Ask HN: Does Shareware still work in 2018?
I put a lot of thought into a software which solves a real problem and could be licenced on a per machine basis. (Small company pays less than a bigger one - i.e. pay per use).
To go for a startup selling licenced copies, I would need an investor but I don't really know how much the product will sell. The investors like to see hockey stick graphs which I can't claim.
I know there are companies that created open source projects on github and then have businesses doing custom work around it, but I don't know if this will work for me. I feel the product itself provides value and I would like to make some revenue off it.
Now I am thinking this is a good fit for shareware - i.e distribute freely - let anyone who feels it is worth pay for it. I can distribute with source so if someone wants to modify sections they can do so.
Please let me know your thoughts.
128 comments
[ 4.2 ms ] story [ 212 ms ] threadNot my parents. They started computers with buying software in boxes, but in the modern age almost all of their digital purchases are done on smartphones. Neither of them have any idea what Windows Store is. One of them uses Steam for games.
Not most power users I know, most of which bawk at UWP apps and the concept of a walled garden store on their desktop computer. Many power users are more likely to use Steam to buy apps despite it being gaming focused.
Anecdotally I see basically no group that would prefer Windows Store except out of ignorance or necessity. Buying on Windows Store means you don't get the Mac version if one exists. Admittedly, the same caveat applies in reverse to the more successful Mac App Store, but it's still something people have to consider.
Microsoft has been pushing Windows Store pretty hard for a long time. They came to my University back when I was in college, a bit after the Windows 8 launch. They talked numbers but only in terms of the whole Windows install base and not even Windows 8 adoption. Definitely no numbers on Windows Store.
To this day, I don't know if we've seen good stats on how much apps that were built and sold exclusively on Windows Store have fared. No doubt games have sold well - Microsoft can use its Xbox brand to deliver exclusives to Windows Store and not Steam - but I sincerely and absolutely doubt that success will translate to app developers.
I use probably about a dozen Windows Store apps with some regularity, though being in IT and dev, I obviously use a lot of apps outside of it as well. But while I rarely will install a traditional app from a no-name developer for fear it might do bad things to my system, I am pretty willing to try new UWP apps.
That's the key perk for someone just starting out/trying to make a living off small software offerings. The sandboxing offers a stand-in for trustworthiness. And of course, you don't have to stand up your own licensing servers, pay for bandwidth from downloads, etc.
I would like sandboxing in general, but as a feature of Windows Store it's definitely not enough to win me personally over.
I still remain unconvinced of the prominence of Windows Store and if I were to sell an app today I would guess mobile is the way to go followed by probably Steam or Mac App Store.
Sidenote: it feels like you're more likely to get annoying ad supported software from app stores too. Even the built-in solitaire is ad-ridden!
Note that while Android and iOS both have numerous examples of malicious apps in their store, AFAIK, Windows Store does not (though there are definitely ad-ridden nightmares in there). I found their reviews annoyingly onerous for a literally 50-line UWP app with a single function when I tried to submit something.
The biggest benefit of their sandboxing though is not actually the security limitations of what they can access, but how it's installed, and more importantly, uninstalled. UWP apps are one-click remove, and do not leave any lingering garbage in the registry, as they have kind of like a "registry diff" inside their own folder.
Apps like iTunes which are notoriously messy for install and removal I prefer over UWP because it's much easier to purge them safely.
I also find it a shame that Canonical instead of redoing the Desktop Environment for Ubuntu didn't invest that effort into just Ubuntu Mobile (or whatever it was called) and lastly, Mozilla almost had a reasonable thing, they really should of produced a Chromebook competitor first though. I yearn for a sane ChromeOS alternative that's fully open source and runs Firefox by default.
In a parallel universe Microsoft and Canonical team up to make a sane Android competitor that overtakes the market, and build the first Linux distro to be able to run Windows applications in a sane sandboxed environment.
Same applies to some of our customers that have done the migration to W10.
