The fundamental problem is that in our fully networked era software cannot _not_ be maintained. Security issues are inevitable, cloud/sync/store systems require servers, upstream APIs, libraries & operating systems are constantly shifting and breaking things.
So OSS aside (which has its own complicated economics) someone needs an ongoing revenue stream for that work to happen. Whether it's through regular release of paid upgrades (and EOL of old ones) or a subscription model is these days less of a fundamental separation and more of a question of cadence.
Take a look at the much-vaunted Campfire from once.com - there's been zero new features since initial release and I'll bet the cost of a copy come Feb next year when it's a year old there'll be a 2.0 for another $300. How long after that will 1.0 be EOL'd? So are you really 'buying once' for $300 or paying $300 a year just with the auto-renew turned off?
Sure they can. If you pay 100-1200x more for your cloud services than you need, maybe not. And I guess that is a theme, if you have subscriptions you don't really need to care about expenses because you have a steady income. Absurd? Yes, but that is how our industry behaves. I'm appalled at what people are paying for the cloud that literally could be served from a raspberry pi. Yes, you need redundancy etc. and that takes time and thought. But have you ever factored in the time and expertise required to manage the cloud? It is incomprehensible.
So, how many subscriptions do you need to be able to maintain your software? Is 10 enough? No? Conclusion, subscriptions are not sustainable?
Similarly, how many purchases do you need to recoup the initial development? You will hardly break even after the first 10 sales. But once you do break even every single purchase goes 100% to maintenance and new features. That is a very good position to be in.
Of course it matters what kind of product you have and how big of an audience you can get. If it is very niche product and you only expect 100 sales then maybe a one off payment isn't particularly appealing. But then again, maybe the value lies in support instead. Or, of course, a subscription.
Paying for upgrades is also fine! But without shafting your current users. Compute is dirt cheap. If you sell someone a piece of software for $300 you can afford 20 cents per year for cloud costs to maintain your relationships.
One of the problems of no-one being liable for security, is that no-one has the incentive to thoroughly separate software you really need to trust, and that you care whether it's vulnerable to hacking, from that which you don't. "everything needs a revenue for security reasons" would make sense if companies were giving an effective security guarantee and investing in fulfilling it, which they aren't. If software liability existed, most providers would recuse themselves from the most security sensitive features and we'd have a few big providers doing stuff like syncing, handling long term data storage, sandboxing, etc. rather like now with credit card data. At which point a lot of software becomes unimportant to update.
Agree. However the core problem with most SAAS products is: the companies are beeing greedy and the products fail to deliver enough value or keeep missing a certain quality bar to make you happy.
You could easily fix that by a base price and a maintenence fee (per year, since
first buy) which would make it easy to opt in/out of a subscription model. The fact that companies are beeing greedy and fail to deliver prevents this from happening.
Jetbrains is the perfect example of a subscription based platform that I will happily pay for without any bad feelings. I pay yearly, after a few years you do get a discount (so I am a happy camper). And if at any point I decide I don't want to pay it anymore, I keep access to the latest version of the software I paid for.
Yes, the multi-year discount and perpetual license are nice -- though the version is the one from the start of the year but that's certainly better than a lot of peers.
I was impressed that when Jetbrains raised prices, it seemed a reasonable amount, they gave a few months' notice, and allowed people to prepurchase three years in advance at the old rates.
I'm hoping IDEA and the all products pack continue to survive basically as-is. Jetbrains looks to have launched a couple overlapping enterprise SaaS products, the AI integration, and increasingly specialized individual IDEs.
Speaking as a SaaS company here: I get what you’re saying, but this isn’t really fair. We have ongoing development so sustain, financial planning to do, and investors down our necks. It’s hard to justify a business plan with lots of fluctuations and uncertainty, when we could also have a SaaS model instead.
I was referring to companies such as Adobe and Autodesk and i don't mind SAAS at all if the value is provided and a choise is offered. Guesss its hard balancing the investors vs the intereste of your clients.
The campfire model specifically is even worse; it becomes a liability to maintain that I can’t request of the vendor, but have to take care of myself. The burden of actively looking for vulnerabilities, and fixing them, is on me.
The idea may be venerable, but I don’t quite understand who would think you could consider a networked online chat server as finished…
> who would think you could consider a networked online chat server as finished…
What about a phone service? What about cable? WAN? Recent "smart" insanity notwithstanding, those things are not updated on end-user side; every now and then, some hardware or wiring might get replaced, but the infrastructure part on the customer side is generally one-and-done, and customers pay only for the actual telecommunication service.
Now that's just moving the goalposts. Comparing a chat server application with copper wiring isn't just Apples to Oranges, it's Apples to a Torx Screwdriver.
The chat server requires constant network communications with various devices that change over time, suffers from possible supply chain vulnerabilities it needs to defend against, talks to databases, mobile apps—suffice to say, there's a million interactions that may be failing in the future due to unforeseen changes in the outside world.
A phone line introduces but a fraction of the complexity on-site, and thus doesn't need as much caretaking. I mean, ...duh?
And even if you want to pull through with this analogy; phone service customers still have an ongoing subscription to cover for the eventual repairs and wiring replacement. Just think of the transition to fibre in recent years. You pay for the service on the phone line and the maintenance of your infrastructure. It's a bundle.
Yes, when you purchase, you get that version + free updates and support for a year.
After the first year, you can still use the software, but if you want support or to get the newest features and update to the latest versions, you can pay to renew your support period.
