Ask HN: Are subscriptions the doom of the tech sector?

I sell software with an inflation pegged maintenance component. It acts as a perpetuity and as interest rates are so low it makes the net present value (NPV) of the perpetuity very high to the point it dominates the initial purchase price. If real interest rates were not so low I could invest the initial purchase price and use the recurring revenue from that investment to cover ongoing costs.

People’s expectations are anchored and it is hard to sell against that, they expect to pay X for the software price with a Y yearly maintenance. So we stick to industry standard where possible. The standard varies by country so we have country specific setups.

I’m annoyed by subscriptions as well but mostly because I dabble in a lot of things and it’s expensive to have a ongoing subscriptions for something I may only use a few days a year. I generally try to get perpetual licenses and stick to old versions.

The software I make is very niche and domain specific, users spend 70% of their work time on it and are vastly more productive with it. The end users also hate change so they’re very happy to keep paying so that someone will care about their problems and they don’t want upgrades for the sake of it.

I personally really enjoy the incentives that subscriptions create for developers. If you expect me to keep paying for your software then I expect you to keep providing value.

The other nice thing is the continuous revenue stream ensures that developer can continue working on enhancements, lowering the risk of it becoming abandoned.

I also get to be on the latest greatest version at all times.

> If you expect me to keep paying for your software then I expect you to keep providing value.

If you make paid software part of your workflow, you're going to keep using it whether they keep providing value or not. When someone else captures your workflow into their own infrastructure, there tends to be a high cost associated with extricating yourself.

This is why I think public cloud and open standards for connecting to them will be so important going forward. Its okay for a developer to capture your workflow into their software, but it should be possible to run it in your infrastructure. You can still pay for the software subscription style, but use your infrastructure and pay metered

Exactly - IMO any subscription service that is seeing flattening growth is likely focused too much on "growth hacking." Changing the text on your paywall can only do so much, as those subscribers will still churn if the product doesn't improve to meet their needs over time.

Having sold both subscriptions and CDROM based B2B software, I can tell you with confidence that it was easier to sell a $2,000 software package once, then a $200 monthly subscription. That was my experience in the B2B world about 20 years ago.

There’s a definite niche in the $5-50 a month range where people can buy b2b software rentals that help their job without much paperwork.

But once you hit consumer signups or bigger prices it gets much harder.

Many companies require receipts for anything > $50 on an expense report. Under that and employees will buy/try tools that help them with their job, and not have to justify it to anyone.

Subscriptions are great model. If something really gives me value I will keep paying for it in return for newer content/features.

Why would I spend $200 on some software one time when it will become obsolete within a couple years? Or maybe I only use it a few times.

I’d rather pay 10-20 a month for one month. And never use it again. Or keep using it.

In this economy I went ahead and cancelled a few subscriptions and kept some. I feel zero obligations and it’s great.

TLDR if it brings you value use it. If not don’t.

Subscriptions for simple usage only make sense for a true service with an ongoing cost to the provider; cloud storage, email, movie streaming, etc. A subscription for a general-purpose application is incongruous; you’re purchasing a finished product with no ongoing costs, like a pair of shoes or a book, and it should cost a one-off fee that reflects the cost that went into producing it.

What subscriptions are theoretically paying for is *updates* — the ongoing cost of supporting new technology, fixing bugs, and adding features. Requiring additional payment for additional work is totally fine, and some apps successfully use this model (Agenda (https://agenda.com) and Nova (https://nova.app) are two that come to mind, and I’d be interested to hear of others).

Requiring ongoing payment just for continued usage is predatory and leads to a perverse incentive where developers don’t actually need to provide updates or new features; they’ve got you as long as you still need the app’s basic functionality. It’s rent-seeking behaviour and it will prove to be as toxic in the digital realm as it is in real life.

From a consumer perspective, I disagree.

If a program costs $200 but is something that is used very infrequently and doesn't worth $200 to me - then it's poorly priced or best sold (or re-sold) as a service rather than a program. Subscription here is just a service fee in disguise, so why not make things straight? Let me pay $20 one-time to process whatever I want it to, and be done with it.