Did this make a difference for most users? No. But, when I wrote shareware, I usually got one or two 4 figure site licenses per year from companies for my product. It was just a sideline but the difference between pure donationware and "you're really legally required to pay for this" did make a difference.
That said, today I'd probably just do it as open source donationware though the money wasn't completely trivial for me at the time.
These days, yeah, but the term shareware started out as OP described. It later became synonymous with trialware.
https://asp-software.org/www/history/the-origin-of-shareware...
A lot of it was on the covers of magazines or you could order a set of disks from people like Atlantic Coast:
https://www.flickr.com/photos/edgecombe/4709059216
Put it out there for sale and test it out.
I literally went from using Kagi for ten years, to using Patreon. At first it's a decimating change because you're going from 'sale of a product' to 'literally giving product away free and telling people they can support you if they like'.
However, if your cashflow isn't too tough, one benefit is that Patreon is a LOT more stable and predictable than shareware ever was. You stop having boom and bust product releases and instead have months where you grow kinda briskly and months where you grow not so briskly.
You're not making revenue off the product, you're making revenue off 'I am the one who makes products such as this'. I've got suggestions for what people should pledge if they would've bought my stuff commercially, but it's entirely voluntary and I'm also MIT-licensing it all. I'm airwindows on github, and as a website.
Look at the statistics [0]: unless you are in the top 100 creators, you'll get far less than your regular FAANG pay.
[0]: https://graphtreon.com/patreon-creators
Publishers/creators can choose to show or hide the number of subscribers and their monthly earnings. This site simply is scraping that freely available information.
https://news.ycombinator.com/item?id=18468332
I would like this type of economy to flourish, as it seems like a very viable way to get good economic support for open-source. However, if this is a common occurrence, it seems developers will continue to be better off working on closed-source.
If supporting someone with X€ is any kind of "vote", it's a vote that they should get X€ more than they currently get for their work. It may be hard to make an educated decision about that, but it's impossible without knowing a creator's current income level to begin with.
No, think of each unit of currency as equaling 1 vote.
It's a vote that, between the various things that they could be doing, they should continue doing what they're currently doing.
For example, in a programmer's case, he could be employed by a company which will give $X votes. He could also be working freelancing gigs and earn $Y votes. He could start his own SaaS business and earn $Z votes. He could do open source work and earn through Patreon $W votes.
Only he will know where he stands in each of those markets, and, on average, you can bet that he will do what gives the most votes (i.e. money).
Not voting because you think he's got enough votes, is like not voting for a presidential candidate because you think he's got enough votes, despite otherwise wanting to vote for him/her. Imagine presidential voting was done such that you can see, in real-time, the votes of your preferred presidential candidate (without needing to vote for him/her) but not being able to see the votes of other candidates. That's pretty messed up, right? You see 10,000 people have voted for your candidate and think, "that's enough for him; he doesn't need my vote", and it turns out that other candidates have 100,000's of votes. It could be that the majority of the voting population would prefer your candidate, but because many of the people that preferred him didn't express themselves through their vote, someone else ended up being picked as president. That means the voting system failed, and that's why such transparency here is disruptive.
If:
- I believe that someone's work is worth 50,000€
- They currently get 50,000€ from Patreon
- Because their earnings are secret I give them another 1000€
- Because of my support, they refuse an offer for 50,500€
then the resource allocation system failed. The creator is now doing work worth 50,000€ even though they could be doing work worth 50,500€.
(For the sake of argument, I'm making some assumptions that are clearly wrong but required for markets to make sense as allocation mechanisms in the first place)
Not all voting systems are created equal. There are voting systems that permit using more than 1 vote, like this one:
https://en.wikipedia.org/wiki/Cumulative_voting
(If you go to the bottom of the article, you can see a blue box that says "Electoral Systems" with "[show]" on the right. You can click that to have the box display a variety of voting systems and related articles.)
> If you don't vote, you don't get to keep your vote and use it for a different election.
I agree that it's different in that there is no formal, delimited election time, like in other voting systems. Here it's like the election is always ongoing in various dimensions. Despite that, I believe the way capitalism works remains very similar to voting systems.
> If:
> - I believe that someone's work is worth 50,000€
> - They currently get 50,000€ from Patreon
> - Because their earnings are secret I give them another 1000€
> - Because of my support, they refuse an offer for 50,500€
> then the resource allocation system failed. The creator is now doing work worth 50,000€ even though they could be doing work worth 50,500€.
I feel that logic is weird. When you go to a service provider (like a freelancer), do you wish you could see what they earned monthly so you can decide whether to offer to pay them to do your job request or to just request of them to do it for free because otherwise you'd feel they'd be earning too much? I would think not. Why do you insist on deciding what to pay based on what their earnings are?
I think what you decide to pay them should only be based on the benefit you receive. What is "fair" is naturally decided by the economic system, not the individuals.
In the current market you will hardly find a private player looking for a return on their investment through profits/dividends; they will be more likely trying to cash out when you create enough buzz to get acquired or IPOed. That's a completely different game from the naive "sell licenses - make profit" option, but that's the game private equity plays in 2018. And no, desktop applications are not buzzwordy enough for that game, sorry.
That said, there's plenty of money to be made selling desktop products to businesses. You just need to approach it differently, get the actual product/market fit, grow your presence organically, listen to your customers, learn from your mistakes and so on. Don't get discouraged by bootstrapping - you don't need a fully functional product to see if there is a market for the problem you are solving. Create a minimal prototype, write a few articles showing it off and tune in on the feedback. Once you start hearing from real people trying to solve real problems with your prototype, it will be very easy to find which next features will make your product more usable and will inevitably bring more revenue.
Build the thing, and sell it (not necessarily for shareware, either). There are new channels for software now (like Setapp[0]) that you can look at, too, depending on who the customer and platform is.
Then you can make decisions based on actual sales, demand, market size, etc, and if investors are needed at that stage, then you have a waaaaay better proposition for them.
[0]: https://setapp.com/
Also it sounds like you are doing enterprise software. Shareware was more of a consumer pricing model. With enterprise you want to make it totally free for small scale grassroots adoption by champions and then monetize large scale adoption.
Though I'm sure it was always a game of chance like everywhere in business.
Wikipedia mentions that “by mid-1994 John Carmack had purchased two Ferrari Testarossas,” so it's possible he was both of the guys. But I'm sure others did well too.
Distribution was not a solved problem back then, and required significant capital. Shareware use a viral engine of growth (though this did mean that there were far more products for which shareware didn't work versus the handle for which it did).
I made one (1) sale, despite receiving e-mails from several people thanking me that they're happily using it at work, and that sale was to a company who wanted customization, and only ended up actually paying the invoice when they asked for another round of customization and I pointed out that they haven't paid their last invoice.
Donationware does not work for companies, I think - the bureaucracy required to make money move from the company to you will keep people from doing it even if they think you deserve it. If it is labelled as voluntary instead of a legally required license fee, it will also be hard to make it happen.
If you're targeting companies, and want to do a shareware model, you should:
* Make it easy to buy (with credit card etc.), but also provide a contact for volume licensing. If you're lucky, this allows employees to pay you for your software without having to go through approvals.
* Make it hard to use permanently without buying (beyond just a nag screen, e.g. blocking the save feature once an expiration time is reached)
Your goal isn't to convince someone to pay for the software. Your goal is to convince the person sitting in front of the computer that dealing with the bureaucracy to pay you is easier than not dealing with it, and if given the choice between a nag screen and the bureaucracy, the nag screen is easier to deal with.
So I'd imagine most people wouldn't risk it.
Except 'IT'.
I can't buy a $5 mouse at Best Buy. Or a Lightning cable to charge my company-provided iPhone. That will trigger long chains of accusatory emails from Accounting and Procurement. Software and 'IT equipment' has to be purchased according to the Policy, which to be quite honest I've never successfully figured out how to do.
It's come to the point that I'd rather buy small things (<$50) out of my own pocket than persecute myself by spending days going through some convoluted process and knowing at the end of it that we paid 3x the market price to get it from our 'Preferred Supplier'.
