide an SLA for $$$. Meanwhile, SaaS companies are frequently started by the core developers of open source projects, and the attention and reputation of having a company exist to promote your project tends to create a virtuous cycle of new users, training materials and PRs. It quite literally raises the tide for all boats.

I don't have any beef with the specific act of complaining about prices in general. People are indeed free to not appreciate all sorts of things. However, much like the rise of conscious consumerism, good managers and good companies - the sort of people and places you want to work for when given a choice - increasingly understand that you have to reinvest back into the infrastructure that makes our lifestyle possible.

So yeah, go ahead and install all of the locally hosted solutions you want, and feel as superior as you can, until the moment where you realize that it's utterly irresponsible for key tech talent to spend time in a growth stage startup applying patches to things in the middle of the night. If you don't figure it out, your board of directors will replace you with someone wiser.


> Second, SaaS tools have displaced entire categories of information workers that you no longer need to have in your org chart. Remember DBAs? There's an entire generation of entrepreneurs that do not.

This is a very good point, but if you use several of these SaaS, you're paying each one for their information workers. This is the same problem that makes expensive to have many streaming services for your home.


You're entirely missing our point(s).

When I send an SMS with Twilio, they a) aren't going to send me a bill for the cost of employing an engineer for a year and b) are a reliable vendor I can take to my board.

To refute your example, I could send messages via three services and still be ahead because I'm paying for what I use.


At my last job we turned a $1M SaaS implementation into over $5M in incremental revenue inside of two years with a team of just 4 people and offloaded an external vendor that was costing an additional $1M/yr. At a company that size (yearly revenue sub $1B) that was a significant return on our investment.


I don't want to be that guy, but since I actually want to understand what you're trying to say: could you proof this and say it differently? It's clear that you think you're making a coherent point, but for those of us not in your head, you aren't differentiating between subjects so it's not clear if you love SaaS, hate SaaS, made 4M, 3M or "sub $1B".


Look, either it's worth it or it's not. Costs may be soaring, but companies wouldn't be buying them if they weren't providing more value than they cost. But:

> enterprise companies are spending an average of more than $10,000 per employee per year on software

Sorry but I can't wrap my head around this. Is that really accurate? Between Office and cloud and HR/payroll and a few other tools I can imagine $1,000/year. And certain employees need SalesForce or Adobe CC or whatever which adds another, say, $500-2,000.

But I just cannot possibly imagine how an average employee is using $10K worth of subscription services?! Does anyone else think this is accurate?


That number doesn't surprise me. Your quick $1000-$2000 estimate is probably accurate for a typical office worker, but the more specialized positions skew the average. I work in mechanical/aerospace engineering for a large company. A lot of the software we use is very specialized and very costly. Here's the yearly costs from the top 4 softwares I use based on a quick google search:

1) $35,000

2) $15,000

3) $10,000

4) $4,500

So I'm costing the company ~$65,000 per year in software licenses alone.


Even for a typical office worker, based on my experience at a large company, there might be 10 to 20 enterprise apps that include:

  * Tracking work hours
  * Communicating within teams
  * Entering and managing yearly performance reviews
  * Managing IT tickets
  * Catching spyware, phishing, etc.
  * Key and click logging
  * Personality and other screenings for job candidates
  * Health insurance management
  * Jira being used outside of software development
  * Database for archiving HR announcements
  * Password management for the above
In general, the prosperity of a business is a balance of assets and liabilities, and a software purchase is both, since it involves a cost. Managers promote purchases based on a variety of factors relative to the business and their own personal ambitions. There's no magic that makes managers so brilliant in the face of a salesman that they only make beneficial decisions, or that purchases are inherently beneficial, just because the product is software.


> * Key and click logging

What kind of dystopian Corporate hellscapes have you worked in?

This is totally unacceptable. I would not submit myself to that kind of scrutiny or micromanagement, no matter how benign the intentions of it's implementor. Nothing good can come of it.


>> I would not submit myself to that kind of scrutiny or micromanagement

In many cases, you may not even realize it is running.


That just makes it super extra illegal.


IANAL, but my understanding is: - lots of allowances are declared by employers in their handbooks in the US. - unless there are specific union restrictions and/or state provisions, it is probably not illegal


OK, so the USA is mostly a dystopian corporate hellscape. Not that surprising, I guess.


Some people do. I still blame the human for the software we have today.


> I would not submit myself to that kind of scrutiny or micromanagement,

Says the gmail, google, facebook, android user.


Keylogging is illegal almost everywhere.

If you suspect that your current or former employer is monitoring people's keystrokes, please contact the appropriate authorities to start a whistleblower process.


Keylogging of any company provided hardware is probably legal as you have no expectation of privacy.


Fortunately, much of the world is more enlightened than this, and its legal systems do favour the privacy of employees over the voyeuristic ambitions of bad management under normal circumstances. In many places, the idea that you have no entitlement to personal privacy just because you are working for someone else would be widely derided.

There are often special cases in the laws for when covert monitoring is genuinely justified, for example when there is a reasonable suspicion of wrong-doing by an employee or there are serious security concerns that legitimately require stronger measures. But intrusive surveillance by default, even of employees using company resources on company time, is not permitted in these jurisdictions.


Beautifully said.

I would only add that in the extraordinary case of a legal investigation, there's still no room for vigilante spying by the employer themselves. Presumably, law enforcement would take necessary steps to intervene, not the employer.



Maybe in your jurisdiction, but certainly not here in Germany.


I do expect privacy, and find the idea of a middle manager or hr person digging through a employee keylogger output file revolting. It's a truly degenerate state of employer-employee relationship if this is even conceivable.


Ladies and Gentlemen no fighting!!

Why bother logging keystrokes and clicks, when there is BlueCoat that in real time decrypts, reads, logs/reports, encrypts EVERYTHING that comes in and out of your PC?


Key and click logging?!?!?


Note, this list is based on things I've read about, but I don't actually know what specific security software is typically used by a large employer.


It is a good list, I would add: * Office software (Office365 or G Suite) * Record Management and Workflow * Regulatory Reporting Management * Contract Management * Financial ledger management * Data analytics / business intelligence * Logging and monitoring, log retention * Cloud hosting * Vendor Management * Travel Planning and Expense Management * Cloud Telephony * Mobile Device Management


But software in highly-specialized industries (i.e. needs a lot of functionality but the install base just isn't that big) has always been really expensive. This is pretty much unrelated to the SaaS boom of the past 10-15 years.

I don't think specialized positions have skewed the change much towards increasing SaaS costs.


But there would have to be 5 other workers for every you. I think it’s more likely that the number is around 1 in 200 for most enterprises. And maybe closer to 1 in 1000 for all enterprises combined.

Note: those numbers are pulled out of my ass.


CATIA is a beast of a bill!


And the licensing is not perpetual. Once you stop paying the annual licensing, you are legally obligated to stop sing the software.


That's per license, per employee? Or company-wide license fee? If the latter, you still need to divide by the total number of users of the product.


if GP works in aerospace, those are very likely per-seat license costs.


Completely out of curiousity -- what software costs $35k per seat per year??


VxWorks is about that I believe. When you reach these kinds of price tags though it's hard to evaluate them on a per seat basis, because they tend to come with built-in consultants, etc.

CryEngine 3 is notoriously $1.2 million, but I believe that's a "total cost" rather than a per seat style license.


Lots of engineering software does. FPGA toolchains, CAD, simulation. Not uncommon.


Having worked in several related areas, but not very recently, the costs for specialist software packages were certainly much higher than people think of for basic office software, but nowhere near the scale we're talking about here. Have things truly become so crazy that five-figure annual fees for these packages are the norm? A permanent licence used to cost less per user than that! But then I suppose this is the point the article is making.


Several years ago my university was bragging about a $200 million donation the engineering school had received from several companies.

Turned out it was just several hundred seat licenses for a bunch of commercial cad, simulation etc. software, which the companies valued at about $50k/seat.

A very cheap way to donate $200 million.


And to write-off $200.00 million.


it is. And especially for SMBs, the cost was way less in the past. Imagine a small mechanical engineering company: 10 years ago they'd have 5 seats of a 5 year old AutoCAD/ProEngineer/Solidworks version, sneakily locking them in but basically no recurring costs. If a new customer needed a new version, they bought one seat for that and kept the old ones, gradually upgrading. Nowadays they pay half the single-seat cost per year because of "continouus updates" (for the UI mainly, so that they can sell trainings)


I heard about 20k per seat for CATIA at a previous employer. 35k wouldn't surprise me if you start adding add-ons (simulation, PDM, etc).


Enterprise optimization software costs that. Cplex, gurobi, Xpress.


That is per seat license.


Do you mean that no one else at your company uses any of those 4 softwares?


Not if they're per-seat licenses.