For the pricing, usually a renewal for an extra year is about 10-30% of the original license price.
Yes. That is why I want a Hardware Product with Software Model like iPhone. I want my NAS / HomeServer with all the OS and Software included. With Security update for at least 8 years. ( Synology is the closest thing we have got )
Apple could have done this, but they were so focused on their Services Revenue they want everyone to subscribe to their iCloud. They could have allow the NAS / HomeServer to backup to their cloud.
I think the subscription model needs to be refined.
500$/year (invented amount) for a Photoshop subscription that I’ll use 3 times in a year? But when I want it, I need photoshop, not $SOMETHINGELSE. Sell me a $5/day license and I’ll be happy to pay every time.
Subscription model is ok for heavily used software, but doesn’t properly address whatever I use every now and then.
There are also SaaS products with usage based pricing. It depends on what SaaS or software or product it is. Different pricing model works for different things.
Fully agree. I am a software developer myself. I can't imagine making a living by charging my customers just once, ever. It just doesn't make any business sense.
Software needs to be maintained and developed, and just following the evolving technology (operating systems, libraries, environments) is a lot of work. On top of that, you need to fix bugs, provide support, and yes, develop new features. This is not feasible with a one-time purchase.
For some reason many people like to play a game of pretending. The person buying pretends that it's a one-time purchase. The person selling pretends that it's a one-time payment. But then it turns out that a new version comes out every two years or so and you need to upgrade, so you end up paying. It's a subscription, but everybody pretends it isn't.
For the software I use and rely on, I would much rather see a subscription model, which is sustainable.
It is not a subscription though. You are not hostage. You don't have to upgrade. You can keep using the old version if it serves your needs.
Imagine your stance as a customer instead. Imagine being hostage to every single piece of software, completely at the whim of another business that can increase the prices and/or pivot away from your needs. You have to keep paying regardless of whether you want or need any of the features being introduced. Does it make business sense to allow for that risk?
Both models work and it is getting tiresome to pretend that anything but a subscription model isn't sustainable. Yes, everyone gets the appeal of subscriptions from the perspective of selling software. But it is pretty disingenuous to not acknowledge the other side or pretend that the other doesn't work. You can't be that lazy with anything in life.
> have to keep paying regardless of whether you want or need any of the features being introduced
This here is the misunderstanding. A better way to phrase it is that you have to keep paying if you'd like the software to work.
On most platforms just keeping up with the OS and library changes requires plenty of work and it is unreasonable to expect someone to do this without constant income.
Now, when I'm a customer, I see things differently: if I use the software and I intend to rely on it, I want the developer to survive, so that they can maintain and develop the software in the long term. Paying a one-time fee will not result in a sustainable business, so the software will eventually die, disappear, or get acquired, none of which is good for me.
Downvote me all you want (HN can be really narrow-minded at times), but when I look at new software that I intend to rely on, one of the main things I consider is whether the developer has a sustainable business model. If they don't, I don't want to invest my time into learning, migrating to, or integrating their software.
Incidentally, I've been running a solo SaaS business for a number of years now. It's B2B only. Seeing the knee-jerk downvoting reaction here only convinces me that not providing any B2C plans was the right choice. I just give my service away for free to hobbyists.
Yes, and often when the OS has had a major revision a new version comes out as a paid upgrade (or more often lately, a purchase includes upgrades and maintenance for a certain time).
Though it is work the developer has to do regardless, if they want any new sales. And maintaining the software is orders of magnitudes cheaper than develop it from scratch so it is kind of weird to latch on to the maintenance aspect. If your sales paid for all the initial development then it will pay for maintenance. A bigger worry than the maintenance aspect is, will I recoup my initial investment?
The cost will always be arbitrary, software costs nothing to "manufacture" so any one-off payments or subscriptions will only match the costs associated costs for a certain scale.
I agree with what you say, sustainability is crucial. Doesn't follow that subscriptions are the only solution. Subscriptions-is-the-only-sustainable-model is the narrow-minded view I mostly see on HN. Undoubtedly because the audience are developers.
> So OSS aside (which has its own complicated economics)
Assuming you're actually referring to free software and not open source software, it really doesn't, though. It's straight up better in every way than service-oriented software and commercial software in every way... except actually compensating developers. I'd work for a pittance writing free software if there were any institutional support for it. But who wants to kill the golden goose, even if it means our lives would all be greatly improved?
That's not "complicated", this is the opportunity to make a shit ton of money by charging people for software despite insignificant marginal costs. Even if it means humanity writes the same goddamn software over and over and over and over and over and over and over again, mostly shittier than the last iteration.
Many SaaS disappoint after a while. They suggest you are paying monthly and benefit from ongoing development in return. Instead prices get increased, and essential new features are locked behind additional pricing tiers. Premium, professional, enterprise, what’s next? The user interface becomes an advertising app for the upsell. It’s an abuse of trust. So the problem is not so much subscriptions and SaaS themselves, but the business practices they enable.
So OSS aside (which has its own complicated economics) someone needs an ongoing revenue stream for that work to happen. Whether it's through regular release of paid upgrades (and EOL of old ones) or a subscription model is these days less of a fundamental separation and more of a question of cadence.
Take a look at the much-vaunted Campfire from once.com - there's been zero new features since initial release and I'll bet the cost of a copy come Feb next year when it's a year old there'll be a 2.0 for another $300. How long after that will 1.0 be EOL'd? So are you really 'buying once' for $300 or paying $300 a year just with the auto-renew turned off?