Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: Why is everything a SaaS product?
141 points by kevivni on Sept 15, 2022 | hide | past | favorite | 179 comments
I understand the advantages of Saas - transparent updates, cloud based, redundancy, subscriptions, scaling, unicorns etc.

Why don't we have more businesses following the JetBrains model specially if the product is not a service that needs to run 24/7. I buy the product once and use it for perpetuity. If I want updates, I pay more.

As a consumer, my data does not leave my perimeter, my data is not sold or used for ads and I am not hooking into a subscription that I am going to forget soon.

1) Personal photos and videos backup and viewer. - Just give me a cheap cloud for backup and a desktop app for viewing. 2) Personal budget - Just give me a desktop app that connects to my different accounts and gives me overview.




1. Recurring Revenue

In some sense, every business dreams of wanting to be a recurring revenue business.

- Gyms want you to pay monthly, and discount annual plans (and sometimes discount again if you try to cancel your annual plan)

- Restaurants want you to come frequently, 6th time is a free sandwich or coffee

- Bars want you to come to happy hour, at the same time, every day

- Content creators often have a daily, weekly, biweekly, or monthly cadence

SaaS is all about recurring revenue. In theory, SaaS aligns incentives because you can quit once the product or service deviates from your expectations or no longer serves your needs. Of course, in practice, SaaS companies want to keep the monthly / annual drip coming and may make it hard to export your data or switch providers. Local first software could be one antidote to this (https://www.inkandswitch.com/local-first/) and perhaps there are others.

2. SaaS loosely resembles old-school semiconductor VC investing

SaaS is similar to old-school tech VC investing because the marginal cost is low but the upfront cost is somewhat high / fixed. A lot of investment houses think in terms of recurring revenue as well (e.g. your mortgage is a monthly payment). So if you want external funding, show a premise of recurring revenue!


You are correct, but I also hate that "recurring revenue" is the right answer.

Not only does a subscription fee act as a constant resource drain (as opposed to a one-time charge) but it further pushes society into the zone of "you won't own anything and like it".

Human life and the ability to interact with others and our environment should not be reserved to those that can sustain a high velocity of money.

Focusing on a consistent cash flow makes financial management easy but makes everything else suck. It optimizes for the wrong metric.

UPDATE: You can also look at subscription fees as a tax on productivity. If I pay once for Adobe software I can amortize that over the entire time I use it, and eventually the monthly cost will reach nearly zero. With a subscription fee I can never amortize the software, meaning the fee is effectively a tax. And if I stop paying the fee I completely lose my ability to use the product.


On the other hand, you can regard one-time fees as subscriptions in disguise, because hardly any useful software can be used forever. File formats change, collaboration changes, operating systems change.

Yeah you could run your bought-once MS-DOS accounting software in a VM forever, but most useful software has a clear expiration date, especially if you try to collaborate on data.


Subscriptions are almost definitely more profitable.

E.g. Photoshop is now $240/year. The last price I could find for a version you could actually buy was $500. Assuming someone uses it for 4 years (which might be short, the few people I know who use PS latch onto a version for forever), it would be ~$1,000 for a license.

A lot of stuff would be fine for about a decade. That would be ~$2,500 for comparable income.

Subscriptions tend to be dramatically more expensive than buying software was. $5/month over a decade is $600. A lot of stuff wants to charge $5/month for something I wouldn't pay $600 to own for a decade.


Even knowing that, I much prefer software I buy and own forever. I have too many subscriptions already and I feel icky knowing there’s all these leeches draining money automatically from my bank account every month, with not enough visibility.

A few weeks ago I was looking at iPad apps to help me learn piano. They all want a $15/month subscription or something, and I hate it. I know what I’m like - I’ll subscribe then forget and end up spending $100 for an app I use twice. I ended up getting an in person music teacher instead. It’s more expensive, but it feels better spending the money like that.

Software I buy then own (or even, pay for once a year for updates like Jetbrains) feels better because each renewal feels like an active, conscious choice.


I have a file manager for my phone. (File commander) They have a pro version that I would gladly pay for but the only way to get it is as a monthly subscription. So I just don't buy it.


Why would a file manager that runs locally and doesn't require tons of flashy features have a monthly subscription?

It's so stupid. I prefer the Jetbrains model of pay subscription or license, but subscriptions become cheaper as you had more time subscribed (IIRC, 30% discount on 3rd year)


For apps like those, I just find cracks online and install those apps with all pro features included.


Consider subscribing, and then cancelling, if the monthly fee is worth a one-time use


Every business is moving to SaaS because they make more money off of the customers that way. So you can't really spin it as being better for the customer. It's better for the investors in those businesses, and worse for the customers.


I understand this perspective and sometimes I want to agree with it, but it requires some kind of competition/innovation/refinement process to exist. One time purchases means the provider needs to entice you to upgrade one way or the other - compatibility with new formats and paradigms, better X, something. Subscriptions don't have that incentive once they're relatively entrenched as de facto standards. Granted, one time purchases also suffer from hubris at that stage (see unnecessary UX changes to Photoshop, Windows, etc.) but it's more obvious with SaaS due to the sunk cost fallacy.


I would guess that in many cases, you could use it longer than the SaaS would exist.


I actually don't mind subscriptions if they are structured correctly. If I had the choice of buying a software package that I could expect to be useful for 5 years (even good software becomes out-of-date over time) for $100 or paying a $20 yearly subscription; I would rather have the subscription.

The problem is that most vendors want you to pay about half of what a package would normally cost for a perpetual license on a yearly basis. Anything over 2 years becomes more expensive.


I do not like the incentives it place on the supplier. With a software package, they are required to compete not only with other companies, but also themselves from years ago. If they do not make worthwhile improvements, users will not upgrade.

SAAS only has to compete with other companies.


Yeah this. I don't mind SaaS in theory. In practice it is just a more expensive and invasive way of a way to get your software.

Too many apps think 10$/mo is reasonable. Thats 120$ a year. Absurd.


It also puts the burden on the user to remember to cancel things. If I stop using my bike, it just sits in my garage. If I stop using the gym, I keep paying for it until I jump through their hoops and cancel my membership.


You would think lump sum is better. Monetizing sooner means you can put that money into other investments sooner that will generate greater return for you overall than having it dripped out to you at a fixed schedule.


No, because the lump sum can't be depended upon, it's unpredictable. Will you have 50 buyers this month or none at all because it's holiday season? Or maybe you'll saturate your market niche. You can't plan anything ahead.


You can say the same for whether or not you will net a subscriber next month or have cancellations. What I'm saying is that if your license is priced correctly, that revenue the consumer pays up front can go into a fund that exists to perpetually support this software indpendent of having to maintain an appeased subscriber base in perpetuity. It's this idea of having a growing pile of money that exist because of what already happened, versus having constant revenue that must continue to be depended upon indefinitely.


It’s much easier to get users to stay subscribed than it is to get them to actively buy every period.


You don't need to get them to buy every period if they paid a lump sum up front. That's the whole point. Money is capable of growing for you, so take advantage of that fact in your business model and unshackle yourself from subscriber dependency.


In the perfect world where every monthly subscriber would be willing to pay 20-100 times (or more) the amount for a one-time, up-front purchase, sure. But that's not reality.


I mean, that's how software used to be? Let's exclude anything cloud-based and assume no marginal operating cost of running purchased software.

Adobe Photoshop cost around $1000 in the 90s, and now it's like $25 a month? Back then you paid a discount to upgrade to latest version, and now it's generally included in the subscription. For professionals, the amortized subscription price is probably the same as the flat one-time free plus the upgrades, if not cheaper.