And if something works for me today and I don't foresee any need for updates then what exactly I'm paying for? Developers will to continue development, that I may or may not need and may or may not care about? Doesn't sound right to me, unless I feel benevolent and charitable for some reason. I get the subscription model if something needs active regular maintenance. Drivers do (heck, I'd buy that rather than seeing companies "we no longer support shit, go buy our new hardware instead"), financial software does, a weather app does. But, say, a camera or note-taking app would certainly continue to work without any maintenance until I upgrade phone, or OS, or want some new features (and then I can buy an upgrade or figure out how to fix it on my own), so I won't buy a subscription (or very much hate doing so if it's my only option besides simply not having something I need at all).

> Why would I spend $200 on some software one time when it will become obsolete within a couple years? Or maybe I only use it a few times.

I spent 80€ on Strongbox, and I think it was a similar amount for canarymail lifetime.

Strongbox is a keepass2 client, canary is an email client. I expect this software to work without any issues for the next 10 years. Neither keepass2-databases nor imap/smtp will be meaningfully developed in the near future, therefore the maintenance for these apps is probably close to zero.

Okay, actually, canary got Sequoia VC money recently and switched to a subscription model. Since then, they added stuff that is useless to me, but I don't care as long as my lifetime licence still works. The core functions likely won't change.

I've grown weary of subscriptions. But I also think people who don't want to pay monthly would definitely be opposed to paying the equivalent one-time purchase.

Remember, Adobe Photoshop was like $500+ before they made it available via subscription.

When I used it, it was $300 IIRC. And I happily paid that price, knowing it would work for me for years to come.

In the same vain, I paid $200 for Logic Pro X three years ago and it still does what I paid for. I’m about to spend $300 on Final Cut Pro.

If Logic Pro went subscription only at, say, Photoshop prices at $20/mo, that would be the equivalent of $720 so far. Even though Logic is seriously awesome, I would not pay it and just go with something else.

Subscriptions do somewhat align developer and user interests. But it somewhat depends on the application. Some small utility that I use for some task a few times a year? I'll be pretty dis-inclined to buy a subscription for that. Something that's a core part of my workflow (or that I generally get regular utility from)? I'm much more open to a subscription--depending on the pricing of course.

> Subscriptions are great model.

Not for the user.

I give $0/yr to digital subscription services with zero marginal reproduction cost (SaaS, video, etc.) and somewhere in the range of $250-1000/yr for pay-to-own software.

Reasons:

1. Pay-to-own software is almost universally of a higher quality than rental software

2. I want to host my own software and data, so I'm not instantly screwed if the software company goes under or unilaterally decides to change the way they do business

3. Subscription management is a pain in the ass (in terms of the effort required to track down subscriptions, cancel them, make sure you aren't still getting charged, etc.) whereas pay-to-own requires close to zero mental overhead

The primary risk with pay-to-own software is that it stops being maintained and the OS APIs the software uses gradually go out of date, but this is much easier to address than the corresponding risk for rental software: that the company folds or the product gets cancelled, and you lose everything instantly!

I know lots of people complain about subscription but for most services this seems a realistic pricing based onto the fact these services have recurring costs in the form of infra, maintenance, employees?

On the other end the service is providing you recurring value otherwise you would just drop it.

> these services have recurring costs in the form of infra

Only because these companies have forced us onto their cloud platforms in the interest of justifying/enforcing continuous payment. I don't want to use Adobe's crappy "infra", for example, and neither do 95% of their customers.

Hosting your own infra is still a recurring cost given that hardware has an end of life. These servers also need employees to maintenance then and electricity to power them.

There's really no such thing as pay once use forever for anything physical.

You're just paying an estimated lump sum for the subscription revenue it would produce over it's life time

I don't really agree with this take.

You're not wrong that owning your own equipment can be viewed in a similar manner as renting with the lump sum payment amortized over time.

But... when you do it yourself you can much more easily tailor costs that make sense for you.

I have essentially stopped paying for almost any online subscription SAAS product that I can replace with a self-hosted version or a desktop install. I have cut more than 200 bucks a month out of my budget (fairly easily, honestly - replacing recurring costs for things like evernote, myfitnesspal, photoshop, google photos, calendly, ifttt, etc).

I threw some old desktops into a microk8s cluster in my basement, and spent about $1000 on a decent NAS + drives.

I broke even in 6 months. Two years in and I've saved more than 3 grand.

The subscription costs are HUGELY inflated for my needs, and there's no way to bring them down in most cases, because SAAS companies lock features to tiers in a way designed to maximize revenue.

Basically - my strong opinion is that SAAS is so popular right now because it's a terrible deal for most buyers. Companies make more money with less work.