Then someone gets Slack because they don't like Teams, and they think that's worth it. Someone else wants Flowdock. Your new exec used Zoom at his old job and wants everyone to have that for video calling. Your IT security team have half a dozen anti-virus tools, two data loss prevention tools, three monitoring solutions because they're overwhelmed by the amount of data coming out of them. You use a third party RAS solution that doesn't work as well as the one built into Windows, but boy does it add a chunk per-user. You're still running a few JVM applications which you now owe Oracle for the pleasure of using an old JRE. No-one uses the three knowledge management tools you've bought over the years to try and organize your documentation, but you can't decommission any of them yet, because you never migrated everything out of them.

Then someone in the finance team decides the way to tackle out-of-control costs is to buy a tool to monitor installed software on all the desktops. It has another agent.

Welcome to the swamp.


You speak the truth!


We - at Okta - publish a report on this every year. Here's the January 2020 study: https://www.okta.com/businesses-at-work/2020

A key finding:

"Over the years, we’ve seen app adoption steadily rise, across all industries and company sizes. This year, the average number of apps per customer reached 88: that’s a 6% increase from 83 apps a year ago, and a 21% increase from 72 apps three years ago. And now 10% of our customers deploy a substantial 200 apps or more. When we look at customers that have been with Okta for more than four years, they deploy an average of 190 apps."

Odds are, a good chunk of those are free but if the average subscription cost is $10/month => $10/month * 12 months * 88 apps => $10560/year aka well within the estimate cited.


Of course this also assumes every app is licensed to every employee.


> but companies wouldn't be buying them if they weren't providing more value than they cost

Companies do stupid shit all the time. It only takes one person with authority to spend money to sign up for a service. Personally I've seen many software subscriptions that are complete garbage.


“We’ve seen companies buy the same subscription over and over again—even within the same department,” he added.

I think that's telling


One of the worst things I've seen at an employer was more than one SSO. That's when you know the people paying your wages are truly dysfunctional. Because of course Fiefdom A doesn't want Fiefdom B's SSO to be used by their systems, that gives the person ruling Fiefdom B power - so instead get a separate SSO, what do you mean that's missing the point?

If I worked for an SSO provider I think I'd be tempted to expontentially scale corporate pricing. Want just one SSO system for (say) 10 000 employees? That'll be $5000 pa. Oh you want a second one at the same company? $20k pa please. Third one? $100k.

My reasoning would be: An outfit which buys more than one is completely irrational and so they are my best target to gouge as heavily as possible because they'll be least able to do anything about it.


I’ve mainly seen that situation in post M&A environments, and while fiefdoms are often a part of the problem, untangling auth across multiple enterprise systems is really hard.


The fiefdoms I'm thinking of were created naturally inside the organisation, not acquired. Somebody's power grows, a department with ten people becomes a business unit with a thousand people due to its success, other managers resent that, and senior management mistakes this as an opportunity for "internal competition" rather than a dysfunction which needs to be crushed, if necessary by terminating those who promote it.

Lots of things are hard about M&A but it's made easier if people remember from the outset that their business exists in the world. You don't have and can't build a business which functions independently from the world, stop trying. On the technology side that means things like using FQDNs in URLs (https://webinar.mycorp.example/2020-New-Products not https://webinar/2020-New-Products)


But the GP's pricing scheme would put a hard incentive on fixing it.


It’s not so surprising, if the owners of the original subscription prevent access to new people because ‘their costs will rise’.

I’m planning to request a new subscription for exactly that reason.


The benefits and pitfalls of multiple payment approvals


And don’t forget the vendor lock in.

why change what works and redo all the work done in that system. It would require a huge expense and time cost to port to another platform.

And there’s the fact that longterm employees and more importantly managers are likely familiar with the old software and will see little reason to even test the waters with something else.

And that’s how programs like emblem stay around.


It's probably total SaaS spend divided by headcount. There are lots of bits you may not interface with directly, e.g. backup & compliance, infrastructure monitoring, advertising. And as another comment suggested, certain specialized applications with sky-high license fees.


I used to work for a company whose licensing fees were mid-five figures a seat. Granted, not every person in the company needed a license; but nearly everyone involved in the production process did. The company also benefited from vendor lock-in, because skilled workers tend to get pretty attached to their workflows and competitors in niche software tend not to have 1-to-1 feature overlaps.

It seems like nearly every profession has high-priced software that's required: artists, engineers, accountants, lawyers, doctors/nurses, research scientists, architects, etc, are buying products from Adobe, Quicken, West Law, Reed Elsevier, Epic. None of those are small companies, yet all of them focus in somewhat niche information technology sectors. I'm struggling to come up with a profession that doesn't touch software anymore; even retail workers use point of sale and inventory software.

I do agree with your sentiment though. The software is expensive because it's so niche. The market for all of this stuff is pretty limited, meaning there's only a few products offered, and the capital required to build competitive products makes it prohibitive for a company to build their own. The companies making the software also understand their target industries intimately enough to know exactly what price the market will bare.

However, it does leave an interesting market opportunity for and engineer + domain expert to tap into the market with cheap, low-feature offerings. Or maybe we will see more companies bringing development back in-house.


Software isn't expensive because it's niche.

It's about case and effect: seat costs for specialised tools are high because they can be. People with limited B2B experience underestimate just how much cash large corporates can throw around without even noticing, and how rarely ROI is assessed on this spending.

The development costs required to create a competitor to an industry standard product can be relatively small. (Sometimes they aren't - but they can be.) The real costs are in marketing and sales.

Quicken doesn't do anything outstandingly clever, but good luck trying to sell your new competitor to existing accountants.

Generally there's only a change if the incumbent gets too greedy and complacent, or if the industry changes significantly and a new generation of companies appears with new ideas about which software is "standard."


Qualtrics, a company I have worked for in the past, specializes in selling huge experience management packages to global enterprises. Deal size ranges anywhere from $10k to $1mil+ (yearly). Then there's SAP which is something like $3k per user per year. These things can seriously add up.


SAP might be worth it, but my first experience with it has been absolutely awful.

Some buttons are hidden behind 3 layers of scrollbars because they’ve put panes inside of panes inside of panes. It’s ridiculous...


What about VMware, backup software, system monitoring software, update software, remote support and meeting apps. Spam filter, Terminal server licensing. There’s more I’m forgetting


Why would any of them be per employee?


Why wouldn't it be? The size of your infrastructure depends on the amount of employees you have.


per-user licensing.


Really every user needs a VMware VM and terminal server, and monitoring license? Maybe they do, but they better do something that makes a lot of money to cover the costs.


I think it may depend on the company and on how these "averages" are calculated.

For a company where almost all the workforce is white collar then yes, $10K seems reasonable. In fact, I'm surprised it's not more.

For a company like Wal-Mart, I'd be shocked to see that much spend on software subscription.


Well, it depends. Only considering Saas, i have a hard time to imagine that as well. If include highly specialized stuff, Catia and so on, SAP etcc.., and devide that by total number of users (not employees) the picture changes.

I am currently at roughly 40 per user. That includes MS Office 365, three mail accounts under Office 365 for the TMS / WMS suite, Wordpress and Slidebean. Hope I didn't forget nything.

On the other hand so, Catia liscences are usually part of the service (for engineering corps.) or part of the product development costs (for manufacturing corps). in both cases costs are covered by customers. Especially in Aerospace these costs are the least f the cost problems.


In addition to that, some equivalent of Gsuite, Jira, Confluence, Trello, Slack, Figma, Expensify, Github, Looker, Tripactions, Greenhouse, and Jenkins all feel like table stakes for an enterprise firm these days.


Yeah, those numbers seem very suspicious. Cherry-picked maybe.

Fifty dollars a month is a decently pricey subscription (like a lot more than Slack, JIRA, Lucidchart, Office, etc.), and you have twenty of those for every employee?


I've used specialized embedded system tools that cost $10,000/year per seat for tooling. Our QNX licenses were about $5,000 or $10,000/year alone. That's just to be able to run a piece of software. Autodesk tooling is $20 to 30k a year. Some of the more specialized tools cost six-figures.

So I guess if you're just doing some basic web development those numbers are suspect. But if you do hardware or embedded software projects $20,000/year per person is cheap.


For such products, there's the per seat license and the per device license. Depending on volume and vendor pricing model more of the cost will either be in the seat or device license. The value of these the seat license is so that you can distribute their software on your devices versus the value of the seat license of Adobe CC is so that you can use the software directly.


I agree, but how much employees at an enterprise have the software?

Like 5%.


Maybe 1% of enterprises do hardware or embedded development.