Some of the really expensive software I paid for in 90s and early 00s still has a flat one-time fee, such as Visual Studio or Final Cut Pro. Adobe and Autodesk have moved to subscriptions, but if anything that has made them more accessible to hobbyists. They were both charging ludicrous sums back then so the normies resorted to warez downloads.


Yes, that's the old pricing model. I'd bet that there are more subscribers on the new subscription model than the old flat-fee model.


> In some sense, every business dreams of wanting to be a recurring revenue business.

As do individuals. The only thing preventing it before was the expense in enforcing licensing terms. Today’s software and cloud hosting makes that much more trivial.


SaaS is also quite easy to churn out as a lone SWE, think most of us prefer to make our own income than toil on legacy CRUD apps all day.


Repeat business at your local pub is recurring revenue. SaaS is just collecting rent.


For 1), what exactly are you expecting to happen to enable this cloud backup to continue working?

You pay once for someone to set up a hard drive in the "cloud" you can put your photos on?

They charge you enough for a lifetime of transfer bandwidth?

What happens when the hard drive dies? Do you want redundancy so that if one drive dies the photos aren't lost? Who pays the tech to go change out the first drive when it fails so it can recover before the others in the array also die?

What's your personal risk aversion for losing this data? Even a two-drive RAID array can lose all of your data in some scenarios.

For 2), you want an app that connects to all of your accounts. How does it do that? Usually by using web APIs or a third party interconnect like Plaid.

What happens when they bank you use updated their API? Who is going to provide you with updates to keep all of these account connections working?

---

Both scenarios you describe have ongoing maintenance. Things weren't SaaS when they existed solely on your local computer and continued working without constant maintenance, but your two examples that come to mind show how our expectations of technology have changed and we expect things to work online, backed-up, and to keep working even as integration partners, OS vendors, etc. update their systems.


I think the point is that you should be able to use one commodity service to handle your data storage and syncing needs. (Or maybe a couple different ones, if you're worried about your account getting terminated.) But that doesn't mean you should have to be nickel-and-dimed by subscriptions for every individual application you want to use.


Sure, I could build a photo app for a one-time purchase that lets you add your own credentials for S3, B2, etc.

What’s the market size for users who will set this up themselves (versus going for a competing app that bundles the cloud storage), how many platforms will I have to develop for (Windows, Linux, Mac, iOS, Android?) and will they promise to never complain when an OS update or API change breaks the app?


We did that for decades. Many products still do. The model is plenty well proven, and the markets well demonstrated, so this kind of rhetorical “well what do you expect?” stuff is unconstructive.

There are reasons a business may prefer to provide their product as an SaaS, but it’s hardly some necessity.


We did cloud backup as a single-purchase pluggable-into-any-provider desktop application for decades? I don't recall anything like that.

The features OP is asking for in their not-SaaS product wishlist just plain didn't exist before SaaS. There was no cloud backup, there was no automatic account connection in your budgeting app. It's not obvious to me that things that require a connection to an always-on server (and some features really do require that) could be sustainably delivered in a one-time purchase model.


Microsoft Money and Quicken absolutely had account syncing abilities in their one-time purchase software license budgeting apps.


Ah! I stand corrected.


Those decades didn't include the features, functionality, scalability, or user-friendliness required by the BILLIONS of people online now using services that weren't able to use them when they required you to do your own setup.


If that's the case you should be able to build a non SaaS competitor and take their market share. In the end if SaaS is really bad for the end user, competition should take care of that should it not?


Yes, but that market mostly existed before everyone expected every app to be online and connected to other services, and since then the market has moved to SaaS. So if customers will pay more for a subscription that is inclusive of these details and maintenance, why would I choose the option that leaves me with less revenue and fewer customers?


Software should be maintained in an almost private trust model imo. Pay upfront for the license which generates a bunch of money for the developer upfront. That money along with all future revenues from new licenses are put into a trust that grows over time and is used to pay salaries for maintainers and to pay for server hardware. Maybe if you own your own hardware you could even have a side business of leasing your spare compute capacity and that perhaps could cover its costs.


Nothing is stopping you from having such a model, but why would everyone need to do this?


Jetbrains can offer a perpetual license[1] in theory because old versions don't continue getting updates. They spend (relatively) little on "supporting" old versions -- mostly just the cost of the download servers.

The two use cases you described don't work that way. Both of them have cloud components, which means they need frequent security patches, and they likely also need bug fixes and other updates.

"Connects to my different accounts" by itself is a huge undertaking that most people outsource to Plaid or a similar vendor.

So to answer your question: a lot (not all, but most) SaaS products are services because they are services. They are updated incessantly by developers and have rolling updates.

1. https://sales.jetbrains.com/hc/en-gb/articles/207240845-What...


To me, the keep your old version forever was a thing Jetbrain did to "soften" the blow when going from buy once, use forever.. to buy every single month.

Still, one of the most valuable softwares I pay for.


A lot of people really didn't like the fact they were going from having a perpetual license to having a subscription. The reality is, the way it worked out is basically the same as before but I could pay monthly. Since I paid for my own license instead of allowing an employer to buy it (so I could get IntelliJ for all languages instead of getting a language specific version) that was nicer.


Completely understand the comments around valuation of a recurring revenue business and ongoing maintenance. Those are all true and accurate.

I've also been thinking about this from an ecosystem standpoint that it is now very difficult to develop a software product that is not SaaS.

If everyone else is SaaS, and you are not, your business will be:

- Valued lower than equivalent revenue SaaS businesses, because revenue is lumpy and comparatively unpredictable

- Have greater difficulty raising capital, because investors aren't used to valuing non-SaaS businesses any more

- Have greater difficulty hiring engineers, as they know the next job they have will likely be SaaS related

- Slower sales cycle due to higher upfront price required (IE - Doesn't automatically fit on an employee card for land and expand)

- Higher customer acquisition costs, as most customers, other than those on HN, are used to the subscription model and prices, and would encounter sticker shock at the high prices required to make a one time purchase work


> most customers, other than those on HN, are used to the subscription model and prices, and would encounter sticker shock at the high prices required to make a one time purchase work

This one has basically come full circle last year. [1] Many "customers, other than those on HN" have been burned by subscriptions in the past and are rightfully using pricing models assuming a 10x service spike in the future. We've found it easier to sell one-time units for $25000 than subscriptions of $250/mo

[1] https://news.ycombinator.com/item?id=28372532


I don't get it. It used to be software licenses were being sold for like $10 or $20 dollars. Now $.99 a month is somehow more palatable than $20 for life? I don't believe it. I think perhaps organizations got too large in their bureaucracy. Times were probably leaner for the business folk when software were sold for $10 and $20 or even $120 dollars for Adobe products.

All that being said, it stands to reason you could structure your business to be lean enough to ask for $10 or $20 or $120 licenses again, because it worked fine a decade ago, and nothing significantly changed to make it not work now other than all the sexier corporate real estate these companies have invested into across the US and expanding administrations.

You could take a share of that profit from license sales after paying your immediate overhead to set up some sort of trust that can pay out maintainers for that software long into the future. It can even be structured to outlast the development company if done correctly.


I know a lot of small business owners love SaaS that reduce their IT department to someone who can buy a new iPad and log into the apps again.

Payroll, POS, inventory, backup, etc and the owners can keep track of everything remotely. Setup a backup mobile internet solution, and your biggest IT problem becomes regional loss of electricity.

