2M users but no money in the bank | Hacker News Reader

2M users but no money in the bank

exercism.org

699 pointsleandot351 comments

This is one of the saddest posts I've read in a long time.

I love exercism. It's beautiful, works in the browser and from CLI. The lessons are really well designed. It's the perfect coding school environment.

So, I'm sad it isn't viable.

But, I think I am really saddened by the comments here, second guessing all the decisions they made. If I were to second guess the people behind most of the comments, I would assume no one has ever run a business with payroll, or built something with this much reach. The comments ring really hollow for the most part and seem really callous given this person has sunk his life into making the world, and my life, a better place. I'm scared for the moment he is in life and sad there are not more positive comments, and even better, comments from people where they would actually help by making a donation.

Thank you for being so kind and supportive. I really appreciate your comment.

I'm ok. I'm hopeful for the future. I believe there's lots of amazing stuff we can do and that Exercism has a great future. But right now it's in a hard, low place. The last few months have been pretty tough emotionally, but I think getting to this place is one of acceptance. The only direction now is up :)

I've gotten so much from your site. I'm sorry you have put so much in of yourself and haven't been able to make it work financially.

Would you email me (chris@extrastatic.com). I've got a few people I think would help depending on what direction you want to go in.

Done, although I emailed the email in your profile. I'll send it to that one too :)

Can you please just add a paid tier before shutting down? Just a $5/mo something that shows a badge on your account, or whatever. Call it something other than "donate", "premium account" or something, and see how many people pick it up. You "only" need 1500 people to do that to cover the costs, which is less than a thousandth of your user base.

Even if it gives a few useful features to subscribers, that's better than shutting down.

They have a paid tier, called insiders, although you can also "pay" through helping other users on the site: https://exercism.org/insiders

Ah, thanks, I looked for that but didn't find it (the "donate" button reminded me of a one-off). I'd recommend changing the link name to "pricing" and A/B testing the recommended donation amount being $5/mo.

I don't know how much effort they've put into experimenting with pricing, but it'd be worth trying to increase donations, especially for such a well-designed site.

Never knew they had this, signed up immediately :) Exercism is worthy of 120$ a year and worth its weight in gold. It has helped me learn Rust, Swift and Kotlin and I'll be forever greatful!

We're not shutting down :)

But yes, rather than shutting down, I'd probably just put the whole thing behind a paywall to try and keep it alive.

Well yes, I mean "maintenance mode" or whatever it is. You've built a really valuable service, if you want some help with maybe reducing costs, I'd be happy to have a chat at some point, I have experience in running things on budget hardware.

Please do not give up. You are doing something that I think is very important in the educational space.

I’ve had things with more users and 1/1000 would have been amazing. The difference between $0 and $1 is gigantic though, and a hard ask for content in most cases.

Yeah, I guess that's true, but it's something that can be worked on and optimized, and I have a feeling the OP hasn't focused much on actually making enough money to pay employees.

I had no idea you were in financial straits and was just using Exercism yesterday to teach myself a new language. I’ve really appreciated the platform and it’s already helped me learn so much.

I’m donating today and I wish you the best.

Thank you so much for commenting. Always awesome to hear about from the people exercism has helped.

I agree with the parent, I love exercism and although not a daily user I am surprised by this announcement.

I hope you would consider a "hard sell" banner ad for donations in the vein of Wikipedia before giving up on the model. I wonder if there are a lot of potential donors out there who would respond if they understood the stakes.

Yes. Like others here, I did not know there was a paid Insiders tier. Just nudging free users with a “consider becoming an insider” hard sell splash screen upon each login with a 10 second countdown would go a long way towards encouraging users to join Insiders and might be enough to avoid going the ad route. Then you are still offering a free option but I think lots will sign up.

Redesign the donate/payment page for clarity. Enable Apple Pay to reduce friction.

They definitely should consult with monetization gurus. Building useful product does not lead to paid users

I submitted the article to HN after reading it and I have joined the Exercism Insiders - recurring $10/month, a very fair deal. Hope that has helped a bit!

As someone already mentioned please consider adding ways for monetizing, ads + remove ads, paywall, paid exercises, etc. to stay afloat.

Also please get in touch, email in bio, I don‘t have anything concrete in mind but would like to be in touch. In another life I would gladly be teaching coding, who knows maybe it works out.