Lots of managers have credit cards and will pay the SaaS fee and ignore procurement processes. The the app becomes embedded in the company and the procurement team has to accept the reality of it.
Very clear what’s going on there once you get to the end. Somebody’s got themselves a nice cash cow sewn up.
At work we have something similar, where every few years the IT department tenders the 'preferred supplier' for general IT stuff. The logic is basically that by promising the vendors exclusivity, they will bid lower. So yes, sometimes that means we'll pay 3x for a keyboard compared to the price we'd pay at the local discount computer store; I think in general the prices for the stuff that is in the 'tendering portfolio' is very competitive, otherwise it's list price.
OTOH, we save money since the billing from the preferred supplier is worn-in standard procedure, and also employees don't waste company time searching the web for the cheapest/best/whatever keyboard.
If you have a business account, everything gets bundled into that account. Sure, if your org is 110 people big, keeping track is pretty easy. My last job was with a state university. They see thousands of purchases a day, and Amazon bins them onto 1 credit card, and in 1 account. For all 8 campuses and over 50k people.
So the preferred vendor means also integrating with your financial system. So budgeting is easier. But Amazon makes this so damned hard. Even being able to tag purchases would help. But I'm sure ol Besos figured out that makes him lose a dollar.
Every other online store sends you a proper invoice. But FBA vendors will send you sketchy invoices that make it clear they aren't paying their taxes. Which is bad, because that means as a business customer you're on the hook....
(And yes I know that's because of terrible design decisions in the intranet itself, but has anyone seriously worked at a large company where that wasn't the case?)
It's not even just the cost of the item itself - I often have to spend hours of time over a period of months to get something purchased, with multiple levels of management also having to spend time on approvals and chasing people up. The cost of all that time can easily exceed the cost of the item.
Sometimes I hate working for a megacorp :(
If you really want a legal way to get companies to pay, phrase it as "free for non-commercial use" and include that in the public documentation as well as somewhere in the tool itself. Non-commercial includes the indirect commercial cases that would apply to the tool.
AIUI that is wrong in USA and UK (don't know about other countries). At least if using the "non-commercial" definition used in IP law.
For example, you create an advert for free for an event that charges money, that's commercial. You give stuff away that impacts someone else's ability to sell, that's commercial. You do a free event, charge for snacks, it's a commercial event.
Use an app privately the results of which you use to benefit your employer, that's a commercial use.
It's pretty hard to answer the general question for "an app" rather than a specific app and associated action.
This is a really important point. Small companies I've worked at generally have had sensible and manageable procurement processes, but big companies have sometimes had mind bogglingly labyrinthine processes even for small purchases. There was once I had to raise an "Approval To Spend" and "on-board" a new vendor, and as far as I could tell the process changed 3 times while I was going through the "process", although I'm not completely sure because while everyone was telling me I had to go through "the process" no-one seemed completely sure what the process was, including those responsible for "the process". All very Kafkaesque. Anyway, it all comes down to your target market - if it is individuals you won't face this problem, but if it is businesses you will face the problem on a scale according to the size of business.
To start with I suggest you give the MVP product for free. The users will be paying you with their feedback.
On the other hand someone who has put their hand in their pocket is going to get a lot of my attention.
Right now OP is making an assumption that his product is useful. It's a guess. Doing the beta for free for a small set of users validates it with small effort.
In fact the first release should be as minimal as possible. Op can charge those users in the subsequent releases.
You should also tailor your business model to the product's market. If you have a broad market - home users, small business, large businesses, etc - some can use it for free, but some will still pay. If it's a smaller market - only small businesses - maybe only a few will pay if it's available for free.
You can also change the business model to be free software that drives paying for something else - like a cool cli tool, but a much more useful premium GUI. Or a free tool that's easier to set up and use as a SaaS.
Feel free to hit me up on twitter @m_herrmann if you have further questions. :-)
1: https://fman.io
2: https://blog.qt.io/blog/2018/11/15/python-qt-3000-hours-deve...
3: https://fman.io/buy
---
[0]: https://imgur.com/a/f9CATyj
Do you just nag or actually disable the product after a period? I was always surprised by the Sublime model, my experience indicates that if people can avoid paying then they will, even business customers.
1: https://fman.io/docs/plugins
Have you even tried to just sell licenses for it? If your software delivers value then that should convert better than a nag screen. Worth giving users a 30 day trial and seeing how that converts.
There are a number of approaches to doing that. After getting burned a few times paying $(100s) full-price for products just to find out that a feature I expected/needed was missing, I stopped doing that. I've since found some feature-limited stuff that lets me 'create-but-not-output' that I liked.
If it turns out to be a hit, you can change strategy. If the audience is limited, you can move back to shareware.
Just make a time limited trial, if they feel it is worth paying for then they will buy, if they don't then you don't owe them anything.
More specifically - allow to use the program without a license for X days and the cripple it in some annoying way until the installation is licensed. Activation should pass some machine details to the licensing server, which will then pass back them hashed and signed. The program should have the public part of the licensing key embedded, so on launch it would read the license, verify the sig and check that the license still matches the machine.
Also, for this to work well you will want to have some basic protection against reversing in place. It should not be possible to NOP the IF in the license check, replace the signing pub key or to side-load DLLs. This is a large subject of its own, but there are off-the-shelf solutions for this (called exe protectors)... though the con here is the perpetual hassle with false positive detection from anti-viruses and such.
PS.
I should add that "pay what you want" model doesn't work at all. It absolutely doesn't work for enterprise software and it doesn't even work with home users. It basically makes the software look not valued enough even by its own creator.
Ditto for the donation model unless the active audience size is in 100s of thousands. Donation model is not really a model to begin with, it lacks predictability.
Differntiating pricing tiers just by the user type - personal/home vs. business/commercial - works very poorly as well. People cheat. The only thing that really works as a price differentiator are the features. Pay more - get more. Also, charging for Windows Server installs 10x the normal price is a perfectly acceptable practice that works well.
If you have any questions - ask, I'd be happy to answer what I can.
So you will need to add some kind of restriction to make people pay for your app.
The restriction depends on the usage type of your app.
- If it is something that lots of people use everyday (like a text editor), you might get away with a nag dialog. Annoy a large number of people just a bit every day, and some of them will buy a license.
- If it is something that people use just once (eg. a file conversion app) you absolutely have to require payment before people can do what they want, or they will never come back.
- If is something that companies will want to customise, and you want to distribute source code, you probably need to charge a lot of money for it. Only few companies can afford to hire someone to write custom software, so you need to make sure you charge them enough that it is sustainable.
As an example Lichess (www.lichess.org) is an amazing service for playing chess. It supplanted services like ICC (internet chess club) and now competes against services such as chess.com and really stands head and shoulders above the rest of the sites. Now let's consider compensation. ICC had on the order of tens of thousands of users and charged around $50/year for what would have required extremely little overhead. They were undoubtedly making millions. Chess.com today is heavily ad-driven and works hard to push their users into a premium access model. No idea about their revenue, but they happily spend 5 figures a year in contracts just to get various players/streamers to exclusively play/stream on their site, and have an extensive paid staff. In contrast to these two sites, the founder of Lichess is able to gives himself a salary of $22,218, about $2k above the French minimum wage. It's not like he's then getting loads in options each year - that's it, that's what he gets for building software that beats out other sites making countless millions offering arguably inferior services.
It sucks, because I wish this model worked, but I think it simply does not. There are exceptions, but they are definitely the exception.
To be specific, it's importing tables from Excel to AutoCAD and Revit. It seems like this is something that AutoCAD and Revit should be able to do, but they handle it very poorly.
http://www.cadig.com/products/
http://dotsoft.com/xl2cad.htm
Licenses are ~$30-50 each and every company I've worked at has happily paid it rather than let employees waste time fiddling with the terrible native functionality.
I don't know how much the creators are actually making, but it seems to be enough to keep them going for the past 10 years or so.