Nothing for employees or others to steal, nothing to spend on tech support labor, and being up and running again is just a matter of replacing the iPad.


It’s also very easy to justify. As long as the product generates more value than it’s monthly price, it doesn’t matter what it costs. And it’s so low risk since you can cancel any time.


So we have settled on one model that VCs understand and manufactured consent to be ruled by that model by relying on VCs

So much for being risk takers and disruptors. Sounds a whole lot like the opposite.

What’s old is new again. Why do we continue to value people who do not do real work but, like a preacher, latch onto a meme that has others work for them?


Because having a constant cash flow fits better the model of a software company since you need to keep a dev team in perpetuity.

You can't hire devs per project like in construction. Developing commercial software requires maintenance, knowing the codebase, etc.


I’m not saying you’re wrong, but you can absolutely hire devs per project. My employer has done it many times. I was in a project that scaled up from 10 scrum teams up to 120 over a year and a half, and then once the product was launched, down to ~25 teams again for the maintenance part of the product cycle. This would have been impossible without contractors.

That said, your first argument is sound.


> I’m not saying you’re wrong, but you can absolutely hire devs per project

You're right of course. I should have written a more nuanced comment that I had time for :)

My point was only that in other types of businesses it's possible and common to hire the whole team per project or season.


The difference is software companies hire programmers while companies that sell physical goods hire engineers. Engineers work hard to get it right the first time while programmers are constantly fixing their mistakes.


Are they fixing mistakes or constantly adapting to new/expanding requirements?

Try adding an extra bedroom or 3 mid-construction on a house that had perfectly designed plans. One might be doable, seem easy even but 3 might require a new foundation or something to carry the load (scale).

What if the ask is to pivot to a commercial property. Now you have to physically move it to piece of land that is zoned appropriately. Or maybe just tear it down and start over (refactor).

Those aren't mistakes, they just weren't planned when construction began yet have insane implications to the builders. Planning and expanding software never ends like it would when you complete a tangible finished good.


> Are they fixing mistakes or constantly adapting to new/expanding requirements?

Fixing their mistakes. Constantly adapting to new/expanding requirements doesn't require SaaS.

But the way it's worked out is that SaaS lowers production costs because code can be shipped fast -- with minimal testing and before there's any real reason to have confidence in it -- on the theory that you can just push out updates later for any bugs that customers find too annoying.


It's a bit of everything. Sure there are mistakes, but OTOH the complexity also grows exponentially.

And it's literally impossible to make software without mistakes. Doesn't matter how much money you throw at it.


True. However, companies used to engage in comprehensive testing before release. SaaS removes the need to do that, saving a lot of money in exchange for a lower-quality product.


Non-SaaS still exists. Why does my OS update every week?


This is a dumb take, product recalls happen all the time if the issue is bad enough, otherwise everything else flies under the radar. All physical goods eventually need upkeep, too.


You probably don't understand how either of them work. Do you remember for how long Apple was trying to fix problems with their butterfly keyboards?


Even if we disregard the recurring-monthly-fee part of SaaS, which is popular because it makes more money, I think the "cloud" part is a main reason there are so many. Nobody wants to deal with creating installable business software, and businesses mostly don't want to deal with installing software across their systems. It's much simpler for a small dev team to have the software live inside a website, and it's relatively simple to bolt on a monthly fee for access to a website.


This is the big one. It's far easier for both sides to use the "cloud hosting" model. The seller only has to worry about deployments on their own systems, and the buyer doesn't have to worry about deployments at all. It really is a win-win.


Time for an anecdote. I have recently witnessed a medium large company with a paranoid CEO, obsessed with business security, owning physical servers only for internal needs, hosting every single needed service, hoarding all the data. And just overnight they started switching these services to SaaS based one by one.

Why? Lack of on-premise and offline based software. Consequently, the innovation today is almost exclusive to SaaS based products. On-premise and offline client software is just becoming an optional expensive side-product of every SaaS product.

But it's a pretty logical step forward. As for the value, SaaS eliminates administration, installation, hardware, technical support and security hastle and majorly cuts the costs for a client company. As for the price, we always praise the always blooming software industry, successful startups, awesome jobs with so many perks and benefits, but someone has to pay for it. I think people don't consider that software is just getting very valuable and too expensive to make and it just makes more sense for companies to rent it then to own it forever.


> businesses mostly don't want to deal with installing software across their systems

And now they have to deal with cloud leaks of sensitive data that they no longer control and various SOC2/ISO "certifications" increasingly look like security theater and a "feel good" measure because they don't guarantee anything.


What is there to deal with? No one cares about leaks, and at most you give out some cheap credit reporting nonsense. A data breach is hardly news anymore.


I won't elaborate on the very real pessimistic comments because those are addressed from others.

What I haven't seen mentioned are scenarios with non-technical customers. In the extreme use case of resource intensive software with completely non-technical customers, who is going to man the ship? It's like selling someone a car who is incapable of driving.

Even if you provide an on-prem system turn-key, who is going to manage and maintain the system? Yes, you can sell perpetual licenses with support agreements, but the delimitation of responsibilities will be difficult for non-technical customers to understand.

SaaS essentially allows customers to outsource their IT operations.


There's a deep and rich ecosystem of IT consultancies that sell, operate and maintain shrink-wrapped software and appliances from the likes of Microsoft, Cisco, VMWare, Citrix. These consultancies sell "SaaS" in the form of, like, we'll have a technician drive to your site, get the keys to the server closet, and spend N hours checking on and maintaining things every month. You can make a living as the SQL Server guy for Duluth or the Cisco phones guy for Milwaukee.

It's hilariously inefficient compared to cloud/SaaS, and no surprise that cloud and SaaS are eating this world's lunch.


Exactly, as much as we love open source in many ways it's also kind-of destroyed the mid-segment of fixed-software sales(the dBase, Lightwaves, Turbo Pascals,etc) and we really need a way to find back to it.

Because not only is people not gaining income from their work (even if for example the Blender foundation exists.. how many people can live off it?) but we're also straining a lot of open source developers with support that really needs to be paid-for in some fashion.

I don't fault Stallman in this instance(not a fan in general), the GPL always allowed re-sales and I think he did see something akin to the shareware world even with free software, but what happened was the combination with the Internets ubiquity made everyone just jump to the source (kind of rightfully due to bad shenanigans by middlemen actors like SourceForge or worse) leaving potential supporting middlemen screwed.


Stallman would probably be happy with the idea of paying an expert who used open tools. It's no different from taking a bus, or hiring a taxi if you don't know how to drive.

My impression is that the key aspect revolves around freedom to participate in society. In theory anyone can run their own email, irc, matrix, etc servers. Anyone is free to use an off the shelf open source product, a paid product, or even write their own if there is need.


Ms exchange and aws, azure come to mind. The cost of paying $10


For your two specific product questions:

> 1) Personal photos and videos backup and viewer. - Just give me a cheap cloud for backup and a desktop app for viewing.

It sounds like you want a one time purchase for backup that lasts forever? I'm not sure who is going to rationally offer that service as there are ongoing costs.

> 2) Personal budget - Just give me a desktop app that connects to my different accounts and gives me overview.

At least in the US that means you have to directly work with banks (ala Quicken), use a third party service (like Plaid) or maintain your own scraping system that perpetually breaks and/or violates the bank TOS. Any of those options is going to cost the software provider ongoing costs. Again nobody is going to take a one time payment and be liable for ongoing costs.