You can pick a industrial vertical and find analogues though. Aerospace, automative, chemical, materials, biosciences. And even within a company each departments like finance, marketing, supply chain, etc have their own specialized software. Inventory management isn't cheap. Marketing software like marketo isn't cheap. ERP systems for the finance department aren't cheap. Accounting and tax software isn't cheap.


We used Epicore Vantage ERP for manufacturing. The starting cost per seat was $175 per user per month. With many added modules it can easily reach twice that per user. Our $30 million a year company was spending about $150K per year licencing this product, and this was only one of our software packages. I think the $10K number is reality for many companies.


It think the point of the article is that with SaaS, the customer invariably end up spending more than she really wanted vs purchased locally-hosted software.

Which is a big reason why there is so much SaaS in the first place, in both the consumer and professional world: the customer _forgets_ about this recurring cost.


> companies wouldn't be buying them if they weren't providing more value than they cost

At first, but value may decrease in time and companies will still be dependent because it may not be possible to take the data out, logistics, contracts, bureaucracy, etc...


I'm with SonicScrub. In my last job, my yearly licenses for me and my group of about 6 were $1 million annually not including hardware. We likely heavily skewed the average workers costs since we used extremely expensive, specialized software


MS office is probably one of them.

Office 365 isn’t worth it but most people don’t seem to know that.


For whom is it not worth it? Software devs? Are they representative of the average white collar?

I couldn’t live without excel.


Sheets will fit a lot of excel use cases for those not doing serious heavy lifting with it.

FWIW, I'm a software engineer and our IT department actually let people opt out of Microsoft office if they weren't using it to save money, and encouraging people to use Google Docs/Sheets (as most do anyways). I'm guessing a good deal of software engineers did that but there are plenty of non-engineers that probably also did outside of finance / some sales people.


I am a big fan of sheets, but it is not Excel. Some of the functions are missing after all these years, not to mention charts and pivots.

I think Google gave up on trying to compete with excel on features.


The overwhelming majority of people who double click the Excel icon would be equally well served by sheets or localc. It's fine if you're an analyst and are more productive with Excel, but if you're the rest of us, it doesn't matter what spreadsheet app you use. For most of us, a spreadsheet is a spreadsheet and don't need anything more complicated than sum().

In most companies, it's silly to have Office installed by default for all employees. If you need it, fine, put in a request and we'll get it done.

(please note there are a few things I'm NOT saying: 1) localc or sheets are as good as Excel for all users. 2) companies should save money by never paying for Excel, even if an employee wants to use Excel. 3) all companies should follow the rule of thumb outlined in my second paragraph.)


You’re right for creating. But many people read spreadsheets that are created by someone who needs those functions. If only 1% of spreadsheets use that function, the entire organization needs it. Training users how to double click icons is surprisingly hard. Trying to train them how to open some spreadsheets in Sheets and some Excel is a support nightmare.

Maybe I could see an argument for localc, but sheets is actually pretty costly at $50-150/year.

I think it’s fine for non-business people to just use the free sheets, but an org would have to not care about excel to use sheets for everyone.

With 365, it’s sort of a no-brainer as while it’s not as good as sheets for accessibility, if I’m going to pay $100/year/user then it’s easier to plan for 365, than sheets, I think.


> Maybe I could see an argument for localc, but sheets is actually pretty costly at $50-150/year.

Where are you getting that from? I see Enterprise GSuite at $25 a user (max) and that includes mail and drive. The cost of sheets here is probably $10 or less.

It does seem like the cost is in line, but if you're already paying for Gmail/Drive or other Gsuite things then saving the office 365 cost absolutely helps and Sheets is essentially a no-cost add-on.

https://gsuite.google.com/pricing.html https://products.office.com/en-us/compare-all-microsoft-offi...


i think you are severely underestimating the amount of people whose entire job is excel


Sheets has had pivots for years. It's become very rare that Sheets doesn't have a feature I want, whereas it's become more common that I'm in Excel and wish I could use Sheets' query function.


Pivots exist, but aren’t nearly as functional.


Sheets has charts and pivot tables. It definitely lacks some features of excel, but it also has some nice new additions that I love so far - e.g. Showing the return value of a formula while still editing.


Sorry, I didn’t mean that sheets doesn’t have charts. I meant that the charts in sheets are missing substantial features that have been in excel for years.


Meh, you should probably be using either python or an RDBMS. Most non trivial applications of excel come from a lack of IS knowledge IMO. It’s easy to do but it makes everyone around you less happy.


I disagree as both a python and rdbms user. Excel is great for millions of one off use cases by users who don’t know how to program and don’t know sql and/or have access to rdbms infrastructure. And I think that’s ok.

If I want to create a reproducible data process then I’ll build it in python and maybe some data share or database.

But tons of work can be done in excel that isn’t worth the cost to build in something else.

I don’t include Macros, if I’m programming in Excel, I should probably do that in something that is better designed.


I use spreadsheet programs all the time during software development. it's a great way to quickly look at data or do some calculations.

I use Numbers. Is it as good as Excel, not really, but it's good enough (although data import is still a bit lacking), and it's not a subscription.


I have a data import task I run about once a week - I've found Numbers does a better job not screwing the date field up I add than Excel does when I export to CSV (for import into our relational database)


Office 365 is a bit more than some office apps though, you have teams, OneDrive, planner, forms, Todo, powerapps, stream and so on... It actually replaces a whole stack of SAAS apps most companies are already shelling out for.


Shutter.. hate the msft sticky apps. They are giveaways simply to not lose market share. Can someone please make a free version of sharepoint already so I can be productive.


All companies I've worked at had dev environments that did not require any licensing. The tools used where mainly free or under the free tier (sublime, eclipse, CurcleCI, AWS...)

For larger products those will require to purchase certain usage since the workload will be higher but that's a company product cost and not per dev.

Unless one is developing MS products or is dependent on MS software I can't see how the numbers mentioned add up


Right but MS has such a stranglehold that no business dare run without office. It's a shame that no one really has competition for them


I know plenty of companies that are on Google apps.


Even with that they're probably a still running windows

once you start talking about SAP and other enterprise software you are talking big $$


and even if it is worth it, Why aren't they switching to equally good subscriptions that are cheaper? that's the real question.


> Between Office and cloud and HR/payroll and a few other tools I can imagine $1,000/year.

Really? You can imagine only #83/mo in software costs for a working professional? I think you're being really rather unimaginative.


I'm not trying to be super-imaginative, I'm trying to be realistic. That's the point.

Considering Office 365 is $10/mo (before volume discounts) and is probably the main tool most office workers use, I am having a hard time how on earth you get to $10K/12=$833/mo.

I'm supposed to believe the average office worker uses a total of subscription software that is eighty times more expensive than Microsoft Office?


Accounting software, payroll software, financial planning software, supply chain management/inventory, health insurance/401k/benefits/PTO software, JIRA/Ticketing, knowledge management (sharepoint/internal wiki/confluence), data warehouse licenses, reporting software (tableau, looker), IT and device management software, expense and reimbursement, business travel scheduling, performance management, contract management, backups and disaster planning, etc. And then any specialized software for actually doing different jobs roles (CAD, photo/video editing (think of the marketing and design departments of any F1000 company), IDE's for the developers, operations software for running a warehouse or factory, etc.


You've listed a ton of tools, but not every office employee uses those, and I hope that payroll software isn't costing $1000 per employee/year.

To get to those large numbers, you'll need to hand-pick companies that have few non-specialized employees.


Lots of the SaaS products we use mean we don't have to hire more employees. So I'm not sure if dollars per employee is a meaningful stat.


For a lot of companies I think it's actually the key to being able to expand and hire in the first place. Most startups aren't choosing between "should I hire someone to do this, or buy a solution?"--they're choosing between "Can I do this by buying a solution or do we go without? (either this, or something else because we need to build this)"


Just an interesting idea, if no one wants to have employees (lets face it, they are a mess, they waste oxygen, they fart, producing CO2, are corona virus friendly, produce road accidents... and you need to pay them so they can live) maybe everyone is having wrong bussiness plan. Whole world should transform to produce only SAAS products for SAAS people as they are the only customers who will in long term still have money to buy SAAS. /s

This is becoming a joke.

In my whole life I didnt buy a one single software product, that is based on subscription model (unless necessary as there are ongoing costs for its development like antiviruses (which I dont use, but this is another story)). Why? As it tells me, the product is at the end of its inovative phase and has nothing more to offer. I am fine with paying for update hunderds of dollars if, and only if, i need the added functionality. Or in different words Microsoft Office 97 was more than enough for 99% of its users (I dare you to install it and find something that you miss, it wasn't perfect but it did get the job done).