Thanks so much for your support :)

Can you detail the $7,500 a month server bill please? What hardware are you paying for and in what quantities? Would love to compare to what we're using at Linode to see if maybe there is a better pricing structure out there for you.

You've built something great that 2M enjoy, remember that!

Fwiw moving away from AWS would probably get that bill from 7500 to 750pcm or lower. Cloud is very expensive if you've not made special arrangements or can burn VC money/free credits for your org.

> I would assume no one has ever run a business with payroll

Or maybe they have and therefore know you can’t sustain a “business” without collecting money to pay for your expenses?

I think we passed the 2020s and in 2024, nobody asks "how will we sustain this" nd just hopes that a VC will give them x$ which ... will them just buy... time.

Yes, I have deep respect for the people building such a site for the better good, but asking the "how will we sustain this" from Day 1 is important (not even how to get rich, just how to get break-even).

Indeed. Who knew free wasn't sustainable?!?

> I love exercism. It's beautiful, works in the browser and from CLI. The lessons are really well designed. It's the perfect coding school environment.

By any chance do you also donate?

When your Internet-site-based business can no longer afford to pay its bills, start charing and lock out everyone who does not pay.

Maybe you can reopen a free tier later, but for now you are in panic mode.

"7,500 users, but a bit of money in the bank because the server bills are somewhat lower"

>I would assume no one has ever run a business with payroll, or built something with this much reach.

Somewhere along the line between 2010 - 2012 and definitely post 2013, most of the HN comments are no longer founder focused or business minded. It is unfortunate.

I don't know if iHiD will see this, but this is Lane, the founder of Boot.dev. One of our students shared this article with me, and it really sucks to see. Exercism has always been a go to recommendation of mine for students who need additional practice or to learn languages we don't yet teach.

I know you mentioned you don't wanna do ads, which makes sense. But if you're getting 1200 sign ups a day thats worth a lot, and id be happy to chat and see if there's a sponsor opportunity thats uninvasive and that could at least keep you going.

anyhow, you can reach me on Twitter @wagslane or via email: lane @ boot dev.

either way, hope things turn around

I agree with the ads model as a way forward where purchasing a membership removes them. The fairest way.

second this, pay to get rid of ads, a fair deal and proved model.

Thirding this.

Another thing to try is to find mentors to train juniors and charge small %

Congratulations on a very successful platform! (Success should be measured in usage/reach, not $).

You've probably thought of this, but it might be easier to charge a very small amount -- say $1/month, which most people even students, could afford, rather than rely on donations. If you get 1% of users paying that, it's still $20K month which is more than you're getting now. (Of course managing the payments themselves costs money).

It can still be non-profit (which is a great thing), so no one is getting rich off of it, and it doesn't poison the mission. But it could pay the bills.

Update: Processing fees are a problem as posters have pointed out, so you'd probably want to make it $10/year, or $5/six months. It's still an easily decision for a very large number of people.

Patreon has $1/month options, not sure how they do it though.

I think this makes a lot of sense. Same model as TryHackMe.

Would $1/month even be worth it? It seems payment processors would take a big chunk out of that $1, which is why I very rarely see monthly subs of that amount.

Yeah, if you want that little, do Bitwardens $10 a year. $1 transactions are ripe for abuse, checking if cards are active.

You'd probably want to charge $10/year (once) rather than $1/month. It's still an easy choice for many people (of course still inaccessible for students in some countries).

Another option is to make deals with universities or schools for access for their students where they pay a certain amount per year and the students can all have access

Uni sales cycles are horrendous. take forever and fickle

Yeah, that would be a __lot__ of overhead. Not worth it. I'd instead focus on trying to find sponsors ("$100 / month allows us to give a free account to 100 students in X countries!")

I've always wanted to experiment with a $5 a month or $10 a year model for low-cost subscriptions.

Some payment processors have "micropayment" fee structures, where they take a higher % (5% in PayPal's case) for a lower fixed fee per transaction.

PayPal's traditional fee (for USD)[1] = 3.49% + $.3, for $1 = $0.3349

PayPal's micropayment fee (for USD)[2] = 4.99% + $.09, for $1 = $0.1349

Using that model would make sense in OP's case.

[1]: https://www.paypal.com/us/webapps/mpp/merchant-fees#statemen...