I do think there is money to be made in the right industry for desktop apps that charge a decent amount ($100 and up) and offer 1 year of updates and then perpetual use with the final updated version for that year of use. This incentivizes the developer to maintain security updates and make feature additions so users are willing (but not required) to pay again in the future.


Software is sold as SaaS for the same reason games are multiplayer. Both are driven by the need to stop your app from being pirated. The recurring revenue piece is just a side effect of adopting the only effective anti-piracy mechanism ever developed.

If all the code runs on the user's laptop, it doesn't matter whether it's a game or a productivity app and it doesn't matter whether it's cheap or expensive, it will be cracked and it will be pirated.

If a critical piece of the product runs on a server that you the business founder own, whether that's multiplayer lobbying or MMO sharding or some key part of your productivity app (like data storage or whatever), piracy ceases to be a crippling problem for your business and becomes a non-issue. You also have the opportunity to get a recurring revenue stream rather than a single payment, which is generally preferable for the business (and better aligned with the fact that you do now have monthly per-user costs because you're hosting part of the app on your servers).

As a data point to back up the importance of piracy in the equation, look at the distribution of start dates for successful software companies. There were lots of successful software companies started in the 80's and early 90's, but almost no successful software companies were started between 1995 (when the first web browsers made piracy easy for the first time because you didn't have to physically know someone with a physical copy of the cracked software, you could download the crack from anyone anywhere) and 2006 (when Salesforce.com showed the world what the games industry had already figured out, which is that hosting part of your app on your own server is a great way to stop piracy and get recurring revenues).

Once Salesforce.com showed the industry that "SaaS" works, all of a sudden people could start launching tons of profitable software companies again. Can you find a successful software company started during those dark ages of 1995-2006, of course you can, these are trends not absolutes. Was it far harder to be a software company in those dark days? Absolutely. That's why we have SaaS today and that's why SaaS is here to stay.


Has piracy really ever been a crippling problem for any business?


Just read up on the famous game dev tycoon mechanism, where the gamedev (player) goes broke with his company due to too much piracy.

97% of players played the cracked game that the publisher seeded on purpose via piratebay and other trackers.

They thousands of forum posts after this because pirates weren't able to realize that this was a copy protection mechanism.

So yeah, I'd argue that piracy is a huge problem.


Yes

https://irdeto.com/news/game-piracy-results-in-over-21-milli...


Mhmm, that's an anti-piracy PR blurb from an "anti-piracy" service selling site. Not sure it is the most unbiased source... I think it has been demonstrated elsewhere that 1 pirated download does not equate to 1 lost sale.

Where I live (Mexico) I know so many people that have pirated Adobe, Microsoft apps and plenty of games, who there is NO WAY they would have bought the software if they hadn't been able to pirate it.


1. Easier to manage updates. If you control the service you can ensure it's up to date. If everyone is using their own version you have to deal with bug reports for bugs that have been fixed.

2. As many point out, subscriptions are best for a business because it's recurring revenue and you can grow. One off sales makes this harder.

3. The Jetbrains model is basically subscription model now. They only allowed you to have perpetual license because of backlash. However, I liked their approach of you can have a peretual license for the version that was available 12-months before your subscription ended. It's what I'm going to be using with my product with a slight modification.

4. As you point out, you're likely to forget about a subscription therefore the company makes money while basically providing nothing.

5. Overall, the problem is that everyone is accepting the current setup and paying for subscriptions instead of buying things so more companies are getting in on the gig.


OP here: Those are fair points.

I get it from the business' perspective and also from a software developer's perspective.

Its unfortunate that there is so little option for a consumer here when they are just wanting to just simplify their lives without sharing their personal data and/or subscribing to a service.


You claim that the consumer has it better with the desktop app, but you're ignoring the number of people who want quick access to their photos or finances from their mobile device, or from the computer at their friend's house when their computer dies. SaaS also implies universal access to a system.


Technically speaking, you could totally imagine a SaaS software that does not require you sharing your data. These two things are fairly orthogonal.


How so? At the very least, you need to have some sort of payment method on file with the company.


Only for automatically recurring subscriptions. If subscribers are willing to to manually renew every pay period, the payment methods don't need to to stay on file (outside of the bare minimum required for tax purposes).


One reason is that SaaS pricing is an easy way for people to "hack" corporate spending restrictions. You need accounting approval for that perpetual $10000 software license, but a $500 monthly subscription payment is totally at your discretion, even if you end up spending more of the company's money in the end.


I am not in accounting/finance, but the world has been on SaaS for so long that it's hard to imagine that the department responsible for controlling spending hasn't caught on and plugged this loophole.


Yeah this is less true than it used to be but it's a big part of how SaaS initially came to be the only game in town.


It's best to think of it as three loops.

1. The loop of just regular maintenance. How long is that version of Node supported? 2 years? Alright assuming you did literally nothing, no bugs, nothing else your software is already needing to be tested against a new major version in just 2 years.

2. External loops, integrate with XYZ platform? Oh version 1 is EoL in 2 years? Guess you need to get on upgrading to API version 2!

3. Someone uses your out of date software and runs into an issue, now they email you, tweet at you, complain that your software sucks and you should support it better. (My first job would get this all the time for software we released 6+ years ago) Oh you fixed that bug in the new version? "I should get it for FREE!!"

Who is going to do this work for free?

Even if you cut the support emails down to a minimum, that alone costs money to automate...

Another favorite of mine was complaining about price, "$5,000 for a perpetual license?!"


Even tech people became like the old school folks complaining about charging for fixing a computer or the ol' "$500 for a website!?".

People get spoiled with so many free services and content provider around them and never consider that someone has to actually pay for it. Even tech people live in this ever-blooming software industry with above average paychecks, perks and benefits, but never think about who is really covering for that.

Software development is expensive, value it can provide is high, businesses provide SaaS to enable companies to attract them and pay for the value they use and get. There will always be attempts at price manipulations, but if software is used and business is profitable - the price is reasonable.


As someone who runs a retail (eCommerce) business for a living, I think it's also worth pointing out that people get far more "rational" about their purchases when you ask them to spend a large chunk of money at once.

The same person who would think "do I really need this?" when faced with spending $1200 on a high-end automatic espresso machine wouldn't even think about spending $5 a day on a takeaway latte (which ends up being an order of magnitude per expensive once you amortise the cost of the machine over the number of cups produced).


There is far less risk for ongoing expenses. So many times I have gone and paid the up front cost and then not used the product nearly as much or for as long as I imagined and ended up making a huge loss compared to having subscribed or rented it.


I really dislike the subscription model and avoid it wherever I can. It serves the software publisher at the expense of the user. If I can't license your product, I'll pay by usage of resources or usage of minutes or hours. I won't pay for access. I will pay to use the product.


As someone who built a SaaS for your 2nd use case (Homechart: https://homechart.app, lifetime license available!), the real reason is any kind of software that integrates with other things has an ongoing maintenance cost that needs to be paid by all of the users to keep it working. Like connecting to your bank accounts, that's not a one and done thing typically.


Because businesses with a recurring revenue/subscription model are valued far higher than comparable "buy" businesses.

And of course, the more guaranteed revenue you can demonstrate, the more valuable your business is going to be for potential acquirers. “Because a high percentage of the revenue of a subscription-based business is recurring, its value will be up to eight times that of a comparable business with very little recurring revenue,” claims Warillow.

https://www.techradar.com/news/how-recurring-revenue-can-inc...


My pet theory is that the rise of SaaS in the 2010s is largely a function of the low prevailing interest rates. Standard pricing says that if interest rates are 0.5%, an annual recurring payment of $100 is worth $20,000; but if interest rates are 5%, like in the 90’s, it is only worth $2,000. With the fed increasing rates, maybe we’ll see a return to pay-once software?