I think it's taking buyers a minute to figure out JUST how shitty the deal they're getting truly is. But trust me - SAAS is a shit deal in most cases.

The only time SAAS isn't a terrible deal is when you can buy a single month, do the task, and cancel. If you need to be using a tool repeatedly over time - you're WAY better off buying up front than you are leaking money every month to a SAAS company.

What kind of infrastructure do you need to run Adobe Photoshop? Other than your own storage I can’t think of anything else. I used the same version of PS for years before they switched to a subscription model and forced users to upgrade.

This is the real problem, Adobe made kick ass software that was so good it cut off their future revenue stream, so they decided to rent it out instead of letting people own it. Fine, that’s their decision, just like it is mine not to continue paying for something I use once a month as a photography hobbyist.

Adobe sent an update that made your old, paid for version of PS stop working?

We used to buy software that ran on our machines. We expected a certain amount of support. Eventually you would have to go purchase version x+1 of the software.

Wonder if it's possible to build a business of reselling subscriptions as one time payments. A sort of reverse loan.

I pay your subscription forever but you have to pay me a lump sum right now.

Welcome to the world of finance! All products are either that, or the reverse. Have fun!

> However, they weren't even bringing enough value without subscription

In theory, anyway, it's a good thing that they get cut if they are not bringing enough value. Sure, downturns are a time when this happens in earnest, but you'll always get people complaining about the service and claiming "We could save X dollars here" if it's not adding value.

In practice -- sure, there's probably a lot of services that are low-value and got "floated" by the general upswing in the last couple of years. But to me, this is probably true of non-subscription applications and physical goods as well. People are going to be thinking twice about buying those as well during downturns, so the impact isn't limited or specific to subscription-based models.

> However, they weren't even bringing enough value without subscription or with ads

This value is objective? Therein lies the issue with this argument. For some people, a spam blocking app is a waste of $2/mo. For another person, it saves them a ton of headache. I'm currently subscribed to 12 apps on my iPhone and I have 20 others that I let expire over the past year - I pay for them because I enjoy them and would like to support the developers and to continue using them. I have a ton of paid apps from years past that are now just junk, because the $5 the developer charged 10 years ago couldn't support development for 10 years of service. Subscriptions are the answer to that problem.

Subscriptions will naturally go away if they don't work, but that also means those services/products go away as well.

Also where is this data on flattening subscriber growth that you have access to?

I'm seeing a lot of basic kids apps on ipads that should just be 'pay once' but instead its a bullshit subscription specifically because they know mum and dad will forget about it and keep paying for months after the kid has stopped using the app

For my business, if it’s a subscription, we actually invest money in building out the capability in house because subscriptions increase over time

Like GitHub for example, I love it, it provides value, but it costs a ton if you have a lot of devs, and you can spin your own git server

If you think the value of GitHub is in the git server, you’re not using it right.

Whenever you have a question of this sort, it's important to note that:

1. I personally don't like XYZ

2. My echo chamber doesn't like XYZ

3. Consumers at large don't like XYZ

4. XYZ will be the death of a company or the industry or the economy

are all very different realities. In this particular case, I'm pretty confident in saying that it is actually 1 & 2. The tech sector will be just fine.

I suspect that quite a few subscription-based companies are going to have a hard time. They have no realistic alternative revenue models so they might fail.

However, consumer-facing subscription models are not the majority of the tech sector. Most tech companies have other ways of making money, and the tech sector as a whole will be fine.

Depending on the numbers, I feel more comfortable signing up for a subscription than a one off payment. With the one off payment it’s inevitable that they will bump the version number and eventually cut me off. With the subscription I can pay the smaller fee for as long as value is provided.

If they can cut you off after having paid a lump sum, you never owned it to begin with.

What's the alternative, if those companies need subscription revenue to exist?

And if customers are unwilling to pay more than a nominal fee for one-time purchase software.

No (betteridge's law). But the tech sector is the doom of the consumer :)

There's plenty of folks out there, doing creative work, using legacy systems, that don't receive security updates, because you used to own your software when you bought it.

Personally, I think we are about to see a huge resurgence in self hosting and returning to "old ways" that worked before subscriptions.

One great example of the anti-pattern: Look at how GitHub has shit the bed since MS bought them. Tons of QoL developer features were pushed behind the enterprise paywall, actions was constantly down and having issues, and they did not innovate anything.

Beware robber barons who innovate nothing and expect you to pay more.