But there is more. I wont pay for cloud products either. Why? I believe that what is giving you the "edge" is your knowlidge. Using, for instance, gmail is taking you the chance to increase it. Yeah sure, time to market, but if you stick to DIY logic, this is not going to be a problem (I can set up a mail server with left hand while coding with right hand... and offer it as a SAAS to those that ignore DIY and earn extra bucks ;) /s ). If you ignore the lack of knowlidge, you will end up always paying to someone else or stuck in a problem that wouldn't be a problem if you wouldn't be ignoring it for so long. Not to even go into lack of inovation and reinventing / copying the solutions that were solved long time ago, but this is just another story.


I like your first sentence, but unsarcastically. Companies should always strive to do more with fewer employees by automating them away.

As for the rest of your argument, most people are not like you, as they will not want to spend time setting up their own mail server. For companies this is doubly true, as they will likely spend more money setting it up, maintaining it, and upgrading it, than the subscription actually costs.


> Companies should always strive to do more with fewer employees by automating them away.

I wonder if those same companies intend to socialize the wealth generated by their automation.


Of course not, there is no incentive to do so. Until there is, nothing will change. I am all for socialization of wealth, as money in circulation boosts the economy more than hoarding it, but I just don't understand how people can say that companies should (morally) socialize their wealth "out of the goodness of their hearts" without some incentive. Companies, like all organizations of individual agents, do not have collective morals. They are merely machinations that have the sole fitness function of increasing wealth, at whatever cost. If we take this to be true, why exactly would, without another fitness function, a company socialize wealth?


I think we've reached peak Hacker News. This is now looking like Slashdot.


I'd actually prefer slashdot's moderation/meta moderation model to the one used by HN, though it's hard to fault the job dang and co. are doing with their hands-on approach.


Horizontal integration of business processes means a lot of external stakeholders have you by the balls.


This. Bunch of crappy apps which meet the minimum rqt to keep you in the ecosystem. Teams is the business equivalent of imessage. Keeps people on msft products vs. switching platforms. That plus finance folks cant not have their Excel for their 30-tab whatevers.


You've attached a sarcasm tag to your first paragraph, but I do think the market is actually going there. Notice how many products around you get replaced by services.

I don't like it at all.


Hmm, there are many companies where what you proposed make sense. But there are also many that I think it's hard to debate that SaaS doesn't offer value.

Running the mail server is simple enough, but then you need to set up storage. Emails are sensitive and often can't afford to lose. Then you need to set up backs, monitor and regularly test DR scenarios to see if your failovers, recovery plans are operational. Of course you need to maintain your data centers as well if you don't want subscriptions.

Then think of login. Employees don't want to log in to 15 different software. So they are going to ask for SSO. You are going to have to deploy and maintain something like active directory. Sometimes they need to be maintained across networks between different offices that are far from each other. After all this, you will still have to buy support packages from principle vendors for times when things go wrong (and they go wrong all the time when you host your own stuff). Shit escalates mate.


But this work scales better than linearly. One sysadmin, if doing their job properly, can configure a robust backup solution once and it will apply to many different programs. Machines in a DC can be set up once and host many programs.

Login is surely one of the best arguments for non-SaaS solutions. SSO is a total mess in the SaaS world. Everyone ends up with a billion passwords that they manage individually, change at different times, have to use a password manager to keep track of (but they all use different ones), etc. Active Directory is actually a solution to that which got lost in the move to SaaS-in-the-cloud.


It feels like the more meaningful measure would be the change in profit over the change in SaaS utilization.


No, total cost would be a better measure since there is no per-person cost to the SaaS provider.

If you have 10,000 employees that's $100M per year in expenses. If you're just renting payroll, purchasing, etc software for that price it'd be far cheaper to roll your own, or pay a few people to work on OSS full time.


> If you're just renting payroll, purchasing, etc software for that price it'd be far cheaper to roll your own, or pay a few people to work on OSS full time.

This is the kind of hubris I come to this site for.


Hubris is half of software development ;)


Is the other half the hole of legacy support you dig yourself into supporting the bad decisions and naive assumptions you made driving with that hubris?


Nah that's when you leave the company and leave some intern to clean it up so they can develop their own sense of hubris over your bad decisions. And the cycle repeats.


Read up on the history Chrysler Comprehensive Compensation System (C3). The team worked for 7 years and although they achieved some technical goals they never managed to deliver a working payroll system.

https://en.wikipedia.org/wiki/Chrysler_Comprehensive_Compens...

Payroll is harder than it looks due to all the edge cases. For example, it is literally possible for an employee to have so many payroll deductions including court-ordered wage garnishment that net pay would be negative. How do you handle that in all legal jurisdictions where the company operates?


> For example, it is literally possible for an employee to have so many payroll deductions including court-ordered wage garnishment that net pay would be negative. How do you handle that in all legal jurisdictions where the company operates?

By issuing a error and demanding that a human deal with it. Obviously.


Surely you jest? If that was a serious comment then you obviously have no understanding of the business and legal requirements for payroll systems.


If you're legally required do your payroll with a buggy computer system rather than manually (and not even "this specific buggy computer that we gave you", just any old buggy piece of crap), then, uh, I don't really know what to say to that. My condolences, I guess?

If you mean it's impractical and error-prone to do payroll manually, well that's the point of automating >99.999% of it. For the remaining <0.001%, someone is going to have decide how to deal with it, and I'm sceptical that a programmer seven years ago will know better than a lawyer today what the legal requirements today are.


The level of aggressive ignorance in HN comments can be quite shocking. There is no legal requirement to use any particular payroll system. The payroll rules are written by domain experts including lawyers and others, and encoded into rules engines.


Well, how do existing manual systems deal with it? Why not just make do with partial automation and fall back on the existing methods for those thorny legal edge cases?


Except for the smallest local businesses, there is literally no such thing as existing manual systems anymore. It's simply not possible to do manual payroll while maintaining compliance with laws, regulations, civil orders, labor contracts, and company policies. Every significant enterprise has either built up a custom automated system over decades, or has customized a commercial ERP payroll module, or outsourced the entire thing to a specialized payroll company.

You could argue that payroll shouldn't be so complex but that's the reality of it today.


Rolling your own payroll is about as good an idea as rolling your own crypto.


However, if you roll your own crypto payroll, the risk of one project perfectly offsets that of the other. It's a perfect hedge.


Ah, but what if they add rather than cancel?


Two negatives make a positive, right?


It depends whether payroll theft is part of your business model, or something you're trying to avoid.


One large company (170k employees) I worked at switched from an in house payroll system to an external month.

Second month was a disaster in the end they had to physically cut the tape on the last remaining partially destroyed copy - about 20% got paid on time.


Switching a large enterprise software system is almost as time consuming, expensive, and error prone as developing a replacement in-house. You would have experienced substantial problems even if it was in-house -> in-house-rewrite or external->in-house or external-a -> external-b.


If rolling your own crypto, you mean cryptocoin, that sounds like a great idea to increase wealth :)


Cryptography not cryptocurrency.


That was the joke.


We wanted to build our own payroll system, but engineering is too busy milking the cows to restock the fridge to learn the tax code for multiple countries.


Has anyone thought about asking engineering to implement or build an interface for someone else (business side?) to manage country level differences in taxes?


I would be interested in engineering something like this, my email is in my profile if you want to share more.


You know what costs way more than SaaS? Hiring, managing, and losing people. Leveraging high-quality people with tooling, both for individual functions and for company-wide process, is a no-brainer.


There is also the option of buying self-hosted products and installing them on your own server. The setup time is usually the same, but you only pay $5/mo for the server and maybe $100 one time for the product, instead of paying $100/mo for the service. This is also multiplied by X, where X is the number of subscriptions you have.


No. Please do not suggest this ever.

Every company I've worked at and that has gone with self-hosted products have lost more money than paying the subscription.

Self-hosted nearly never gets upgraded because there's no budget and it can always be cut from the budget because it'll keep functioning.

Corp IT who usually get stuck managing these become the gatekeepers. They have nearly zero incentive to upgrade because the risk is they take down the whole thing for longer than the maintenance window.

During this time the product innovates & none of this is usable until the upgrade. I've seen whole teams created to build out tooling on top of Jira only to find out if the company had spent $100k they could have upgraded and gotten the features.

Nobody knew said features existed because nobody had been exposed to the most recent version and we had a 4 year old version.


The problem is not the self-hosting aspect - it's not taking into consideration the lifecycle.

It's worth self-hosting if you have a good reason for it, and the software is old and boring (aka "battle tested") and packaged by someone else so the lifecycle is someone else's problem.

If it can be installed/maintained/upgraded by running `apt` or `yum`, and isn't a pile of custom hacks then go for it. Server software like Apache/Postgres/MySQL or even fairly nontrivial apps like Jenkins fall into this category. You can literally run these for years, staying updated, with an insignificant amount of maintenance.

Atlassian products with the "download this tarball and run a script" update method aren't as nice as the `apt-get upgrade` you get with the items mentioned above.


We self-host everything and have refused requests to update to subscription products. Our systems are our responsibility, so outsourcing to a SaaS that will die or get hacked tomorrow is enormous unquantifiable risk compared to the little opportunity costs of self-hosting.


On a less serious note, self-hosting also feels more like you do own your business, have full control over your products and are not just a merchant buying low and selling high.


I don't think either absolute is the answer. The decision to outsource anything often depends on a particular company's operational strengths/efficiencies.

Maybe it makes sense for a software company to self-host some software, but they might hire a paving company to pave their parking lot. Likewise, it might make sense for that paving company to pave their own parking lot, but use SaaS software.


I agree with your point but that's just astoundingly stupid to put a team of people on building your own features on top of Jira without first spending 5 minutes to look into whether your new features have already been developed in more recent versions.


Is the issue with staying on an old version security, or lack of new features?

Because if it's just features, then maybe the features actually aren't worth the cost of an upgrade, if no one feels incentivized to do so.

Your anecdote about Jira tooling feels like the company not doing adequate research before hiring a time. (In other words, they should have considered all options—what is the cost of upgrading versus building tooling?)


At the previous company I worked it turned out pretty well. We used Grafana (which is OpenSource) for our high volume in-app analytics and servers monitoring. It took a bit to implement all the graphs, but we had exactly the graphs we needed. Even if we went for a SaaS, we would still have had to implement those custom graphs in their interface.

We have also paid (a lot) for DeltaDNA to have more insights into monetization, but their platform was a lot slower than our self-hosted Grafana and they even forced us to use our own S3 buckets to store the data (set them up and link them) due to GDPR (they said). Not only that, but creating custom graphs in their platform was very tedious and a bad experience overall.

I am not saying it makes sense to always go self-hosted, you are right that depending on the application you want to host you might need less or more technical knowledge and maintenance.

I would never recommend (for the average individual/company) to self-host email servers, CDNs or Slack chat server alternatives, but I do think for some applications it's worth the time, for example static pages (instead of using a website builder get a theme and host it yourself) or analytics.


The cost of insourcing a SaaS type product is that, when it's broken, you have typically 1 person or small team; and that 1 person is typically responsible for other SaaS apps that may fail at the same time.

Setup time may be similar; but if something goes wrong, the cost is a loss of whatever other project that person is working on to fix it (+ the mental overhead of preparing to handle fixes)


But now you need to hire an in house IT team that can manage and install and update those servers and software


Just a small remark: I think usually not the X multiplier is the important factor in making the decision to go self-hosted, but the price, privacy, lack of features or poor performance or availability of the service.


This is usually a bad idea.

The times is a good idea is something like Confluence or JIRA where the functionality of the cloud hosted version is hobbled compared to the on-prem version. (In that case, due to plugin security.)


SaaS is a disaster without controls. They make it damn hard to automatically disable users who are no longer active. Then you have users picking competing duplicative tools, and now users can't work together without having duplicate accounts in each tool. Once they get you in, you've got no leverage--they'll just keep increasing prices because you're locked in. Companies are pushing you from perpetual license to cloud not for your benefit.


Name some examples of price increases that were unfair?

Price increases due to inflation are a necessity.

I can't see a world where I would buy a software once and use it forever, so paying for upgrades is necessary at some point.


There's a hell of a gap between "paying for upgrades is necessary at some point" and "my supplier can choose when I have to pay more".


At the startup I used to work at the MD spent quite a bit of time dodging calls from Cloudinary. Every few months we needed to be on a higher tier package. We'd generally be tasked with optimising whichever metric was causing the problem a coupe of times a year (deleting / consolidating versions of various images etc).

It felt pretty shady (both his behaviour and their pricing structures).


Some of us put accounts on pause if they've been inactive for 30 days. It's not the norm (yet), but the good PR and happy customers were well worth the extra business logic.


If you are already spending >$200K per employee, spending another $10K to make them even 10% more productive is a bargain.


Exactly this. Unfortunately, way too often I see execs saying "that's too expensive, we just have to do it manually" instead of using good tool.

Then what happens is somebody making $100+/hour would spend several hours to do the job that would be covered by a tool that costs under $50/month.


Then what happens is somebody making $100+/hour would spend several hours to do the job that would be covered by a tool that costs under $50/month.

It's never that simple, though, is it? Suppose that tool solves a problem that a $100/hour employee would have solved manually in a day. How long did it take the employee to identify and choose the tool, arrange the purchase, and then learn to solve the problem using it? Probably a few hours too. So that $50/month tool had better be useful for replacing that job several times in a year or it's unlikely to be a net win in terms of time and money. It's certainly possible that good tools are far better than that cost/benefit ratio and using them is easily justified, but I'm guessing that in these organisations with hundreds of different SaaS subscriptions only a few of them are in that class.

Of course the elephant in the room is that there used to be another alternative, which was buying a tool outright for say $1,000 and then using it indefinitely, which would be a financial win compared to an equivalent $50/month SaaS tool in well under two years. Yes, there are factors like CAPEX vs. OPEX to consider so this isn't so simple either, but ultimately paying much more money for the same thing is still paying much more money, however you slice it.


The other gain you get from having the employee build the tool is that now the employee is better at building these tools, so you fundamentally have a better employee than the one that had to spend the time learning the non-transferable one-off tool.


Which is valuable, if your company builds those tools... otherwise you just trained your employee to work at that SaaS vendor.


On the other side of that equation, what you would hope is that your $50/month is buying you a subscription to steadily increased functionality over time. That's something your $1000 up-front payment wouldn't get you, and it helps offset the initial on-boarding costs. It's also probably true that once your one employee has gone through the selection, learning, and purchasing process, getting subsequent users on-boarded is much quicker, because there's someone sat next to them to say "just do it this way".


On the other side of that equation, what you would hope is that your $50/month is buying you a subscription to steadily increased functionality over time.

Some people might hope for that. Personally, if I'm buying a tool for professional use, I want to choose my preferred one and then have it be reliable and future-proof. Things like security updates are one thing, but the last thing I want is random changes in functionality or UI being forced on me.

This worked just fine in the traditional model where you bought a version of some software, and then if there was another version released later with something more that you wanted and adequate compatibility, you bought the upgrade too. The user gets stability and new developments if they want them. The developer gets paid for new developments, though only if they are actually valuable to users. No-one gets forced into anything unexpected changing, or going missing, or breaking compatibility.

Tragically, the whole SaaS, insta-deploy anything we feel like culture that has evolved in recent years has utterly destroyed that stability and reliability. I regard this as possibly the biggest retrograde step in the history of personal computing (and that "possibly" is mostly because the walled garden culture that has also become so powerful in recent years might be a larger backward step).


I don't disagree: there's a reason I prefer Debian Stable on my servers. But when done well, and in the right place, it can be a worthwhile avenue. Github is a pretty good example: they've not stood still (ok, they also haven't moved that fast either), but the core of what worked 10 years ago still works today.


It's easy to say "yes" and difficult to say "no". Becoming a financial steward comes by first saying "no" and only saying "yes" when the ROI worksheet is filled out, completed, and approved. If you want your manager to approve spending some money on tools you can help them by filling out your own preliminary ROI worksheet to show them why you think it's a no-brainer.


Buying a $5/mo subscription by accident, forgetting about it for years, and getting 0 value in return is cheaper than training on a formal purchase system, filling out a ROI sheet, and ping-ponging it back and forth in a few meetings before "responsibly" deciding not to buy.

Giving individual contributors purchase authority and then supervising spend -- the model implemented in every major cloud -- is a much better compromise, because it lets management prioritize where they spend ROI calculation cycles which are not free. Very not free!


What if that SAAS is data mining you though and your company secrets are leaking out? Or claiming some sort of ownership over content you post within it? I don’t know anyone who reads the privacy policies and TOS of those apps but tech companies just let their teams use what they like.

It’s hard to find a good SAAS privacy policy actually, and they target businesses who are theoretically more “serious” than individuals


I am far too apathetic for that.


Someone should build a webpage that does this:

1. Enter salary of engineer.

2. Enter cost per engineer of product (subscription, one time, etc.).

3. Enter required minutes to make worth your time (setup per employee can be factored in).

This may already exist, but if it doesn't it could be a useful tool for Saas companies. Something you could drop into emails to potential customers ("see how much you would save").

Would pay for this monthly if it works.


Can the tool be used on itself?


So, a spreadsheet?


Yeah but built with React and deployed on a Gcloud kubernetes cluster.


It should also include "machine learning"


That would cost you $73 per month, at least!


Seeing as you asked, here is my attempt from an afternoon's work: https://isitworththecost.com/


Very nice job.


I have a spreadsheet at work that does this for manual tasks. It calculates minutes to perform the task X # employees X yearly frequency of task.

Once I've plugged in the values, I go back and ask if we really want to spend 30 hours a year updating intranet employee profiles or if we should use that time to fix some bugs or code a new feature.

Its quite effective to pushing back on unneeded tasks.


I've seen an app called Kanban Mail whose pricing page has such a simple calculator, although for individuals rather than companies,

https://kanbanmail.app/pricing


Should just be able to do that in your head.


It's not for you, it's for execs on the fence. Visually seeing potential savings would probably drive higher conversions, just a guess.


I don't think the problem is people at the executive level. It's that all kinds of teams across the org are paying for Saas products and it adds up over time. The other problem is that 15 teams might be paying for small group Slack (insert Saas here) licenses when they should be on a single corporate plan. Ironically I believe there is a startup that looks at accounting and located duplicate licenses.


So something like this: https://xkcd.com/1205/ ?


Now I'm wondering if I can automate my VPN login. Apparently I can spend 3-4 days on the automation and come out ahead!


Because most business in the United States is done despite the management, not because of it.


Where’s the list of awesome companies that succeed without managers?


Hard to find for sure. There was a time though when we didn't pay CEO's & professional managers 500x as much as the lowest paid employee. When we did that we still managed to go to the moon, split the atom, develop computers, create antibiotics, and lots of other neat things.


All of the examples you listed are either NASA or academic research, don't see what that has to do with CEO pay.


Do you have evidence for your assertion? Of what significance is the ratio of highest to lowest paid employee? Shareholders of a particular company pay CEOs, not whoever "we" is.

Here is evidence that claim CEOs produce more value than they are compensated for (i.e. they are underpaid): https://faculty.chicagobooth.edu/steven.kaplan/research/kceo... https://econpapers.repec.org/article/inmormnsc/v_3a60_3ay_3a... https://papers.ssrn.com/sol3/papers.cfm?abstract_id=1686068


Producing more value than you are compensated for ... hmmm what does that sound like? You could say it sounds like the human condition. Which employees in the pyramid do you think have the worst scalar here?


The CEO's, it's pretty clear.


This may be true for CEO’s turning dysfunctional companies around on personal authority, but I find it hard to believe for companies that are already running perfectly well.


I have yet to encounter a manager who motivates me to get up and find a way to eat that day.

Just because they are there doesn’t mean the achievements are due to them.

There have been a number of attempts, tens of millions of dollars spent on one study alone, trying to quantify what technologies and people generate value.

The meta-conclusion considering all the studies I know of (traveling and not looking them up), is the math becomes so Byzantine it’s a pointless measure and we should just stick with ideology to avoid blowing up society in people’s minds (an idea that was peddled for thousands of years already, reaffirmed by math, glad we spent the time on it.)

Just because we’ve emotionally conditioned ourselves to engage our mechanical agency towards these ends does not imply it’s because of management.

Manufactured consent is a thing.


Hm. Allocating credit for achievements is a tricky thing.

Management is necessary to allocate resources (including labor) efficiently at scale, but it comes with overhead.

As such, at smaller companies, startups, small businesses, the overhead of management can be outsized, the pain felt acutely when compared relatively to the actual output of ICs and such.

At larger companies, the massive cadre of management which is necessary to keep the machine running is also very obvious, as the overhead of a large organization is very large, even if the management itself is excellent and highly efficient.

As such... How does one allocate credit? An analogy would be like financing or investment. Necessary. A filtering mechanism. Sometimes helpful in an advisory mechanism. But are they the ones creating output? No. But do they deserve some credit? If something cannot be done without it, then they must deserve SOME credit, it's just a question of how much.


Why does society have to optimize at scale for assigning credit to a minority? If anything the studies show value is emotionally subjective.

A healthy society takes a village. IMO there’s plenty of literal history to show isolating a minority from the demands the rest of us face is ripe for abuse at scale.

One person did not invent languages, lay down the highways and invent computers.

All the people we hold up relied not just on the historical invention to push them forward but society giving them space and not killing them. Every individual inventor is outnumbered.

IMO that space to be and do is what we should optimize for. Not a tether to tradition of emotionally wanking off a handful over what is ultimately a linguistic twist on an idea that was discovered/defined collectively


Would be really interested on this paper. When you have time, if you can, mind linking to it?


What a pile of pretentious bullshit.


Yeah, well that’s just like... your opinion, man. Not a very interesting one. Par for the course online.


Valve and other flat organizations arguably qualify. Various tiny small startups may qualify depending on what people mean by managers.


Valve is, by all accounts, a toxic cesspit that's ruled by various influential cabals. They still have a hierarchy, they just aren't explicit about it.

At least for Medium, "Holacracy" pretty much imploded. Zsppos switched from it to an internal marketplace - which is, oddly, controlled by excecutives.

Flat organizations beyond a certain size either move away from flatness officially, create unacknowledged hierarchy, or fail. I'm really not aware of a single large-scale example that still works.

This is much different for small orgs - there's something about org size that requires some sort of centralization. (I think it's communications overhead + complexity exceeding the limit of what a single person can keep in their mind. I'm not 100% convinced, but these are the pressure points I usually see)


The video game sector is unstable and toxic. IMO, Valve compares favorably to other video game companies as for one thing it’s still in business.

Flat organizations are unusual, but survive about as well as other similar organizations. The difference is when a giant reorganization/buyout etc hits, they more obviously change into something else.


“without” is the wrong word: you need managers but they need to have incentives which align with the actual business and accountability. Most companies have problems with at least one of the two, which is a loss but usually indirect enough that it's easy to excuse or deflect.


Early Google. It worked for quite a while, but hit scaling limits.

Github made it all the way until 2014 (and was a massive success by then) too, IIRC.


You mean the companies where executives use extensive corporate speak and pull deadlines out of their asses, without checking in with the experts on the ground, triggering office heroics down in the trenches. We'll do it - we all get rent to pay.

Yeah, you can fake through it more than half of the time, comfortably. Just be eloquent, tall, and loud.


I don't know, where's the list of people who are healthy without disease-causing bacteria? (Have you washed your hands today?)


Valve is the only one that I've heard of but without manager there need to be a lot of internal politics to get your work done.


Has Valve done anything noteworthy in the last 10 years besides milking the Steam store? Their half-baked console seemed like an attempt to do more of the former.


I know you are trying to be snarky but Valve has accomplished a hell of a lot more than most startups ever will. They are far from perfect (hell, they’re known for bugs and their famously poor timelines, and I heard their game Artifact was not good,) but it’s pretty bizarre to suggest they do nothing. Maintaining Steam alone is huge. It’s hard to find a platform with more vigorous fanboys for good reason.

But that all aside... I mean they have a new Half Life game on the near term horizon (later this month apparently) and they have done a ton of work on VR and Linux support.


Fascinating article by Zvi (a very accomplished Magic: the Gathering player) about Artifact. https://thezvi.wordpress.com/2019/10/08/artifact-what-went-w...

Summary: the game was brilliant ("uniquely amazing") if you were really good at it but was essentially impossible to get into, and its economy was badly constructed.


Discussed at https://news.ycombinator.com/item?id=21199134 a few months ago.


Steam has a run rate in the billions! Maintaining something that strong that was built before the last decade is a significant innovation?

Most Startups are at least trying to change the status quo.


Yes, it is. Keeping Steam growing and well-liked is an achievement in itself. They never really stood still with it.

Steam certainly challenged the status quo when it was released, and several times in its life.

Hell, who else is working on Linux gaming right now? They were relatively early on game streaming, VR. Steam Workshop? How cool is it that a large amount of one of their flagship games, Team Fortress 2, not only integrated a lot of third party content, but also paid creators back? How cool is it that many games like Counter Strike began as Half Life mods?

I don’t really use Steam much anymore, so it’s not that I’m personally attached much, but I will admit to being pretty impressed.


Is this sarcasm? Steam has been the only consistent video game Cloud SAAS in the past 10 years.Can you name one time that the service was down without warning/proper recourse, had a data breach, exceeded SLAs, etc?

Also, valve's ther ventures are....massively successful. The microtransaction model in gaming originated with Valve, and now the entire mobile and freemium gaming markets use that business model.

So in this case...I would say yes, everything valve has done as a company trying to make money in the last 10 years is par for other successful SAAS platforms.


> team has been the only consistent video game Cloud SAAS in the past 10 years.

Xbox Live.


The title of this thread is "Companies fret as costs soar for software subscriptions", and this comment thread delved into cloud SAAS providers with a flat org structure.

Steam/valve was brought up because they have a successful cloud SASS model with no subscriptions oand a flat org structure.

Xbox live has a subscription fee since day 1, is maintained by MS (the definition of "non-flat organizational structure"), and is vendor locked. Not even remotely comparable.