Are you saying that because for the SaaS company, $100 in income can service $20,000 in debt?


That’s one way of looking at it. From the investor (and thus valuation) point of view, you need to pay $20,000 lump sum to obtain a recurring interest payment of $100.


1. Recurring revenue - revenue is repetitive, with, ideally, a low churn rate. 2. Revenue forecasting (being able to predict growth with higher precision than ever before, and also check out how fruitful your marketing campaigns are). 3. Flexibility. If users like your product (there's a real value that your product or service provides to them), they'll stick around. If not, they won't. 4. Profitability. The SaaS model tends to be a lot more profitable (especially in the long term).


Personally, I find it a lot easier to address bugs on the producer side of SaaS. Back in the Bad Old Days where I worked on a proprietary database delivered to a customer to run on their servers with general-purpose staff, getting high quality bug reports and working our way to resolution was a major issue, and a very expensive part of the development process.


It's just that having a recurring revenue allows you to offer features that you wouldn't be able to otherwise, and people WANT those features.

For example, as a B2b software company, there is no way I'm offering customers any complex cloud functionality, especially involving data storage, if I can't guarantee that I will be able to cover the costs in the future. Thus recurring revenue is a must.

On the other hand, if you just need an offline app that can work fine without being updated frequently, probably a single upfront payment is better suited.

This is nothing new by the way, multiplayer games have worked this way for a long time. In general a multiplayer game that asks you a subscription fee has a much higher chance of being kept online for longer. See WoW. Without subscription fees they would have probably shut down a long time ago.


Almost all the comments are focused on the "recurring revenue" from the side of the business.

We're looking at subscription from the point of view of the customer in our start-up.

We're in the neurotech/sleeptech business, and competitors (Dreem, Philips) were selling products for $500+, and we see other start-ups in the space apparently doing about the same.

We're looking at a subscription cost because it lets us get into the hands of more customers, which means we can bring our production costs down, and everyone wins.

Yes, our LTV, should we have churn right, will be more than the up-front cost, but as we say "I can't help somebody sleep if I can't get them to try the product".

So please don't ONLY look at SaaS from the perspective of the business. For many businesses, it makes more sense from the consumer side as well.


I have a software product that began with a traditional SaaS model before I realized that usage was project-driven/sporadic and it was better for the customer to do a pay-as-you-go model with no subscriptions.

What happened after I made the switch away from subscriptions?

I received a bunch of support questions asking for the subscription options back. People were truly confused/upset that hosted software wasn't being sold on a monthly or annual subscription basis - even though per-project was more flexible and cheaper for them.

My takeaway is that - for better or worse - the SaaS subscription model is just so expected among buyers these days that to defy those expectations is just going to shoot yourself in the foot.


Jet brain moved to subscription first.

The reason is that you do not want to start over each month sales wise.

Again, as a consumer you can "consume" what is offered in the market.

I.e. if you have an option to not buy, the seller has the option to maximize its profit.


> Jet brain moved to subscription first.

This is why I stopped using their software.


Interesting that you chose the "JetBrains model" as example cause everyone I know who uses one or several JetBrains products uses cracked versions. I know like 10 developers who use their products and I don't think a single one of them paid for it. So I guess one of the advantages of the SaaS models is that you can't actually use the product without paying for it. And if you detect that an account is fraudulent (hacked account, account sharing, fake banking info, etc) or late in its payment, you can just temporarily or permanently disable it with a simple UPDATE query.


"You will own nothing, and be happy."

I don't know about the second part, but they sure are trying to push the first. As many others here have noted, the reason is money, or more precisely, greed and control.


Large companies are growing to levels that are arguably unsustainable. When you have billions of dollars in baseline overhead costs, an amazing product may not even get you in the black. You'd need to do that every year, and likely multiple times per year. It's simply not realistic. So the solution these companies are choosing is rent seeking.

I'd add that this issue is affecting many other industries as well. Entertainment being a visible one. Hollywood has big budgets, big overhead, yet increasingly difficulty finding the big ideas. So we get "Men in Spandex Save the World #83." AAA games are behaving similarly. There seems to be less of an effort than ever to even feign that "Next Big Shooter" is essentially last years next big shooter, reskinned.

This is why I hold the certainly uncommon view that we are probably nearing 'peak big business', at least until the next great frontier emerges. It's difficult to imagine these increasingly widespread "solutions" being sustainable in the longrun. One issue, among many, is a tragedy of the commons with many solutions. Charging rent sounds great for the individual company, but the more companies that charge rent - the more intolerant consumers become of it. See how online streaming went from a golden goose to a bloodbath for everybody once "too many" companies got involved.


The questions feels directed more towards B2C, but let me share a B2B take:

From a business perspective, a SaaS solution can be a ramp to platform lock-in. Once your customers have committed, and your software is part of their business workflows, it's expensive to move away from it.

Another reason is that (for sufficiently complex solutions) the software portion is just part of the service - the business can also offer consulting services for onboarding / migration and customization. That's an entire new revenue stream.


Companies make more money Consumers put up with it And devs like how it's friendly to major rewrites and tinkering.

I use a lot of cloud stuff because there's no other choice. There's no P2P Tile or YoSmart. I'm not interested in hosting something as important as email myself when there are entire teams that focus on that full time. They kind of have the upper hand.

What's the alternative to google docs? Does it involve maintaining a server or saying the words "Ok, just download this app and type in this IP address"?

Open source and (Non-cryptocurrency) P2P is probably always going to be the best for this. Commercial software seems to have clearly decided that it wants to be SaSS.

Maybe there's a niche market for non SaSS, but so many people who want local stuff are concerned with privacy, or have a very low budget, and prefer open source if it's available.

Obsidian seems to be doing well as proprietary local first software. Now that SyncThing exists, it's a lot easier to develop and use this kind of thing.


It is all about money and ease of use. What you are describing is how the software model used to work - it is simply easier to outsource everything. But as you note, of course, this usually makes things more expensive.

Cloud providers and software engineers have of course come up with tons of new tools to make this outsourcing (or you could also call it renting + outsourcing) cheaper (autoscaling being a huge one) and this incentivizes moving to SaaS or PaaS infrastructure. However as these new products mature companies are also realizing they can use those same innovations in their own "on prem" infrastructure. Most of the time it comes down to how big the company is (do they hit the threshold where hiring a bunch of people to run their own datacenter or machines makes sense) and is their business long term focuses enough that they care more about the big potential cost savings vs the advantage of (from an accounting perspective) recuring regular costs per quarter vs larger occasional purchases of hardware.

Personally I think we are going to see a huge "private cloud" movement in a few years once the market matures and companies realize they did not need to give absolutely everything to AWS. I have always liked the hybrid cloud model and then using SaaS only for things that are a huge pain in the ass to manage (since with SaaS you are not just renting/outsourcing the hardware but also all of the software work as well). Email is a great example of this - its a huge PIA to manage yourself and often not that costly to just pay someone else to do it.

In my opinion the holy grail for many companies is having their big "regular" workloads (especially stuff computationally expensive or requiring special hardware or large storages) run themselves on their own hardware + seamless integration with a cloud provider for autoscaling that process up as needed.


Capital in the 21st Century by Piketty claims the trend is labor -> capital in terms of power, I.e. labor becoming worthless, capital becoming more valuable.

Makes sense then that software among other things becomes a service.