[2]: https://www.paypal.com/us/webapps/mpp/merchant-fees#micropay...

I did not know if Paypal had micropayments fee, thanks fir sharing!

Well, it's $12 a year. I always pay a sub for a year or two if it's that low.

You could use crypto for micro-transactions (it's one of the main problems that crypto was meant to solve). LiteCoin, for example, apparently has quite low fees[0] ($0.03-0.04) Monero much less than that. But enabling this in a frictionless manner for users is probably difficult (I don't own crypto myself).

[0] https://coinwire.com/cheapest-crypto-to-transfer/

Hasn't LiteCoin mining largely subsided and therefore is subject to 51% attacks?

Don't most payment platforms have a constant minimum fee per transaction? That would hurt you at a $1 price.

Yeah their minimum is $10, which for me, I can only justify if I daily your project, or use it frequently. I love the site, but I don't use it nearly enough to justify this cost. $5 I might forget about. There's a site I rarely use that I pay $3.99 for, because its <= $5.

I'm guessing fees make it not worthwhile, maybe they should negotiate lowered fees from their payment service, or figure out someone else who would give them drastically lower fee rates.

There will still be plenty of students (mainly from outside of the West) who will immediately lose access if anything gets paywalled by even a cent because of various reasons.

Yes, for sure. You could make it free for users in X countries, based on IP. Yes, people elsewhere could use a VPN to get around it and use it for free but you're charging so little that anyone who can afford or has a mechanism to pay won't go through the trouble.

> Congratulations on a very successful platform! (Success should be measured in usage/reach, not $).

Most organizations aim to survive as long as possible, so financial sustainability ($) is crucial.

> Success should be measured in usage/reach, not $

What? No.

If a business isn't able to sustain itself with its revenue, it has failed and will need to shut down. That's the definition of failure.

Lots of things can be "successful" by your metric by using unsustainable pricing.

This entirely. You can't eat using "good feelings" / "impact" alone.

Also the whole non-profit angle is immature in my opinion.

Non-profits denote nothing about their financial good-standing or moral good-standing. You can be an amoral non-profit, a moral for-profit or vice-versa. Your profit structure has nothing to do with the morals of the people inside of it.

Additionally, there a lot of non-profit executives that are profiting very healthily off of their work.

Non-profit simply tells people what the incentives/motivations are or are not.

The average for-profit company can't be trusted/expected to do what's best for anyone outside the profit structure (e.g. users or the public at large). Of course you can have exceptions in either direction, but on average for-profit orgs will likely be willing to fuck over anyone else to increase profits - intentionally or not - because profit growth is literally their purpose.

'Non-profit' advertises that you don't have to worry about that specific, powerful motivation of the org contending with your best interests.

Unfortunately, in practice, this isn't really true.

It's a less powerful motivation, in that there are not shareholders who will benefit from growth.

But in practice, executives and managers end up making more, which provides an incentive for growth.

Completely agree. In the end, it's all about the motivation and ethics of the owners. For an extereme example: OpenAI is "non-profit".

Not everything beneficial to the world has a viable business model.

In the best case, you get to change the world and make a bunch of money. Sometimes, you just have a viable business selling sugar water. And other times, you're doing something really good that's just not going to get you paid.

There's nothing wrong with the last thing, other than its potential lack of sustainability.

I'm now a teacher; I could easily make 10x my comp elsewhere; possibly 100x. Measured on your scale, the only value is what I'm getting paid. But it turns out that those good feelings are worth something to me.

I don't think anyone is arguing that the good feelings aren't worth anything but rather that we are all humans with mouths to feed at the end of the day. If your comp went down 90% or even started costing you substantial money, at some point you too would probably reconsider being a teacher.

> And other times, you're doing something really good that's just not going to get you paid.

This. I feel that this has been missed in the thread a little.

Yeah, you can’t deposit vibes at the bank, and employees can’t pay their mortgages with them either.

A sustainable business demands some kind of ongoing income.

From TFA, they’re not a business, they’re non-profit.

Nonprofits are businesses.

They aren't. Look it up in the dictionary.

https://www.dictionary.com/browse/business

Someone's profession (their "business") can certainly be working at a nonprofit, but that's because they are getting paid. The nonprofit organization itself is not a business. You might be looking for the word "corporation", which can include nonprofits.

I don't need to look it up because I've started two nonprofits and more than a dozen corporations.

Nonprofits in the US literally are corporations, just with a different tax status. They need revenue that exceeds their costs. Many of them sell things. They can even turn a profit. Ikea operates as a nonprofit in the US.

I know nonprofits are corporations. That’s what I already said above. You claimed they were businesses, which is incorrect and which I disputed.

Requiring more revenue than costs does not a business make. You starting a nonprofit does not give you special insight into the meaning of words. I am the treasurer on the board of a nonprofit. I have had to go through the books quarterly, confirm we’ve got enough cash for a year of runway, etc. The understanding of what nonprofits do is not the issue. This is a boring dispute over words (semantics).

Remember, you initiated this dispute over words when you made an argument “by definition” and you criticized KolmogorovCamp’s correct use of the distinction between nonprofits and businesses.

It may not be successful long-term with this model, but it's been very successful so far.

You do know Twitter never made a profit, right? And Lyft is 14 years old and is not profitable...And GitLab does not make money...And Zillow is not profitable...And Snap is not profitable.

And that does make it the correct approach somehow? How will this business continue to exist in 1, 2, 5 years?

Someone has to pay (paid) for this business a.k.a. is actively loosing money.

How does it not "poison the mission" if it means 99% of users drop the platform?

Does that even matter if the mission is dead?

It matters for integrity. They promised 100% free forever. But anyway, death and betrayal aren't necessarily the only options. As others have suggested, they can try soliciting donations more forcefully.

If that doesn't work, they could introduce premium features. Stuff that obviously has to cost money would make the pill easier to swallow, for example, features that depends on third party services (only LLMs come to mind right now), or human assistance (real-life instructors). Admittedly, this would be a compromise. One could argue it fails on the 100% part, but another could argue that 100% of existing features remain free.

that's one percent more success at the mission than you have if you're forced to turn the servers off because you can't pay the bills. Money's to an organization what's oxygen is to a body, before you go and plan any missions you need to be able to sustain yourself.

Fair point. I used 1% at the very low end -- if the amount is small enough (in the $1/month range) I think the opt-in rate would be much higher.

I would then give a free account to anyone who asks. ("We know that even $1/month is a big ask for some folks in certain countries or they may even have a method by which to pay it. If that's you, let us know and apply for a free account here.") Then basically auto-approve (with some bot-screening mechanism) accounts that apply. Sure you'll get some abuse and people lying, but not that many people are going to lie just to save $1/month.

Another alternative is charging $10/month, but allowing students or anyone to request a waiving of fees. They just need to take a minimal speed-bump action to request it, e.g., sending an email once per year. Anyone who isn't willing to do this probably isn't getting much value anyway, and a non-trivial fraction of people (more than 1%) would plausibly pay $10/year just to avoid the speed bump or to support the mission.

>(Success should be measured in usage/reach, not $)

measuring success by how much people are willing to shift scarce resources in your direction is a perfectly good way to measure success. It's "weighted usage+reach". the remarkable effectiveness of markets in allocating resources this way is a huge achievement for civilization. I really had to scratch my head at their surprise that a non-profit oxymoron model, i mean, business model wasn't viable.

I was doing some Exercism tasks the other day and then saw this so if only felt right that I make a donation. Looking at some of the comments there seems to be a number of people who have also just made a donation.

Given the state of every service becoming subscription based, I can definitely see a lot of subscription fatigue. Maybe an alternative route you could try is the annual wikipedia "We want to keep this free. Please help us buy donating". I think a lot of people appreciate the service but a lot of the times it's easy to forget that running a service like exercism can get expensive, and would be willing to donate, given a compelling enough reminder of this.

Personally, I think this approach of asking for donations is the right move for a non-profit. Subscription fatigue is real and is becoming a major problem. I am now at the point where I will go without software I would absolutely pay for one time, but because it is a subscription I refuse.

I'm sure when all is said and done, there just are not enough people like me to make a difference, but I sure hope things swing the other way. Subscriptions for everything and tips for everything are two cultural things that need to die in a fire. They are ways of squeezing more out of people, especially by taking advantage of human tendencies. It's gross and at this point I think it is unethical. Maybe not all implementations, but there are plenty that are.

You could use the time value of money formula to convert between one-time prices and subscriptions. just multiply annual price by 30 to get forever price.

So my $6 per month VPS would be about $2,000 upfront.

And a $500 game console would be a $1.50 monthly subscription. (Obviously glossing over tons of differences in say, a PS2 and a Stadia subscription)

It feels easier to ask customers for $1 per month than $360 up front... Especially since I don't know if my service will be up forever

I don’t think a perpetuity is correct in this case. Software depreciates rapidly without updates and the norm was to charge the one-time price for each major version, which was always somewhat analogous to a subscription assuming you wanted up-to-date software.

What is the math on multiplying times 30?

If pricing were consistent, I would fully agree with you. However, it is not. People will charge what they think the market can bear. When you're asking for it all up front, you can't charge a huge amount because nobody would pay it. When you mask the request by dividing it up into small chunks, people don't realize how much they are actually paying. In many cases now with automatic billing, people may not even realize they're still paying. The whole subscription thing is an exploitation of human psychology.

If pricing is too expensive the next month, then just cancel. It’s that simple. Monthly subscriptions simply give more freedom both for the user and the builder.

It's hardly annual price times 30. You must account for inflation.

This kind of thinking is why sites like this one, get into financial strain. Why do you expect to get value for free? Would you be the same kind of user who complains about ads, or a highly priced one time fee? I’m glad more products are subscription based since it filters out users who are not serious.

Voluntary donations don't work. At least not for most things. The vast majority of users don't donate, and that includes the most demanding ones.

I offer a ton of free information online. I received raving feedback from people, including some that said I saved them thousands of euros. Just this week readers jokingly suggested giving me a Nobel prize or a statue. Still, donations are one percent of my revenue on a good month. They would not cover my groceries.

This is fine for me, but it's always on the back of my mind when people suggest donations as a funding method.

I'm assuming that you are talking about "All About Berlin"? (I am deliberately not linking, because you didn't, and there is probably a reason for that).

I run a number of apps and services that don't make a dime. In fact, they cost me thousands. I don't mind.

Monetization is a real pain in the ass, and I stay away from it (for these services).

I have the ability to make a lot of money, if I choose, but that's not much fun. I don't really need the money, and I like having full control of everything I do.

I am regularly approached by folks with monetization ideas. Some, are quite good. Others ... not so much ...

Maybe I 'll pursue something, eventually, but I'm in no hurry.

Thanks for your service. Looks like a great site.

I agree that monetisation is a pain. It's the least favourite part of my job because it takes focus away from providing value to my readers. I don't like putting my business hat on.

On the other hand, it allows me to work on what I love full time. This would not be possible if I relied on user donations or even public funding.

> I have the ability to make a lot of money, if I choose, but that's not much fun. I don't really need the money, and I like having full control of everything I do.

Thought of handing off some of those money-making opportunities to others, after you vet them beforehand?

I could use an alternative income stream and I am very tired of the employment grind. But I never networked and that has been biting me on the rear for the last 10-ish years.

Apologies if that's too direct. I am keenly feeling the lack of freedom and the lack of "frak you" money in the bank.

> Voluntary donations don't work. At least not for most things

In my (limited) experience they mostly don't work because the people/projects that need donations are way too coy about asking for money. Donation buttons get tucked out of sight and/or cloaked in twee language like "buy me a coffee!"

I like to point to https://archiveofourown.org as a counterexample. Hardly the most critical project in the world - it's just fanfiction - and yet they raise a few hundred thousand dollars a year in public donations (solidly surpassing their budget goals) because they run an actual, clear, focused fundraising campaign twice a year.

You are right. I'm rather coy about it too.

Personally I don't want to guilt trip my readers into donating. There are affiliate links on the website. It's a business and it works well enough to support me. It feels like double dipping.

Above all, There's a point where it operates exactly like an ad and annoys people just like one. At the moment, American tipping culture is taking over in Germany and it feels like another instance of that.

The nice middle ground would be to ask after doing something especially nice. "That tool sure saved you a lot of time. Donate?" I think that the link in my email signature works especially well because of that.

In my experience asking for donations generates about 1000x less money (within that exact order of magnitude) than directly asking for money up front until you get access to the product.

Yes, but they're not trying to make 1000x more, they just want enough to maintain their existing staff and servers. They're not looking for unreigned financial growth, it's a nonprofit who wants their service to be free and accessible for everyone.