Vulkan


Any small business, really.

This is what 'agile' is supposed to emulate for larger businesses.


> Because most work in the world is done despite the management, not because of it.

FTFY


Had to double check if I was on HN or on ca. 1999 Slashdot...


Is it different in other countries?


We got well over $80k into manpower on optimizing our self-hosted app for resources so that we didn't have to tell less than a dozen big customers to upgrade their hardware before people stopped laughing right in my face for suggesting we just buy them some new rackmount equipment as a gift...

That project died the following year, and I believe the opportunity costs of that work were substantially responsible for the demise.


Lol, I do this. My customers would often try to run my software on dusty old servers and then complain that it's slow, so now I just send them the hardware for free.


Different team, we did a cluster install and they were getting 10% of our advertised transaction rate. Something was clearly off. I commented “did they plug the whole thing into a 10Mb hub?”

They had. We sent someone eight timezones away to find that out.


Most employees don’t make nearly $100+/hour


The number to look at isn’t the employee’s wage, but the employee’s cost to the company. I cost my company about $100/hr more than I make.


>Then what happens is somebody making $100+/hour would spend several hours to do the job that would be covered by a tool that costs under $50/month.

Sure, but then there's the slightly more complicated economic calculus: do we build our own tooling now, for a significant up front cost, or do we buy the SaaS which we might use essentially forever for an unknown future TCO?


How often do you build in house tooling or applications that have no ongoing operating commitments? It's upfront capital investment plus generally unpredictable maintenance vs. More predictable but maybe higher operating expenses


How often do you build in house tooling or applications that have no ongoing operating commitments?

Absolute zero ongoing commitments? Very rarely.

Negligible ongoing commitments compared to the outsourcing alternative? Many, many times.


I don't know if more predictable is all that accurate. SaaS providers change functionality and UX regularly and that often breaks workflows. Then you have the ones that shut down or get bought out and then substantially change their offerings. Then there are those services that would have been private to an intranet, but are now public, and they get compromised.

I don't think it's clear-cut. Some SaaS has worked out remarkably well for me and others I wish I had just built the thing in-house. Also, few providers give a way to actually get your data back out in a usable fashion, so you tend to get locked in without substantial cost to back out.


Nothing complicated about it. You start with SaaS, validate it's a good tool to have/actually used, evaluate monthly costs vs opportunity costs, build your own when/if it makes sense.

The slightly more complicated question is "do we invest in optimizing SaaS costs/utilization? Or use those resources to start building our own?". But even for this question the right answer is usually fairly obvious.


>Nothing complicated about it.

If it was so easy, why are the companies in the article "fretting"? Because putting values on "good", "opportunity" and "makes sense" (not even counting the fact you have no control over the SaaS raising prices or going out of business in the future) is harder than we can spend $400/month or $50/month.


It is complicated at scale and over time though. Look at Saas infrastructure: How you do it as you grow constantly changes the results of any evaluation.


"spending another $10K to make them even 10% more productive is a bargain."

$200K is not a good benchmark for what companies spend.

Median salary in the US is about $50K [1]

Average overhead per/employee is 18-26% [2]

So a $10K spend for 10% increase in productivity on someone earning about $60K ... is not a bargain.

But there's a problem with this, because SaaS can be

1) Individual productivity 2) Operational productivity 3) Product spending (i.e. like GMaps integration as a feature in your product)

So 1/2 and 3 represent a very different type of calculus to the point wherein that kind of spending should definitely fall into different buckets for accounting.

[1] https://www.thebalancecareers.com/average-salary-information...

[2] https://beebole.com/blog/how-to-calculate-the-real-cost-of-a...


Median salary for a software engineer in the US is not $50k...

And the productivity increase using certain SaaS tools is not 10%, it can be an order of magnitude in certain instances.

https://money.usnews.com/careers/best-jobs/software-develope...

https://www.glassdoor.com/Salaries/software-engineer-salary-...

https://www.payscale.com/research/US/Job=Software_Engineer/S...

https://www.indeed.com/career/software-engineer/salaries


This is not about 'software developers' it's about industry at large paying subscriptions for a variety of services.

The vast majority of users of SaaS are not software developers.


The parent of the parent comment clearly referenced engineering with his “If you’re already spending >$200k/year per employee then a productivity increase of 10%...”


> And the productivity increase using certain SaaS tools is not 10%, it can be an order of magnitude in certain instances

well... my productivity isn't always increased, regardless of what tools someone wants to buy. and... sometimes it's a drain on my productivity, but using tool ABC increases someone else's productivity (at the expense of mine). How do you account for the productivity increase of -5% for 30% of your employees, but 20% for the other 70%?


SaaS also comes with

- Process mismatch / impedance : The software is unlikely to be able to do exactly what you want how you want. The human likely is able

- Vendor lock in: If the SaaS decides to change how it works, or the prices you now have a leaving cost

- Expenditure on someone else's competitive advantage: If you use tool X likely there is no reason your competitor cannot. You both use it, no net advantage gained over competitor. If you develop the expertise in house (likely the harder thing) then you have an advantage over your competitor. Of course this has to be balanced with what will be your competitive advantage and what will not

- Harder to manipulate: You cannot fire 1 of your SaaS and tell the other 2 to do the work


It isn't about what the spend per employee, but about the revenue earned per employee.

> Median salary in the US is about $50K

The median company isn't necessarily spending $10k per employee for software.


There is also opportunity cost - what could the employee have worked on that would have made you more money?


My bigger concerns is that it creates dependency, prices go up, and you either deal with it, or you change systems which is hugely costly.

FOSS as much as possible.


FOSS is free if your time is worthless.

There's plenty of situations in business where using it is an incredibly good idea. There's even more situations where you're better off just paying for a service.


Nobody truly believes "FOSS is free"... sure there's cost, you just gotta figure out what the cost is for you! (that's the trick a "free" product has a unique price for each "buyer", here in lies the beauty and the opportunity of the game). There's no "market price" so it's tricky, but when you can compute the cost and it happens to be very low for your team (and preferably high for competitor's teams), you have a competitive advantages.

OSS is all about leveraging "hidden" advantages agains competitors that have clear and obvious financial advantages over you. If you get it right, you win!

Of course, if you don't like risks and don't like "uncomputable" prices, or "prices unique to each buyer", then sure, be conservative, and stay away from FOSS.


You can pay for third-party support based on FOSS, and you'll still avoid being locked in to a single supplier.


1. You can, and it makes sense for FOSS projects with real support ecosystems, like RedHat, or for freemium FOSS projects, that explicitly sell support.

2. You can still be locked into a single-supplier when you pay for third-party FOSS support. In fact, most third-party FOSS support comes from a single supplier.

3. FOSS that doesn't have a sustainable business model attached to it rarely serves the needs of enterprises. There's FOSS alternatives to Sharepoint, but what's your FOSS alternative to Tableau? Salesforce? Slack? Will you actually save money by using it, compared to just paying a vendor?


> There's FOSS alternatives to Sharepoint, but what's your FOSS alternative to Tableau? Salesforce? Slack?

From a purely technical POV, I'm pretty sure that feasible alternatives exist to all of these, e.g. I don't think Tableau's data viz product does anything that couldn't also be done with relative ease in R. Of course, these commercial offerings extend way beyond the purely technical domain where FLOSS makes the most sense; but that has little to do with "sustainable business models" - R itself is plenty sustainable on its own - and a lot to do with extremely niche or obscure "needs of enterprises" that FLOSS projects are either not clearly aware of, or not very interested in supplying. Niche and obscure problem domains have always been problematic for the open source model, this is nothing new.


"I don't think Tableau's data viz product does anything that couldn't also be done with relative ease in R."

It's zingers like this that keep me coming back to threads like this one! "No wireless. Less space than a nomad. Lame."


I'm indifferent on the supplier lock-in. Mainly because our biggest supplier is AWS and they're >%50 of the market.


You're getting downvoted but you're absolutely right.

I use FOSS when it makes sense but 90% of the time the answer has been, 'just use what AWS has.'


Managed FOSS is the way to do it


They screw you on the upsells.


This is reductive. If "10% more productive" translated into >= 10% more revenue as a percentage of their wage, it would make sense but it varies so wildly, that it's not a useful statement.

Some will translate to increased revenue, a few times over, (which incentivizes the vendor to increase the price accordingly) and some will be net negative.


You say with certainty that which is uncertain to the prospective buyer.


What about the ones that cost thousands to decrease productivity?


Immediately made me think of Jira.


That's funny, I was thinking of all of the things people pay even more for to replace Jira.


Jira's awful except for all the other options :3

It's a lowest common denominator product. Which is fine. I use it for anything I need to share on and Trello for organizing, "in my own head."

I won't even look at sharepoint.