Problem is same old conflict: I work for money, they own my work and take money, I resent that, they like it, then violence or legislation or something.


Because people pay for it and companies like money, it's a bit of a dismissive answer but it's just what it is really.


1. everything is mobile: people consume software from phones, tablets, personal laptops, work laptops, desktops, corporate servers. How do you imagine yourself installing your software 5-6 times for a single customer???

2. now imagine updating your software with new release/bug fix. royal pain. these two factors are enough to choose saas


One important reason is the ease of delivery. If you're selling a desktop app you either need to build for different platforms (like JetBrains do), or limit yourself to a single platform (as attested by the many MacOS-only or Windows-only tools). Additionally, you need to support mechanisms for remote updates, and even with that there are many customers who fail to update for various reasons, so you need to support old versions.

With SaaS, you have full control over your product, at the cost of being limited to the in-browser (and mobile) front-end. Your backend can be written in any technology you want, and you can release new updates with any cadence you want, even continuously. And you can even replace the whole backend overnight, with your users being none the wiser (assuming it has full feature parity with the old one).


If people actually looked at what they were paying for MS Office over a decade they'd wonder wtf they were doing.

Small numbers are easier to sell. Before a company spends $20k on a PM tool, they would discuss it to death. But signing up for $299 a month and giving it a shot is easier.


To add to "they would discuss it to death" - yeah now imagine you need to start a company and you have to shell out $300k up front for multiple software licenses.


> Why don't we have more businesses following the JetBrains model specially if the product is not a service that needs to run 24/7.

JetBrains would have gone the SaaS route if they could get away with it. Seven years ago, they made an announcement[1] to go pure-subscription/SaaS, and the outcry[2] was so massive, they reversed course over a matter of days,id memory serves.

1. They proposed licensing term changes included bricking your IDE the moment your subscription lapsed, switching from their previous terms where you'd be stuck on a working then-current release forever.

2. https://news.ycombinator.com/item?id=10170089


It is very hard to develop on premise product.

First you need to works with different Linux OS (ubuntu, redhat, etc.), and nowadays, product may contain a lot of microservice, then you need a kubernetes cluster, which is very hard to maintain. And then, you need to find a good storage (object/blob storage) for kubernetes.

Then, you may need to often troubleshooting server issues (like, disk full, slow disk issue, out of memory, kubernetes network broken). And you need to often check whether the data backup system works.

It is much easier to operate as SaaS, and you can pick your OS and your machine spec, and your programing stacks


Rental fees, so to speak.

Software as a service becomes a rental property and owners can milk consumers for more money over time as well as introduce features to encourage lock-in. Basically, you're teaching frogs to boil themselves.


JetBrains tried to switch to the usual subscription model when they got new management and due to a massive backlash they ended up with a gobbled subscription similar to their superior previous licensing model.


Money and power. It lets companies squeeze more cash out of their customers over time and it lets them make whatever changes they want to something people have already paid for, leaves the door open for things like continuously collecting their user's personal data, and can help make it more difficult for users to migrate away from their product.

It's terrible for users who value their money, security, and privacy, but vastly more profitable for companies so most companies will offer it increasingly leaving users with fewer options for anything else.


Subscriptions are the ultimate rent-seeking business model for software and online services.

When you need to increase profits to justify insane valuations, eventually you need to start wringing your customers dry.


BTW, subscriptions/recurring charges aren't "rent seeking" per se.

https://www.investopedia.com/terms/r/rentseeking.asp


As I read the article you link to, the vast majority of SaaS appears to qualify as rent-seeking. The exceptions would by the things that can only usefully operate "as a service". But that's a minority of the SaaS offerings.


> As I read the article you link to, the vast majority of SaaS appears to qualify as rent-seeking.

The word "rent" can be confusing. Knowing that "rent" in this case means economic rent and not the recurring payments you make to a SaaS provider, which SaaS vendor are you thinking of, and how are they rent-seeking?


By making people use SaaS when there's little to no actual benefit to them for doing so, it makes them do all their work on the company's server. Getting recurring payments for this is one obvious way this increases the company's income.

But it also allows companies to derive income even if they don't directly charge for the service, because it allows for comprehensive collection of the user's data. (Note: I'm not talking about the data the user is processing -- although that's a risk, too -- but data about the user's machine, how they use the software, etc.)

And it locks them in to such a system.


See: Adobe Cloud

I cant think of a more exploitative “SAAS” off hand.


They buy good software and then neglect them until they fall into disrepair or kill them off. Extracting rents the whole way.

Slumlords of software.


Rent seeking behavior is being adapted into all products with the goal of replacing ownership with controlled access. Once people use your service they are likely to be locked in and stay


Imagine a world in which most middle-class families have a cloud VM where they run server-side stuff (Mom's blog, Junior's minecraft server, etc), and when a software vendor sells you software with a client/server model, it's normal for them to deliver the server component as a container to run on your server.

That's the world in which you can buy the software you described without SaaS. As long as the vendor is running the server side, you're paying by the month.


Then there is stuff like Adobe Cloud, where it's forced on you, and you don't really use the cloud features anyway.


I wonder why SaaS has not spilled over to other businesses, ie. food as a service, car as a service, fridge as a service etc. It exists in some cases, but it s not popular


"Car as a service" is really popular, usually in the form of leasing. You can own the car at the end, but many people replace it with a new car after the period ends as there are incentives to do so.

"Food as a service" is called restaurants. But even for groceries, this is by nature a recurring business. They don't need you to sign a contract, they know you will come back if you want to stay alive, they have loyalty cards though.

I never heard of "fridge as a service", or any home appliance for that matter. But it was popular for a times with phones: subsidized phones are essentially that. You pay your phone with your subscription, and after some time (often 2 years), you get a new phone, you got to keep the old one though.

The new thing is recorded entertainment. Streaming services have took over music and video. Video games are a mixed bag, but we are starting to see "all games" subscriptions. Books are stubbornly sold on a "per item, keep forever" basis for some reason though, even on Kindle, which is ironic since "books as a service" have been a thing since the early days of civilization, in the form of libraries.

Thinking of it "buy once, keep forever" is not that common. I just took a look at my bank account and most of my payments are services and/or recurring: utilities, gas, groceries (most of it made up of repeat purchases), subscriptions, taxes,... And usually, when I can buy something forever, I take that option, and I pay cash.


I think you're looking at it backwards, other stuff as a service is the OG, and SaaS is the relatively new thing.

Food as a Service: Restaurants, Cafes, Uber eats, Door dash, and other services that rich people use where they order food to be made and delivered by a private kitchen (not an open restaurant)

Car as a Service: Taxis, Yellow Cab, Uber, Lyft, car rentals, and car leasing

Fridge as a Service: there are stores that would rent you a fridge, it's not as common, as most rental places come with fridges, but I've seen many in Houston without. I've also seen Laundry machine and Dryer rentals

[Anything] as a Service: they exist, but not common or unknown to people like you and me. For example there are dog walking services. Another one that appeared in the past decade is related to clothing where you get a box every month that contains some clothing articles, or others that would assign a personal stylist to help you shop for clothes... Etc


car leasing is a subscription payment, everything else isn't


Not sure I agree. They're services. Uber is a car as a service. You don't pay monthly, but you use it as a service.

I don't pay a subscription for Gmail or their productivity suite, but those are SaaS.


I mean saas as subscription, i don't think it's applicable otherwise. E.g. hotels are not a service as are not airplanes. Neither is uber, you buy itemized rides from individual riders, the fact that uber mediates it doesnt make it a xaas. Otherwise, every grocery store would be a xaas


SaaS business model is nothing new. it just another form of contract.

my employer is in utility management/service. they sign yearly contract with city and MUDs.

i don't see the difference between monthly fee (contract) and yearly/multiple years.


Well, you can rent appliances and cars and there are services that will cook and deliver food for you for a monthly fee.

So yeah. All those things are available on-demand or as recurring service.


It has.

I'm subscribed to a weekly milk and veggie delivery.

Car leases for rentals has been a thing for decades.

Ok I can't think of fridge as a service... But what about extended warranty plans and servicing?


Housing as a service has been around for 1000s of years


Yeah that is very old. i am wondering, if XaaS was so successful for software, why havent we seen other kinds of products turning to subscription-based. Some companies are trying to make car seats-as-a-service but that s just ridiculous


you just described Rent-a-center, and they've been around for awhile.


Isnt bmw offering monthly subscriptions for extra car options?


Food as a service?

hellofresh.com

Car as a service?

zipcar.com


Because enforcement of their copyrights in the digital asset itself, that you would prefer to by, is extremely difficult. SaaS sidesteps the piracy problem entirely.


Delivering the software is much easier:

* Do you want to build it for mac? And windows? And linux desktop (lol)?

And from the clients side, it's also easier:

* You just need to open your browser you're good to go.

* No need to install updates. Just imagine what kind of hassle rolling out updates for desktop software is in a company with more than a few employees. This used to be a dedicated job.

And feature wise, it's also better:

* Compare working with google sheets to excel files being sent around as email attachments.


As others have said - recurring revenue models are easier to structure businesses around. It makes business planning and employing people easier. It also makes VC funding easier because the books are more consistent and easier to understand.

It may or may not be what best serves the customer or represent consumer preferences, but it is (increasingly) how things are done in order for a business to compete (or exist).


Ben Thompson nails it, as usual:

"The brilliance of paying on a subscription basis is that a company can buy exactly what it needs, when it needs it, and no more."

From https://stratechery.com/2019/microsoft-slack-zoom-and-the-sa...

He's written about this in other posts too.


In many cases, it's easier to support. You control the hardware serving the application to your users. You have access to more information for troubleshooting issues your user may have. You can better predict how many paying users you will have and be able to allocate support staff more effectively. You can ensure more of your users are on the latest version of the application.


>I understand the advantages of Saas - transparent updates, cloud based, redundancy, subscriptions, scaling, unicorns etc.

You answered your own question.


There are a few reasons:

1. Higher total revenues. If you compare the subscription fee with the sale fee and how often sales happen.

2. More predictable revenue. There are many reasons people want that.

3. The more information (metadata/telemetry) they have on you the more they can use that for something else. That could be details on how you use the app. If they host your photos they may use those photos to train an AI. Things like that.

4. Finding and tracking unlicensed users is easier to stop and manage. Everything is tied to a SaaS. Less stealing software which cuts into revenue.

There are other reasons, too.

In a capitalistic society where the goal is constant revenue growth, SaaS provides a means to do that more effectively. Look at the goals.


I’d also add: it’s easier to operate and iterate.

If you have a thousand different installs that you need to provide support for, that makes for a very challenging support and maintenance responsibility. People need to be able to operate the software themselves, and may not upgrade when you want them to.

SaaS on the other hand means you, the vendor, are in full control. Sure, it means you need to be able to scale it, but you and only you completely own the infrastructure, which makes things so much simpler.


I just recently cancelled a $5 a month subscription to an Android App/website service for storing food recipes. I was subscribed for about 2 years to "support" the small developer, but really could not justify paying any monthly fee for a program... I replaced it with a self-hosted RecipeSage install.


Besides the recurring revenue, SaaS is also a lock-in device, you can surveil your users to find new ways to make money off 'em, and many people are more apt to pay $10/month because they figure they can cancel it at any time. All of this is more attractive to venture capitalists than the JetBrains model.


They slap anything they can find with a subscription because it is an easy way for them to rent seek and grift as much as they can.

They don't want to admit that they are part of the griftopia that they are creating; especially building a SaSS on-top of something that does not need to be a SaSS or that can be done for free.


You can’t sustain a business on one time sales. Operating systems change, markets change, and products need continuous updates.

Besides, managing servers is “undifferentiated heavy lifting”, why would most companies invest in infrastructure that “doesn’t make the beer taste better”


Yeah but it's not like that software had lifetime free updates either.

You pay for that full version (3.x to 4.0, or 3.2 to 4.2). Want updates? Pay for an upgrade.

This is how a lot of software still works, and so the devs aren't perpetually working for free to existing customers.


In that case you are holding back updates even though they are ready for the “major” update. What if the “update” is just compatibility/features for new browsers/mobile OS versions? Besides, there is no “upgrade pricing” in either mobile OS.


That's a good point, but essentially what Waves Audio does with their plugins (OS compatibility forces you to upgrade).

Mobile is a nightmare, no disagreement there. You have to keep all APIs up to date (new OS release every year?) as unpaid labor for fixed-price apps. You're sort of forced into service model, not goods.


It's simpler than most people think. The Present Value of $10/month for 5 years is ~$450 at around 15% Discount Rate. For consumer software, many are not willing or able to invest that kind of money upfront.


Software piracy. Why make something good when people will just torrent it? Far better to become a SaaS and accept all of the drawbacks and limitations that come with it.


Everything is SaaS because rent-seeking is hugely in fashion. Companies want you to pay them on an ongoing basis.


This has to do with both a technological and economic change. Let's say it's the equivalent of music and movies we go from "owning" through physical devices, then to virtual and now streaming. We are moving to a society where we will no longer own anything and will only use services. This is not only happening in software. We can have car-as-a-service, food-as-a-service, and the list goes on.


In general SaaS vertically scales to increase demand which means factors of increased revue with no new work.


Better business model (definitely for business owner, arguably for customers)


Jetbrains wanted to go full subscription too but their userbase revolted.


Providing software for users to host themselves is quite difficult. We do it at Pachyderm. Our software is basically data version control where you can incrementally process data as it's added (or your code changes). This is much more business-y than the use cases you list, but I can describe why it's tough even when your customers are compute platform teams.

Pachyderm started off as a self-hosted offering, but we did try doing a fully-managed hosted solution, because it would be super easy for users to get started. You write some automation to make a perfect install in your preferred cloud environment, and then users click a button and have a perfectly configured, monitored, and backed up environment. If there's a problem, customer support can go play with the underlying resources (always with explicit user consent, of course; we were super strict about this), and that means that users get a solution instead of a fun debugging problem. Less surface for them to configure, less chance for configuration problems. Ultimately, this turned out to be unappealing to users, for many of the reasons that you list. People want to own their own data and not have to trust even the nicest and most competent people with it. For that reason, we doubled down on the self-hosted option.

There are many pain points with self-hosting. From the start, it's difficult to offer something like a free trial. We'll give you the software for free (hey, it's open source), but you're on your own to find a computer to run it on. You'll need to setup DNS. You'll need to get a TLS certificate. For our software, you'll need to run Kubernetes. Before you can even see what value we offer, you have to do a lot of grunt work. What this means is pretty long sales cycles (while we assist you in navigating all those requests through your organization), and a dropoff among people that just want to have their problem solved today. (You can see this yourself. I bet you'll click any link on HN that sounds interesting. But if I just said "download and run this .exe", you'd probably be quite scared, and it would take some convincing. And rightfully so! That's the difference between SaaS and self-hosted, at the earliest possible phase, "do I even want this thing they make?")

From a technical standpoint, the amount of work you have to do increases dramatically. We have to write code to talk to every variant of object storage. We have to provide configuration documentation and advice for 3+ cloud providers. (For example, we used to depend on etcd pretty heavily. etcd's performance is very heavily dependent on disk iops, and the standard storage classes in your average cloud provider's are woefully inadequate. So we have to get people's data from the inadequate storage to one that is fast enough, and that's just a lot of work.) We have to support customers' internal security initiatives. If you thought keeping up with changes to one cloud provider for one instance of your application was a lot of tedium, multiply that by 3 cloud providers, by ARM and AMD64, by 6 ingress controllers, by 5 Kubernetes versions, by Local SSD, networked SSD, spinning rust, by corporate-mandated HTTP proxies, etc. It is a ton of extra work to make software that fits the customer's environment over just making your own environment that is perfect for your software.

Users dictate the upgrade cycle, not software engineers. I have worked on teams where every commit to master just gets deployed to production within 5 minutes. It's awesome. All you can do when you self-host is to make available another version; if users want it, they will upgrade it. Upgrades are always risky, especially when you wait a long time. So every release has to provide value worth the risk, or the same bug from 2 years ago will be reported again and again. Because of the risk, we have to maintain stable and testing release channels. Customers that want to try a new feature in their dev environment can get a nightly build with the latest features, but we still have to maintain a stable branch for customers that are happy with the featureset and just need security and bug fixes. We have to do work to lower their ops risk. This is honestly twice as much work as just pushing master to prod every so often. Backports cleanly apply most of the time, but when they don't, you're just implementing the same fix against slightly different code. It's work. (I'm a "if the tests pass, there are no known bugs in the code" kind of guy, so I've never been worried about deploying master to production. Obviously, some care in your software engineering is required and this doesn't work for many people, but it does work for me with a very good track record of uptime. None of that matters when you self-host, nobody would simply trust my assurance that the new version is good to upgrade to.)

Even ignoring all of that, it's quite the change in how you as a software engineer operate. You can't reach out and touch the system, ever. You ship a build artifact, suggest that the customer install it, and wait for their feedback. If you need certain information to show health/non-health, you have to make that understandable to the operator (who is often different from the user, who has other health concerns). If you need to be able to debug something, you need to add code to your app to do everything that you would think of doing if you were debugging a live environment, and make the results understandable to the user.

The tooling around this is pretty much the wild west. I recently investigated using Bazel to build our software, because we have three parts that are pretty much completely coupled to each other bidirectionally; Go, Typescript, and Python. There just isn't tooling around all of our requirements (we support both AMD64 and ARM deployments, as customers like those cheap ARM instances and developers like M1s without binfmt_misc hacks), because big tech companies that make these tools have the liberty to dictate their requirements. As a result, we spend a lot of time hacking together tools that meet our customers' requirements, but as a startup, don't have the resources to build something perfect, so developer experience honestly isn't as nice as, say, writing an internal app at Google.

Finally, the hiring situation is complicated. Pretty much nobody is building software that users deploy themselves, so it takes a certain rare type of developer to be productive in this environment. Your code has to show its work; operators that didn't write the code will be running it in production, and your software has to lead them to a solution to their problems. (Even things like "my job ran out of memory" are hard to detect and convey to users and operators.) You can't just push a fix you think will work to prod to see if it fixes things; you have to reproduce the customer issue in your own environment and fully test it before your code goes into a release. It's slow going, and care is of the utmost importance. Code review has to be thorough; you need to understand your customers environment and ensure that a fix for one doesn't break some other. (Yes, we have tests for this sort of thing, but great care is the last line of defense against regressions.) It's not just software engineers; marketing types always want to run A/B tests, and customers don't want to be subjected to this sort of thing. We have to make the right decision in advance, we can't give 10% of users an experiment and see if they rage-unsubscribe over the change.

All in all, it's a very different world. If you feel like your job isn't hard enough just running one instance of your software in an environment you control, selling your software for self-hosting is the job for you. It's hard. It's a slog. But, it's great for your users.


Because gross margin is obscenely high and investors like that.


Everything is a SaaS product because that is what users want. That is the simplest answer. Innovation always comes back to the end user. If SaaS was not a business model innovational that the user wanted it would simply die out. Instead, it’s growing.


Because the MBA's decided they could make more money that way.


??? JetBrains definitely offers monthly plans, SaaS-style.


People have already mentioned all the technical aspects, or blamed capitalism... but I really think it's just the best fit for the benefits that software delivers. It provides lower barriers to try for consumers and higher long term rewards for the company that builds the software. You can certainly find companies that have a flat rate or fixed price, but the incentive to build those projects from the beginning is usually outcompeted by the incentive to build a subscription product.


> I really think it's just the best fit for the benefits that software delivers.

How so? The only "benefit" you cite for the user is lower cost of entry, but the total cost of using the software over time is higher, so that's not a clear benefit.

From a user's point of view, SaaS does have some other benefits, certainly, but it also comes with some rather large drawbacks. So it seems to me that, on the whole, it all just adds up to software costing more.


Capitalism, baby! Companies have figured out they can take more of your money if they extract it little by little instead of in chunks.


Rent seeking behavour


Because money


gotta pump those ARR numbers baby


We have modelled our company/product specifically on JetBrains. We're also bootstrapped, self-funded, etc.

We build a tool for Apache Kafka (https://kpow.io), it's not a SaaS product, it's a single docker container or JAR that runs in air-gapped environments, our users install it in their own network and it just runs. That's what we thought engineers wanted, and tbh it turns out we were right in plenty of cases.

Our licensing is an annual subscription however - but then again so is JetBrains (well at least for the Intellij product that I happily pay ~$150 a year for).

I think for us the recurring revenue is really important, we wouldn't be able to sell you a perpetual license as we have commercial costs that are ongoing and related to maintaining the quality of the product. Not only is it better for us commercially, but also we have only had one customer in the last three years request a perpetual license, and we we explained we don't offer that model they bought a subscription.

So I completely agree on the non-SaaS JetBrains model, and I love that you mentioned that company. We often get asked about Confluent in our area of expertise but quite honestly from day 1 in 2018 we've been aiming to be the JetBrains of distributed systems. We even followed their dark branding style.

Edit: Just to add one more bit of context now I think of it. Don't underestimate the power of the hive-mind.

We have spent four years building a boostrapped non-SaaS product. We spend all our time talking to customers, shipping features, squashing bugs - living the dream basically. That's a really rare path to follow this decade.

We've also had four years of often well-meaning people trying to intro us to low-information 'startup investors' or god-forbid another startup 'accelerator'. We stopped talking to all of them about two years ago, but for a while there we got told repeatedly to build a SaaS product.

And the single worst piece of advice I've ever received which nearly made me puke, when we were in year one of our product and already had a reasonable number of users / clusters:

"Just grab all their information and stick it in an S3 bucket, that information is what everyone wants! Number of clusters, users, version, etc - you can sell that!"

Some people just don't understand that you can sell a tool that does something valuable without making your customer base some side product that you sell on the open market.

It has been hard at times when we see startups raise tens of millions, but we're now in a position where we are miles out in front of the pack, have a rock-solid product, a great roadmap yet to go, and s stack of great customers who we respect. At no point would someone else's cash or terrible advice have left us in a better position - though we might have ended up with a SaaS product instead..


Money


Margins.


It’s a model that makes money printer go brrrrr

Most customers like it, it keeps the company able to push newer features and maybe up sell them.




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: