Open Source Apps (2013)
stratechery.com@patio11's recent tweetstorm touched upon the core point
> Most open source software is written by programmers who are full-time employed by companies which directly consume the software, at the explicit or implicit blessing of their employers. It is not charity work, any more than they charitably file taxes.
https://twitter.com/patio11/status/936629310785437696
The actual monetization, as the post explains, comes from related products and services. The presented Oracle analogy to IAP is fascinating:
> Oracle offers a basic version of MySQL under a GPL license, with additional closed source – and expensive! – add-ons that better fit specific businesses’ needs. The allegory here is in-app purchase: a basic experience that gets the user invested in your product, with value-added options that are more attractive because of said investment
Apps don’t “want to be free”; customers want to try before they buy and there are not a lot of sane options presented for doing that in App Stores.
Why the hell isn’t there a simple “download now, pay if you keep more than 4 hours” kind of option, for example? Why no simple “upgrade to version 2.0 for 50% off” option? There are lots of common scenarios that app stores should provide, saving developers from having to hack special purchases into the product. Worse, stores basically don’t distinguish between those types of purchases and the many, many scams so after awhile potential buyers equate “any in-app purchase = scam” and that option disappears. And on top of it all, good, reputable apps can eventually be updated with scammy features that you never used to have.
Terrible app search engines. Terrible browsing with unchanging top-10 lists. Terrible monetization options. Endless garbage apps competing for space. Yet 30% comes off the top unless apps are free. So they’re free, and you get mostly ads.
What a complete waste of potential in the app ecosystem right now.
> Why the hell isn’t there a simple “download now, pay if you keep more than 4 hours” kind of option, for example
So, one thing I've seen a _lot_ of new apps do is basically this; but sadly, includes some annoyances itself.
Instead of "download, pay after x hours", they instead have a subscription model gating most of the features of the app. They give you 1 (week|month) free, as that's easy to do with in-app payments automatically. Of course, now you have to remember to cancel this subscription if it turns out it's not what you're after, which is annoying, albeit pretty simple on iOS.
>>> Why the hell isn’t there a simple “download now, pay if you keep more than 4 hours” kind of option, for example?
There is on Google Play but the window is only 2h.
> What makes the software market so fascinating from an economic perspective is that the marginal cost of software is $0.
I see this claim a lot.
The marginal cost of software is actually its support cost﹡.
If some customers bought the product without a support contract, then the marginal cost of the good that is sold to those customers is the (non-zero) cost for support staff to categorize them and turn them away.
If none of the license includes support, then the vendor doesn't need support staff or activities, and the marginal cost is indeed zero.
Only some software products — especially among commercial software products — don't include any support at any price point.
﹡ Also marginal increases in distribution costs. Also costs that stem from marginally greater exposure to product liability issues and claims.
Having more users even on the free plan gives many benefits: - more testers means that bugs are detected earlier. This improves reliability. - it gives more opportunities to try to sell the support and convert them into paying customers - it increases the visibility of your product. It is free advertising.
IMHO, the marginal cost is generally better than zero.
It's a trade-off.
The benefit is "reach" which is beneficial for (a) marketing (b) OSS community (c) bugs and contributors.
The cost are the (a) lower conversion to payment (b) costs from a 'shared' community e.g. community management and user support.
It's not that apps want to be free, as much as apps that any reasonably competent programmer can make are driven down in price because many reasonably competent programmers do make them.
If you look at applications that require a large amount of domain expertise to build (CAD, audio/video editing, etc.) they're still expensive and have only recently begun to come down in price, in part because there are some free versions that, while noticeably inferior, are still useful to a substantial population of users.
Can a monthly subscription support a collection of open source apps?
Netflix style. Maybe give donors the option to vote on features. I am interested in testing if there is a market for this. I've started writing clone android apps without the ads and freemium.
It might depend what you mean by "support", but in general, I don't see it.
For a non-open-source example, there's https://setapp.com/, but nobody has been able to explain to me how the economics are supposed to work. They're bundling thousands of dollars of software for $10/month, and then paying creators based on usage. Instead of $50 (a common price for these apps), once, developers will get a fraction of $10, repeatedly.
You'll break even if someone uses your app 1/2 of the time (of the Set) for about a year. If they use more apps or don't subscribe that long (which are pretty much the main reasons to use this), you're worse off.
Bundling, open-source, subscriptions -- at the end of the day, users simply don't want to (directly) pay what it costs to create software.
I think it is about lowering the price to increase users. The user experience is also improved as you just install and go.
Thanks for the info.
I think it can, and to your second point, Steam had a great system with Greenlight which let potential customers vote on what their games would look like before hand
> Can a monthly subscription support a collection of open source apps?
Kind of. In the WordPress world there are several companies that ask you to pay $50/year to get access to their entire portfolio of themes and plugins.
In theory, you’re only paying for support, since the code has to be licensed under the GPL.
For web apps, there is https://cloudron.io. Pay 15 per month to install apps and get automatic updates.
As a counterexample, Dark Sky is one of the few apps I pay for that has an endless array possible alternatives. The last time I looked all substitutes were ugly, clunky, bloated by ads, or all of the above. I pay so I don't have deal with an inferior experience in a frequently used app. So, while the details of what an app accomplishes may be easy to replicate... maybe there's always room for a premium experience?
Indeed I would say that with an app like Dark Sky the value is almost entirely in the experience, in a way.
The data is provided by the government for “free” but weather.gov is not as well designed and presented as Dark Sky and other weather apps.
This applies to a very, very narrow segment of software. If I have to run a server, then yes, I do indeed have costs. And those costs go up the more popular my app gets.
And if the app gains any kind of serious pickup, I am going to need a client service team to handle all the bull crap people will be flinging my way. (Seriously, how many open source projects die because the maintainers get sick of managing the issue board?)
If your costs are servers, your marginal costs are still $0 for the app itself. IMO, you're better off with a $0 app and a subscription model, than capturing the revenue once when someone purchases the app.
You are technically correct - yes, an app is not the server it is deployed on. But practically speaking, if I saw a spike in my traffic and got hit with a huge bill, I'm not going to say to myself, "gee, it wasn't the app's fault the server got really popular...".
I think you missed my point that this article applies to a very, very small category of software. As soon as you have a subscription model, or any sort of payment structure, now you need an accountant! Those guys don't work for free.
It's like, calculator apps and notepads that this philosophy applies to.
There are too many counter-examples for this to be true:
- Microsoft has been charging for years for Office before getting into the recurring payment model. The reason they got into the recurring payment models is that most office software requires server support and server support is a recurring cost
- There may be future competition but let's not forget that software still costs money to develop. ie. if an entrepreneur sees a market with 2-3 strong providers, they will hesitate to enter in direct competition. And why would they, their profit margins will be minute in comparison to the cost of building the software.
- The market may have software that is so good, that to catch up, competing players would have to spend years building the software (see point 2)
An example of this is Linux. It is free but it's rarely used by non-IT professionals. One can argue that Windows is effectively free these days, but it still charges a license and the price of a laptop is still affected by having Windows installed.
You've answered your own implied question. The price of MS Office is still high because it's in a category of its own, with no competition in sight. But all it takes is one well-motivated competitor to achieve feature parity and that would be the end of MS Office's cash cow position.
MS is already offering Win10 basically for free. Not because of competition, but because it doesn't cost anything to make it free. They can risk changing their business model without being forced to throw money down a hole.
It's tough to make that argument.
Features are only a small part of how users consider change - and 'parity' definitely doesn't cut it.
Office is a good example ...
StarOffice originally and then RedHat, Suse, Mandriva (and the previous versions), Canonical and a host of other commercial Linux vendors I'm forgetting, plus Sun, IBM, Linux Foundation, etc have all put resources into competing.
OpenOffice is feature equal for most users - but, it didn't make much of a dent.
For me the Windows 10 situation is partially about competition, but it's mostly about protecting their core market. For the most part they make money from corporate users, not users buying 300 dollar laptops - those are just market capture for the serious money. Any time you hear a parent say "Jonnie is learning MS Office" (not word processing) that's a future corporate buyer.
> Because all PCs ran Windows on top of Intel chips, there was no differentiation, and prices dropped.
Umm, no. PCs ran Windows on top of expensive Intel chips with little competition.
The lack of differentiation caused the prices of much of the other stuff that goes into a PC to drop.
And ironically, that other crap, like network cards and whatever, going into PC's, is differentiated.
That doesn't help because it's not differentiated in a way that the user cares about.
If processors were differentiated they way network cards, keyboards or mice are differentiated, it wouldn't make a difference, except making PC's even cheaper.
The same would apply to any digital content (zero-copy cost) which can be substituted. For example, why watch film A when film B is available more cheaply. Of course, it's difficult to have substitutions for entertainment (for various reasons)... right now.
However, I do think the long-term trend here is downwards as well and I'd use music as an example. With streaming services, the cost of music is much lower compared to purchasing music. The streaming service acts as an aggregator for different content and has an incentive to pick the cheaper content (with the opposing force of consumers wanting particular things). But really, if the quality is high and the music chosen to be played is automatic (via recommendations) the underlying artist is less significant (in general - I know people can be fans).
The marginal cost of software is not always zero. You may have cloud cost, support erc.
(2013)
Thanks! Updated.