Alternatives?


Depends on the context, I guess. I'm generally more comfortable with lightweight things - like GitHub projects for software and Trello for everything else.


i'm sure there's like a bazillion lower cost alternatives.


The person who convinced me to [pay for a personal license for] Jetbrains instead of torturing myself with Eclipse just passed on the questions that had convinced him to do it:

How much does it cost you for Jetbrains? How many technical books is that? Is in worth that many books a year for a decent tool?


I think maybe I misunderstood the headline then.

Is the expectation that your employees cost > $200k or is that not qualified?

In some companies / countries you spend $10k per employee.


If your SaaS services more than 10x their prices (which happened to many SaaS like Segment), suddenly $10K/yr -> $100K/yr and the calculus changes


how many 0.0001% of employees does that represent though ? 200k is more than what the average CEO makes (gross) in france (150k€)


Include health insurance, retirement match, employer side payroll taxes, equipment, etc. I don’t make nearly that much but I easily cost my company more than that.


> Include health insurance, retirement match, employer side payroll taxes,

all of these three are included in the gross in france


Question is how much would it cost to host in-house, taking all costs into consideration (personnel, regulatory, etc)


They are just going to pay their employees less. Companies have to be profitable, temporary exceptions notwithstanding, and the money has to come from somewhere.


As long as the purchased software is worthwhile (i.e. it makes employees more effective; the company gets more output per employee expense dollar), this trend should actually increase employee pay. Higher output per employee allows a higher salary per employee while still being profitable.


It also allows to pocket more money instead of paying it to employee.

Remember, many people in USA believe that corporation has duty to increase shareholder value. And similar stupid ideas.


> many people in USA believe that corporation has duty to increase shareholder value. And similar stupid ideas.

No more stupid than the idea that companies are supposed to exist solely for the benefit of the employee. Shareholders are the ones that put the money up to start and grow the company in the first place, so yes, there is an obligation to shareholder value.

Don't like that? Then you are free to start a company that doesn't take investment.


Shareholders are only able to make contracts because the government uses guns to arrest people who break contracts. Don't like it? Move to an offshore oil rig, I guess.


Shareholders take a risk in their investment, mitigated by making it more fluid as they can easily trade the shares compared to more traditional investments.

There's however no law that holds a figurative gun to the heads of corporation that their duty is to increase the value of those shares.


Not sure if you are being facetious, but that's in fact the law. The board is a fiduciary of shareholder (investor) money. Is that a great idea for society? Not sure, until it changes we live in the world we live in.

And on the productivity point, that makes a good theoretical case but it is just not demonstrated by historical data. In the US, in the past 50 years, it even has a name (the pay-productivity gap or the wage gap). You can look at productivity vs wages since 1970 and see how they have not correlated. Further, you can go into income and wealth inquality growth over the same time (https://en.wikipedia.org/wiki/Income_inequality_in_the_Unite...) and see how the wealth generated by the productivity gains has been highly concentrated in the top 0.1% and 0.01% wealthiest. Some people will talk about real or nominal wages (adjustments with something called the Implicit Price Deflator) that show wages purchasing power increasing, but I think one can look at costs of home, college, and medical care and other critical things in social mobility that have increased in cost far exceeding inflation as complications to that idea.


No, that's not what "fiduciary duty" means.

Fiduciary duty means that the board is supposed to be honest with the shareholders and not actively work against the corporation. That does not mean, in fact, that they are required to put growth of shareholder value to any respect.


It's more complicated than that.

https://en.wikipedia.org/wiki/Dodge_v._Ford_Motor_Co.

> Dodge v. Ford Motor Company, 204 Mich. 459, 170 N.W. 668 (Mich. 1919) is a case in which the Michigan Supreme Court held that Henry Ford had to operate the Ford Motor Company in the interests of its shareholders, rather than in a charitable manner for the benefit of his employees or customers. It is often cited as affirming the principle of "shareholder primacy" in corporate America. At the same time, the case affirmed the business judgment rule, leaving Ford an extremely wide latitude about how to run the company.

[...]

> In the 1950s and 1960s, states rejected Dodge repeatedly, in cases including AP Smith Manufacturing Co v. Barlow or Shlensky v. Wrigley. The general legal position today is that the business judgment that directors may exercise is expansive. Management decisions will not be challenged where one can point to any rational link to benefiting the corporation as a whole.

The caveat is important. If a company says, "Customers don't like it when rich people in suits treat <x> like shit. Therefore, if we spend a small amount of money treating <x> better, this will make our customers happy, and they will buy more of our products, recouping the cost and bringing in more profit for our shareholders." Substitute the environment, animals, customers, employees, people in Africa, cancer patients, the children, etc. It doesn't matter if you're completely full of shit, you just have to make the argument.

So while you're correct that companies have a primary requirement to increase profits for their shareholders, in practice, this requirement is incredibly loose, to the point of not actually being a requirement at all.


IANAL, but it doesn't read to me like a requirement to increase profits for shareholders, but requirement for "good stewardship" of the investment.

But that's beyond my knowledge on this matter :)


Doesn't higher output per an employee lead to a change for demand of that type of employee, lowering pay?

It always seemed to me the value an employee brings is the upper bound on pay (demand drops near 0 over that), but the pay is determined by supply and demand for that labor.


> Doesn't higher output per an employee lead to a change for demand of that type of employee

Yes, pretty much so.

> lowering pay?

Nope. Reality is way more complex than that.


Companies don't pay employees what they can, they pay what they have to (decided by the market). If they can't afford to do that, they won't survive.


That's not how supply and demand works. People aren't going to accept lower wages because your company pays for a dashboard.


What if you spend $50K per employee?


Big if.


Most enterprises don't spend >200k an employee. Do people really believe that is a normal salary?


The fully loaded cost of an employee for a business is not the employee's salary.

"The fully-loaded costs of employees are much higher than their salary: exactly how much higher depends on your locality’s laws, your benefits package, and a bunch of other HR administrivia, but a reasonable guesstimate is between 150% and 200% of their salary." -https://www.kalzumeus.com/2012/01/23/salary-negotiation/


Most employees' salaries are not 100k.


> Most employees' salaries are not 100k

I'd hypothesis that most employees on whom companies spend $10k for cloud software are making $100k+.


Obviously the fully-loaded cost of an employee goes beyond the cash compensation.

But even so, the average salary in the US is < $60k. Use whatever multiplier you want, $200k is way high.

Companies are spending far less than $200k/employee all-in.

So to slightly rephrase the OP's question:

> Do people really believe that $100k is a normal salary?


Yes but companies that are spending $60K on an employee aren’t spending $10K on their SaaS.


I don't know much about jobs outside of tech, but I know most of them at the entry level to a few years of experience in accounting, finance, real estate, etc. do not pay much over $60k (if that) and pretty much all of those jobs are going to use things like Salesforce or other expensive SAAS tools and software.


Salesforce "Enterprise" is $1800/user/year [0], and presumably large enterprises are negotiating better rates than the published ones. Every single employee would need to be using five and a half packages like this at full list price to hit $10k/year.


In Sales you would have SFDC plus some kind of enablement tool like Salesloft/Outreach plus a prospecting tool like Sales Navigator or DiscoverOrg plus enterprise Gmail plus a communication tool like Zoom; and then there's the analytics stuff, the HR/finance parts, etc...10k is probably way too high an estimate but 5k is believable for a B2B startup with a scaling sales team.


I won't quibble too much with the precision of the $10k number, which will vary widely.

But it's pretty obvious that a basic "customer support" combo like Office 365 + Okta + Zendesk + Intercom + Slack + Zenefits starts to add up to a meaningful percentage of the employee's take-home compensation. (Remember these are positions where $20/hr is a good rate.) And remember, the comparison figure for managers over 50 is roughly $0, which approximates the spend on these functions as far back ago as 2008.


The assumption is a $100K salary. The other $100K is benefits, management, overhead, etc.


I don't know about most enterprises but I'm guessing any given cost per employee includes a lot more than salary.


In London, I'd guess the median salary for most enterprises is around £30-40K. National insurance, a desk, and a few other benefits are probably £20K on top of that.


If you're talking central London, that should be a fair bit higher.


I'm going on my recent experiences in the City (i.e. the financial district).


Recruitment, payroll, HR, facilities are other costs.


Factoring in all compensation, that is still a ridiculous figure.


Fully loaded costs (health benefits, 401k matching, cost of computer, HR business partners, etc.) for an employee are roughly 2x what their salary is. So yeah, 200k is a good estimate.


That's still very high. Average salary is much lower than $100k in the US. Even among college grads with 10+ years of experience, $100k is still a high-end job paying a high-end salary. It's nowhere near the average.


Take out all the low wage workers that don't use any apps, and the average goes up.


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

Search: