The problem with free (2019)(vanschneider.com)
vanschneider.com
The problem with free (2019)
https://vanschneider.com/blog/the-problem-with-free/
45 comments
Yes: this piece kind of lost me with the opening. Consumer software barely existed in 1999?
Even ignoring games, have you ever seen the contents of a PC World cover CD? It's all over the map: there was consumer software for every conceivable reason and purpose. Commercial, shareware, PD, and open source were all alive and well - in rude health, in fact - in 1999.
Going further back, I remember the 80s. Even the little market town I lived in had at least half a dozen different places you could buy software, and the mail order scene for software, and weird hardware devices paired with software, was very much alive and well. I'd go as far as to say that, budget games aside, mail order was my default mechanism for buying computer software until well into the 90s when online became more prevalent.
Chevrolet, of all companies, used to literally distribute a program that was effectively a virtual and deeply pixellated car showroom to consumers in 1987: https://www.youtube.com/watch?v=I89vPzhpNNA.
Consumer software back in the day was widespread, diverse, weird, and wonderful. Going back into the 70s and early 80s, much of the early innovation in consumer software for personal computers, came from the very consumers - the hackers, hobbyists, and tinkerers - who bought them.
Remarks like "Consumer software barely existed just 20 years ago" simply come off as out of touch and ignorant of history, which has the effect of undermining the credibility of any other point the piece is trying to make (especially as an opener - WTH?).
Even ignoring games, have you ever seen the contents of a PC World cover CD? It's all over the map: there was consumer software for every conceivable reason and purpose. Commercial, shareware, PD, and open source were all alive and well - in rude health, in fact - in 1999.
Going further back, I remember the 80s. Even the little market town I lived in had at least half a dozen different places you could buy software, and the mail order scene for software, and weird hardware devices paired with software, was very much alive and well. I'd go as far as to say that, budget games aside, mail order was my default mechanism for buying computer software until well into the 90s when online became more prevalent.
Chevrolet, of all companies, used to literally distribute a program that was effectively a virtual and deeply pixellated car showroom to consumers in 1987: https://www.youtube.com/watch?v=I89vPzhpNNA.
Consumer software back in the day was widespread, diverse, weird, and wonderful. Going back into the 70s and early 80s, much of the early innovation in consumer software for personal computers, came from the very consumers - the hackers, hobbyists, and tinkerers - who bought them.
Remarks like "Consumer software barely existed just 20 years ago" simply come off as out of touch and ignorant of history, which has the effect of undermining the credibility of any other point the piece is trying to make (especially as an opener - WTH?).
This rankled me too. I understand the world was different, but "barely existed" ?
> "barely existed"
Maybe the term is better understood in terms of amount of "offered software then vs. now"?
Maybe the term is better understood in terms of amount of "offered software then vs. now"?
It is true that however you measure it there is a lot more software now.
But the phrase "barely existed" has connotations of struggle. As though it was close to not existing at all.
No one says "cars barely existed in the 1930's" or "humanity barely existed pre 1600" (world population ~0.5 billion).
By 1999 consumer software definitively existed and was thriving just fine even if it was comparatively early days to what we have now.
But the phrase "barely existed" has connotations of struggle. As though it was close to not existing at all.
No one says "cars barely existed in the 1930's" or "humanity barely existed pre 1600" (world population ~0.5 billion).
By 1999 consumer software definitively existed and was thriving just fine even if it was comparatively early days to what we have now.
I distinctly remember buying consumer software 30 years ago after I got out of the army early and they accidentally sent me an extra paycheck which promptly went to buying a Macintosh.
There was a thriving market of consumer desktop applications stretches back to the mid-1990s. Windows 95 spurred further growth of consumer software (also far more variety).
Perhaps the author is referring to the rise consumer-facing SaaS (Software as a Service).
Perhaps the author is referring to the rise consumer-facing SaaS (Software as a Service).
Yeah 22 years ago as a young kid I was obsessed with reading the SOFTWARE catalog of my local computer store (the hardware stuff was all video cards that all looked like each other, the software was exciting stuff like space simulators! 3D graphics! video editing!).
And this was after the days of
https://en.wikipedia.org/wiki/Microsoft_Home
were quite over.
And this was after the days of
https://en.wikipedia.org/wiki/Microsoft_Home
were quite over.
There was plenty of commercial software in the 80s as well. Go find some C64, Amiga, Apple 2, TRS-80, Atari 800 emulator online. All of that software used to be in stores on shelves
"Consumer software barely existed just 20 years ago."?
I'd say not only it existed, but it existed more than today, if you exclude games.
Eg, things that used to be for sale, and aren't really much of a thing anymore: various system tools like Norton Utilities, desktop toys that did junk like animated cursors in Windows 3.1, file compression tools, disk compression tools (Stacker), replacements for the Windows program manager, special purpose tools (PawSense is a funny one), memory managers (QEMM), PartitionMagic, Web browsers (Netscape was a packaged, paid product), mail clients (Eudora), possibly antiviruses (it's my impression that these days Windows Defender does the job)
The likes of DOS and Windows 3.1 were barren of almost all functionality out of the box compared to a modern system. If you wanted to get something done legally, you had to buy a bunch of stuff. And then often times you'd want a bunch of misc tools on top.
These days pretty much all of that has been absorbed into the OS, or has perfectly good free alternatives, or is functionally obsolete and not really a concept anymore.
I'd say not only it existed, but it existed more than today, if you exclude games.
Eg, things that used to be for sale, and aren't really much of a thing anymore: various system tools like Norton Utilities, desktop toys that did junk like animated cursors in Windows 3.1, file compression tools, disk compression tools (Stacker), replacements for the Windows program manager, special purpose tools (PawSense is a funny one), memory managers (QEMM), PartitionMagic, Web browsers (Netscape was a packaged, paid product), mail clients (Eudora), possibly antiviruses (it's my impression that these days Windows Defender does the job)
The likes of DOS and Windows 3.1 were barren of almost all functionality out of the box compared to a modern system. If you wanted to get something done legally, you had to buy a bunch of stuff. And then often times you'd want a bunch of misc tools on top.
These days pretty much all of that has been absorbed into the OS, or has perfectly good free alternatives, or is functionally obsolete and not really a concept anymore.
I have no problem paying for software. The problem is most software is total crap these days. And I refuse to pay subscriptions. There are great examples of stuff I'm HAPPY to pay for. iStat menus, KeyboardMaestro (and I used to love 1password before they went subscription).
Don't understand the hate for subscriptions. Some companies moved to subscriptions purely for a money grab, but for a lot of companies a subscription model makes more sense and a lot of them are really good value.
1password for example is only $3 USD a month. I don't understand why you would stop loving 1password because of a $3 subscription.
1password for example is only $3 USD a month. I don't understand why you would stop loving 1password because of a $3 subscription.
You want to build a stable system. Set it up so that the lower level pieces work and will continue working if untouched so you don't have to think about them and instead focus on the higher level stuff.
A software that is subscription based means your computer system seen as a whole will stop working at some point in the future if you don't remember to take a specific action (like updating the payment method or some such). It means uncertainty, it basically means you can only guarantee your system will work as-is for the next month. The more software of this kind you have, the more complex your system becomes to maintain in a given state and it burdens your mind.
I think it's fine for software you only intend to use for a few weeks/months at a time, or for software where you always want to be on the bleeding edge of tech, but for things as low level as a password manager? You want something really stable. It should also not update itself. So if we are not interested in updates what's the point of paying monthly?
A software that is subscription based means your computer system seen as a whole will stop working at some point in the future if you don't remember to take a specific action (like updating the payment method or some such). It means uncertainty, it basically means you can only guarantee your system will work as-is for the next month. The more software of this kind you have, the more complex your system becomes to maintain in a given state and it burdens your mind.
I think it's fine for software you only intend to use for a few weeks/months at a time, or for software where you always want to be on the bleeding edge of tech, but for things as low level as a password manager? You want something really stable. It should also not update itself. So if we are not interested in updates what's the point of paying monthly?
My main gripe with 1Password is that it used to be a great product. It worked well, was relatively fast and stayed out of the way.
All those things have gone in later versions though.
Now it's very laggy and keeps getting in the way of my browsing, and I have lost passwords for new websites just because the modal to create the entry would "freeze" although not completely - I could still press the cancel button which would then ... cancel the creation, but also delete the temporary password item.
This modal stops you from using anything else on the browser as well, so there's no way out of it.
All those things have gone in later versions though.
Now it's very laggy and keeps getting in the way of my browsing, and I have lost passwords for new websites just because the modal to create the entry would "freeze" although not completely - I could still press the cancel button which would then ... cancel the creation, but also delete the temporary password item.
This modal stops you from using anything else on the browser as well, so there's no way out of it.
Why do you refuse subscriptions?
I always thought they made more sense, especially for software that needs a lot of maintenance.
It should be up to the company/developer to choose their business model, and it should be up to the market to accept or reject it.
Subscriptions do have some advantages for the customer: they often come with a free trial, and if you later on decide that they're not worth the value, you can (hopefully, easily) cancel; whereas getting a refund for a one-off big purchase might be more tricky, or even impossible; your resale ability might also be limited. The subscription model also helps align incentives: the developer must continue providing good value, or risk losing subscribers.
But on the other hand: people want to own things they pay for.
Subscriptions do have some advantages for the customer: they often come with a free trial, and if you later on decide that they're not worth the value, you can (hopefully, easily) cancel; whereas getting a refund for a one-off big purchase might be more tricky, or even impossible; your resale ability might also be limited. The subscription model also helps align incentives: the developer must continue providing good value, or risk losing subscribers.
But on the other hand: people want to own things they pay for.
Counterpoint - DaVinci Resolve is a professional high-end video editing, compositing, and colour grading package. The free-as-in-beer version omits some specialised tools (like 3D camera tracking and support for timelines larger than 4k resolution), which are present in the full-fat paid-for Studio version.
It costs about 200 quid, or you get a licence for it free with basically any of their cameras or edit controllers.
Once you've paid for it, you own that copy, and so far in any of their releases your key will work with all the updates. Version 18 is about to drop, and still even a key for 14 or 15 will work just fine.
I get that they make their money from selling hardware and Adobe don't, but less than a year of licence fees for Premiere would pay for a copy of Resolve. If you waited 18 months you'd be able to afford the Speed Editor package and then you'd get a cute wee edit controller with a proper jog/shuttle wheel and proper clicky buttons with your licence.
The 27 quid a month for a Premiere licence would be one of my larger monthly outgoings.
It costs about 200 quid, or you get a licence for it free with basically any of their cameras or edit controllers.
Once you've paid for it, you own that copy, and so far in any of their releases your key will work with all the updates. Version 18 is about to drop, and still even a key for 14 or 15 will work just fine.
I get that they make their money from selling hardware and Adobe don't, but less than a year of licence fees for Premiere would pay for a copy of Resolve. If you waited 18 months you'd be able to afford the Speed Editor package and then you'd get a cute wee edit controller with a proper jog/shuttle wheel and proper clicky buttons with your licence.
The 27 quid a month for a Premiere licence would be one of my larger monthly outgoings.
If that company, the one that develops DaVinci Resolve, only produced software, would that model of £200 fixed one-time payment be sufficient to be financially viable?
Let's say the company's headcount to develop/maintain such software, plus a bunch of other required roles such as marketing, sales, hr, etc., is around 25 people averaging £80,000/year each (not that far off considering employees cost more than just their salary) to make it an even £2 million/year.
That means in order to not go bankrupt, i.e. break even, they need to sell 10,000 copies to new buyers every single year. Is that realistic? I don't think so, a company that tries that model will soon go in the red because we aren't even considering other costs that a company has, which aren't negligible. Even if you cut the estimate above in half, e.g., let's say they cost £40,000 each instead, that's still 5,000 new buyers of the software each year.
In contrast, a subscription model of £100/year requires them to have 20,000 ongoing subscriptions, which is substantially easier to achieve. I believe there's a reason that a lot of companies have shifted/are shifting to a subscription model, an alternative being monetisation models such as advertisement-supported programs.
Let's say the company's headcount to develop/maintain such software, plus a bunch of other required roles such as marketing, sales, hr, etc., is around 25 people averaging £80,000/year each (not that far off considering employees cost more than just their salary) to make it an even £2 million/year.
That means in order to not go bankrupt, i.e. break even, they need to sell 10,000 copies to new buyers every single year. Is that realistic? I don't think so, a company that tries that model will soon go in the red because we aren't even considering other costs that a company has, which aren't negligible. Even if you cut the estimate above in half, e.g., let's say they cost £40,000 each instead, that's still 5,000 new buyers of the software each year.
In contrast, a subscription model of £100/year requires them to have 20,000 ongoing subscriptions, which is substantially easier to achieve. I believe there's a reason that a lot of companies have shifted/are shifting to a subscription model, an alternative being monetisation models such as advertisement-supported programs.
Other companies which sell (only) software based on one-time payment model:
- ACD Systems (ACDSee)
- Serif (Affinity)
- Freron Software (MailMate)
- Sublime HQ (Sublime Text)
I use them all. Wouldn't pay for a subscription. I don't want a service. I want a product.
I also pay for services, but for other purposes: mail (Fastmail), online bookmarking (Pinboard), music streaming (Spotify, mostly for discovery; after I discover something I like, I usually go and buy it to have it myself), VPN, VPS. If there existed a decent video streaming service, I'd probably pay for that as well, but there isn't any.
- ACD Systems (ACDSee)
- Serif (Affinity)
- Freron Software (MailMate)
- Sublime HQ (Sublime Text)
I use them all. Wouldn't pay for a subscription. I don't want a service. I want a product.
I also pay for services, but for other purposes: mail (Fastmail), online bookmarking (Pinboard), music streaming (Spotify, mostly for discovery; after I discover something I like, I usually go and buy it to have it myself), VPN, VPS. If there existed a decent video streaming service, I'd probably pay for that as well, but there isn't any.
I'm not saying it's impossible to sell software with one-time payment model, I'm saying it's noticeably harder. There's always going to be companies that make do with such business models, but what's the ratio of successful companies with one-time payment out of all companies that have tried that? How does that compare with companies that go for subscription model (or ad-supported model)?
And are there any behemoths (i.e., hugely profitable companies) that do one-time payment software? Because as far as I know Microsoft would have been one of the few or only such cases (and obviously it only applies to a subset of their products, many others are using other business models) and they are also moving towards a subscription model, e.g., for Office, which considering they probably have a legion of financial analysts, accountants, etc. I assume they have done their due diligence to figure out it's worth it.
And are there any behemoths (i.e., hugely profitable companies) that do one-time payment software? Because as far as I know Microsoft would have been one of the few or only such cases (and obviously it only applies to a subset of their products, many others are using other business models) and they are also moving towards a subscription model, e.g., for Office, which considering they probably have a legion of financial analysts, accountants, etc. I assume they have done their due diligence to figure out it's worth it.
Adobe used to be such a behemoth. They didn't switch to subscriptions due to unprofitability. They just wanted more and more. Same with Microsoft. They're a behemoth switching to subscriptions now. But that's not because they're running low on money. And most games are still one-time payment today. So you can count Ubisoft, EA etc. Of course there are other sources of payment for them than just the nominal price of the game, but most gamers only pay for the game and ignore everything else.
Besides, I don't think I want a market dominated by behemoths. If popularity of one-time payment model would lead to a market dominated by <25 person companies with an odd behemoth here and there, then I'm in.
Also, most software companies today are startups which exist only thanks to investor money. It's hard to say what a customer-driven market would look like. Although maybe recession is going to clear this up.
Besides, I don't think I want a market dominated by behemoths. If popularity of one-time payment model would lead to a market dominated by <25 person companies with an odd behemoth here and there, then I'm in.
Also, most software companies today are startups which exist only thanks to investor money. It's hard to say what a customer-driven market would look like. Although maybe recession is going to clear this up.
As I said most of their money comes from cameras. £2M per year would be a couple of thousand cameras, which feels like a fairly easy number to hit.
Yes, I know you said they do sell hardware and that's probably how they survive.
If you look at my comment I said "if they only produced software", because I wanted to use it as an example (just from your £200 figure) with back of the envelope estimates to show why many companies move to subscriptions, because if you have a niche product, having new customers in the thousands or tens of thousands every year is incredibly difficult compared to having a reliable set of recurring customers with subscriptions.
If you look at my comment I said "if they only produced software", because I wanted to use it as an example (just from your £200 figure) with back of the envelope estimates to show why many companies move to subscriptions, because if you have a niche product, having new customers in the thousands or tens of thousands every year is incredibly difficult compared to having a reliable set of recurring customers with subscriptions.
The reason things tend toward free in software... at least consumer software... is that capitalism is really optimized for things that have significant marginal cost. Over time, supply and demand tend to cause the price to approach the marginal cost. (the cost to make one more product)
But software has effectively zero marginal cost. Meaning normal competition and the free market just don't work.
Say a company makes some software that costs them a million dollars to make. They figure their best chance of making a profit is by pricing it at $100, and selling at least 10,000 copies.
Now another company decides to compete as well. It costs them a million dollars as well to develop their competing product, which we can imagine is every bit as good and serves the exact same need.
But they figure their best chance of making a profit is to sell it for $10. Sure, they make less per copy, but the lower price should allow them to sell a lot more. So they'll need to sell at least 100,000 copies.
That's bad. There is no reasonable way of saying how much value a single copy has. It's not like making, say, a bicycle where the bulk of the price is to cover the marginal cost.... i.e. the cost of making one more bicycle.
Anyone developing software must realize that they could be massively undercut by competition. It's just not a stable model.
Does anyone remember when Apple's WebObjects product was reduced in price from $200,000 to $699? Then a year later, bundled with the operating system?
But software has effectively zero marginal cost. Meaning normal competition and the free market just don't work.
Say a company makes some software that costs them a million dollars to make. They figure their best chance of making a profit is by pricing it at $100, and selling at least 10,000 copies.
Now another company decides to compete as well. It costs them a million dollars as well to develop their competing product, which we can imagine is every bit as good and serves the exact same need.
But they figure their best chance of making a profit is to sell it for $10. Sure, they make less per copy, but the lower price should allow them to sell a lot more. So they'll need to sell at least 100,000 copies.
That's bad. There is no reasonable way of saying how much value a single copy has. It's not like making, say, a bicycle where the bulk of the price is to cover the marginal cost.... i.e. the cost of making one more bicycle.
Anyone developing software must realize that they could be massively undercut by competition. It's just not a stable model.
Does anyone remember when Apple's WebObjects product was reduced in price from $200,000 to $699? Then a year later, bundled with the operating system?
Oh, that's easy: just require that every copy sold/downloaded incurs a 5¢ tax on the the publisher. Now the marginal cost is non-zero and we can yet again let the market sort things out.
This assumes an infinite market. If you address a market with only 10000 consumers, it's unlikely competition can succeed with prices assuming 100000 sales.
The size of the market generally scales with your marketing and distribution capacity.
To the point of the original article, smaller companies have a hard time competing due to less reach. That's missing. If I run Disney and can reach 4 billion people, and your 5-person shop can reach 50,000 people, I can either price my product 5 orders of magnitude lower than you, or invest an extra 5 orders of magnitude into development.
That's expensive. If I want to comply with Khazak tax laws, provide support there, and reach distribution channels there, I probably want to pay for an office there. That office might have four people, regardless of whether we sell 1 product or 100 products.
A lot of what app stores do is provide that kind of reach to all businesses, driving pricing down.
Personally, I'm a big fan of government grant funding for free software. It seems like the OS, office suite, video editor, and similar tools ought to have a baseline versions available to anyone for free.
To the point of the original article, smaller companies have a hard time competing due to less reach. That's missing. If I run Disney and can reach 4 billion people, and your 5-person shop can reach 50,000 people, I can either price my product 5 orders of magnitude lower than you, or invest an extra 5 orders of magnitude into development.
That's expensive. If I want to comply with Khazak tax laws, provide support there, and reach distribution channels there, I probably want to pay for an office there. That office might have four people, regardless of whether we sell 1 product or 100 products.
A lot of what app stores do is provide that kind of reach to all businesses, driving pricing down.
Personally, I'm a big fan of government grant funding for free software. It seems like the OS, office suite, video editor, and similar tools ought to have a baseline versions available to anyone for free.
Which can also create another problem.
The second company might think there is a market for 100k users, and price accordingly at $10, but only sell 10k copies and go out of business.
Now, however, people see the product as only being worth the lower price point of $10, and won't pay $100 for it anymore. So company A can only sell 1000 copies at $100, and also goes out of business.
The second company might think there is a market for 100k users, and price accordingly at $10, but only sell 10k copies and go out of business.
Now, however, people see the product as only being worth the lower price point of $10, and won't pay $100 for it anymore. So company A can only sell 1000 copies at $100, and also goes out of business.
It does not assume an infinite market, it just assumes a larger market than your example. Everything still applies if you just adjust the numbers downward or upwards to match whatever particular market you are speaking of.
This is definitely an article that hit (some) right ideas based on incorrect assumptions or an inaccurate historical context. Don't know if I would recommend it to a broader audience.
tl;dr: If the product is free, you are the product.
tl;dr: If the product is free, you are the product.
> If the product is free, you are the product.
I used to contribute to a GPL licensed program which is apparently doing pretty well in its industry these days and is just free. They pay developers a living wage and have a plug-in ecosystem which also generates income for people to develop them (I assume, don’t know if people live off plug-in income).
I used to contribute to a GPL licensed program which is apparently doing pretty well in its industry these days and is just free. They pay developers a living wage and have a plug-in ecosystem which also generates income for people to develop them (I assume, don’t know if people live off plug-in income).
Even if it's not free, you may still be the product (Windows)
Isn't windows sort of quasi-free these days? They've been aggressively pushing free version upgrades for Windows on me several times.
Upgrade is free but you still need an original copy
But if you are resistant to advertisements then many products are free for real. For example, free games are "not free" since they push you to buy microtransactions, but if you can easily resist that sort of temptation then the product is completely free to you. This isn't especially rare either, most people playing free games doesn't pay anything for it so for most people those games are really free.
So a revised statement would be "If the product is free, some users are the product.", as many users aren't easily manipulated and therefore don't become products.
So a revised statement would be "If the product is free, some users are the product.", as many users aren't easily manipulated and therefore don't become products.
I feel there is the wider problem that consumer goods are increasingly being separated from commercial/industrial goods, and consumer goods generally are just crap. This is a trend that needs reversing before we can really consider the specifics of software
Some tangible solutions to the supposed issues proposed in this article: paid technical support for your free product, votes for new features, consulting with companies that have problems your software can solve, and commercial vs consumer licencing.
By no means a new or revolutionary take but important nonetheless. I do, however, believe that this model is just too profitable to go away. People don't really care about these issues and are instead learning to live in the panopticon.
With the rise of privacy protections like GDPR, I wonder will this emphasise oligopolys. Never thought about it in that context before
Gordonjcp(1)
I kinda hate to be that guy but there was a whole store at the mall called Babbage's that sold consumer software. Sure, half was games but there was still plenty of other stuff.