Subscriptions are a pain for sure, but I think a big reason for that is that they are always something like $10-20 a month. For something that people might only use occasionally that's just too much. If they were instead maybe $20 a year and easy to cancel, I think a lot of people would sign up.

Yeah that's a fair point. A modest annual fee is probably the way to go, especially since exercism is a service you dip in and out of (at least I do).

That’s around $1.5 dollars a month. No one can build something full time with that kind of pricing model. That’s more of a side income pricing model.

Donation fatigue even is more real than subscription fatigue.

The only way I can see this working is a recurring donation with a low enough amount that people will just set it once and forget it.

So instead of a $1/month subscription, you could try a $1/month recurring donation.

Donation fatigue? The only time I'm prompted for donations online is Wikipedia and internet archive in December. That's about it. I can't think of any other service that asks for donations.

Very much this. This is the gold standard for non profits. Wikipedia does it, the Internet Archive does it. And the banners are very large and intrusive during donation month. This suggests that they must be having an effect, or they wouldn't be making them quite so large! I think if I was in their shoes, this is probably what I would try. December is generally giving month.

I just had an idea that is a shot in the dark but you could try cold calling large companies with complicated SDKs to see if they will pay you a large monthly fee in exchange for exercism hosting a tutorial they create for their SDK.

You already have the sandbox and the support for the most popular languages.

It may take some development to build out the feature but you probably wouldn't need too many to sign on to cover your costs and you could still leave your core offering free.

This is a great idea, especially when you consider the untold millions that companies like Twilio and Stripe already spend on educating developers. Perhaps along with "language tracks" there could be "API tracks".

Brilliant, genius idea, defo worth trying.

I'm just thinking out loud here, but for 2 million users learning to code and improving their skills, there is a chance they want to change jobs or are already looking for one.

Have you thought of adding a hiring feature? Where people upload their CVs, and companies search (anonymously) for talent?

There is a hiring model called PPA (Pay Per Application), or monthly fees to search CVs.

Your unique selling point would be companies are not getting hundreds of custom AI CVs flooding their application; they search for what they want, and you're helping both sides.

Good luck!

edit: grammar

even easier, just a job board (like stackoverflow), where companies pay to advertise a job.

the job board could even pitch jobs in the language an exercism user is currently learning

This is a tarpit monetisation method unfortunately. (I learnt the hard way) It's common for student ideas too (as in ideas with students/recent graduates as endusers/customers).

The general argument is - we can specifically identify talent and motivated learners/skill-possessing people, who will look for jobs. That's exactly what companies need and companies pay $5k-$10k a head during good markets for this rare STEM skill, like programming!

Problem - The people on these platforms are 70% likely to give up within 6 months or go into another field

Problem - Experience means 10x more than education to employers, especially fundamental experience

Problem - Your information on potential hires as the provider becomes outdated fast and supply-demand changes in hiring market can reduce your commissions by 50-75% in one year (as happened from 2022-2023)

I would go as far as saying these users - because they are typically earlier career - are actually worse for connecting with jobs boards/hiring than literally any random iPhone user for advertising purposes. And companies don't pay much to advertise to random iPhone users. I wish it wasn't so, but it is so according to how companies act.

If OP wants to monetise this way they should consider an affiliate programme to a CV upload or Jobs board site, it will test any appetite for hiring these people at the moment much quicker.

It's also worth pointing out the Exercism isn't for coding newbies - it's for developers learning new languages. This often gets missed.

I think this is a major selling point. If developers could upload their CVs, exercise could package together the users yoe + industry experience with the users learning veracity + main language. I think that would be a solid product that hiring people would pay money for.

Shouldn't that make it even easier to monetize? I appreciate the resolve to keep things free for those who can't afford to pay, but a lot of your users are presumably actively employed making 6 figures.

You could follow the Sublime Text model and basically have an honor system—if you're actively employed making more than $X, please pay us. If you're not, you're welcome to use it for free!

The Tech Resume Inside Out [0] uses this model (though not on the honor system) for similar reasons.

[0] https://thetechresume.com/

Yes. Good idea.

Another thing. There are several podcasts who have a free feed and paid one. Sam Harris’ Making Sense comes to mind with their free feed and a paid one for full access to all content. But they understand not everyone can pay. If one is unable to pay, they can simply email to ask for a free annual membership.

Pay what you want albums on Bandcamp are another example of a sliding scale. Suggested price $10, but artists can define their minimum.

That's a particularly difficult vertical to enter. Most large companies (and a lot of small ones) would consider them an "agency" and wouldn't consider any of their candidates. I know that Recurse Center, which is fucking amazing, ran into problems like this when trying to place their participants.

Maybe, but as a company a place that appears to be designed to teach beginners how to code isn't where I'd want to look for candidates.

I would look to reduce server costs instead. $7k/month for 2 million users seems like a lot. Do they execute your code on their servers? I couldn't really tell from their website.

My marketing department this week asked for my approval to spend $50k placing our logo in an IT newsletter with 300k subscribers.

If you have an email list of 2 million… a lot of companies will pay $50-100k for a mention.

What does "I've lost faith in the nonprofit business model" mean? Non-profits have costs they pay. Many famously outlive their founders. Many, like co-ops or credit unions, don't depend on donations/endowments. They have revenue greater than expenses as a matter of basic accounting?

Non-profit is OK. Non-revenue is a bit of a non-starter.

Yeah, that was a big weird red flag for me as well. He is blaming the failure of his dream on "the non-profit model"? Wait till he tries the for-profit model...

Then he goes on to say that he is upset that 96% of people stop the program. Dude, coding is BORING, but coding is hyped so much that people have high expectations so you get delusional people coming and signing up to your website.

It sounds like his model was: "Make a non-profit org to do something useful for society, get funding from orgs that have money and want useful things done"; and it sounds like the problem he had was that orgs that want useful things done are often too specific.

E.g., there are orgs to promote computer literacy among women; and among poor people in India; and among middle-schoolers in Africa, and so on. And he has users who are women, and who ware poor people in India, and who are middle-schoolers in Africa, and so on. But none of the orgs will give him a grant, because they can't guarantee the money will only be spent for their own target groups.

I can see why that would be kind of discouraging.

You can still work out a deal where you could identify some portion of your users as fitting the mission.

I read the comment more narrowly. I think he's lost faith in Exercism's specific non-profit model, not non-profit models in general.

He wrote:

“I think it's fair to say that at this stage I've lost faith in the nonprofit business model working in a way that allows Exercism to reach any of its potential.”

Which provides additional context. Why’d you remove it? It makes it clear that he’s talking about the non-profit business model as applied to his company.

The only way out for something like this to stay afloat is to be subsidized by a big tech company.

$7,500 monthly is nothing to a large tech company. It sort of does makes sense for big companies to subsidize this so that the quality and amount of software engineers that join their company increases.

This value prop is intangible, but I feel like at least one big company would go for it.

Well... they could charge for it.

The $7500 jumped out at me, but for a different reason. This seems awfully high for a site with 2 million registered users. I feel like some backend optimization could get than number way down ($1000/mo?).

What is all this computing power going towards?

I imagine they need to run code for their exercises on a server. There are probably lots of ways to optimize that, and perhaps even ways to get the code running in the browser via webassembly or something, but all of these things require technical expertise and time.

That's interesting. Well, they are coding gurus, so maybe it's feasible :)

It is going straight to Jeff Bezos pocket selling them servers at at 10 to 20 times markup then reasonable

> 10 to 20 times markup

I don't want to argue that AWS is cheap, but I call hyperbole here. Where can I find an m5.large (2 core 3.1GHz CPU, 8GiB RAM, 10Gbps network, $70/month) for $7 USD a month?

The plain servers aren't that bad, but then they get some extra billing in. Traffic is pretty nuts, if you sustain sending on that 10 Gb/s link it will cost you approx $160000 per month.

Then there are all the services that ain't plain servers, those are generally priced at at least a 10x markup, some of them much more than that.

agreed. this can be optimized by the community. Sometimes, and I'm not saying this happened I am just saying that sometimes engineers have a strong desire to over-engineer and that might be costing money. Also want to add that maybe you can get approved for AWS for startups for example and migrate some compute there.

They are already on AWS, and that is the problem. Probably wouldn't cost 1/10 with normal server leasing, but they have to code their way out of any AWS specific services they rely on before they can switch provider.

How are they managing to spend 7.5k a month in server expenses? Are they using AWS?

> Are they using AWS?

Yes!

You’d think they’d use a cheaper hosting service if that’s literally the only thing preventing them from having positive gross profit.

honest questions, is there any ECS-like hosting that's much cheaper?

> is there any ECS-like hosting that's much cheaper?

Most have adopted EKS-like services i.e. kubernetes.

There is fly.io that's closer. Hope they improve on the reliability aspect.

for compute of containerized payloads, in house servers is a no brainer for cost.

almos zero sysadmin troubles. might even reduce the troubles of working with eks/ecs.

now for storage and db, that's a different story.

Docker Swarm. After building a whole CD PaaS for ECS, I came to believe that swarm is a much more reasonable place for teams to start, so I built a free tool for deploying single-machine swarms called Rove. Do your own research though, and remember you don't have to use one tool or platform for everything.

> if that’s literally the only thing preventing them from having positive gross profit

It’s clearly not if the founder isn’t taking a salary and they just had to lay off their only employee..

$50/month gets you a 12 core 24GB VPS with unlimited traffic on a 1Gbps link on Ionos.

https://www.ionos.com/servers/vps

$7500/month would get them 150 such servers.

Maybe they should cut the AWS costs and hire their developer back.

Id be really interested to hear the breakdown of their AWS bill. It would be a crime if they were blowing what money they have giving Amazon 9 cents per gigabyte egress.

If they are self-managing all the extras you get with a decent cloud setup (backups, node failover, load distribution and auto-scaling, multi-region or at least multi-DC for availability beyond single node failures, …), they are going to need an infrastructure person as well as that developer. Preferably two so the one isn't effectively on-call 24/7. And for that multi-DC for availability thing: you might need someone (or assign time from existing people) to manage the accounts with your various providers, you won't want tens+ of VPSs from just one provider like that. Of and on backups & failover, you need person-time (and other resources, but the people are probably the expensive part from the business PoV) to regularly test and adjust all of that, so you can be reasonably sure it all works when actually needed. And you need to manage replacing those people when/if they decide to move on to something new, etc…

Also note that a lot of the things you are paying for (CPU cores, traffic, network throughput) in those nodes are shared resources (that Gbit link especially) and/or have “fair use” policies attached to them, and while the same might be true of cloud providers those policies are often either more generous or (perhaps more important from the business stability PoV) at least better defined.

“Cloud” is still expensive compared to buying and managing individual nodes, even if you add in all the above and the things I no doubt forgot to mention, but it does give a lot more than the same cost in individual nodes than this sort of comparison suggests. And sometimes just not having to deal with all that, keeping the business more focused on its core competencies, is worth the extra expense.

In DayJob we use Azure a lot, and sometimes I see the costs of certain things¹² and balk, and we do still have infrastructure people to manage the platform, but overall it works better for us than managing our own resources more directly. We have an extra complication due to our client base (regulated companies like banks and insurers, who are storing PII of both their own people and their customers with us) in that we have to give a lot of assurances on security and such which would be more work (it is already a _lot_ of work as anyone else in that sort of B2B arena can attest) if we self-managed everything.

----

[1] $2,400/yr for SFTP access to a storage account if you need it available 24/7?! Especially given we have at least one such account per client as their requirements understandably require that level of separation. I think we'll keep using the relay & management dashboard I setup in a few cheap VMs, thanks…

[2] and the performance given the costs: AzureSQL³ I'm looking at you!

[3] though again, some of that cost is in things like the scaling flexibility and other infrastructure convenience, which the business finds worth paying for

> If they are self-managing all the extras you get with a decent cloud setup (node failover, load distribution and auto-scaling, multi-region or at least multi-DC for availability beyond single node failures, …)

History has proven that most of the time these reduce availability than increase them. Any sort of failover and the complicated setups to get it going introduces bugs and issues more than the redundancy it provides.

Have we forgotten the number of large single server applications running on single linux machines that never needed an unplanned restart or had a crash for years? And you can't beat AWS us-east-1 or Azure or GCP in outages lately.

And I doubt any service like this needs auto-scaling. Most services barely will use up a proper single server i.e. something with >96 cores >1TB of RAM.

> “Cloud” is still expensive compared to buying and managing individual nodes > And sometimes just not having to deal with all that, keeping the business more focused on its core competencies, is worth the extra expense.

There are ways to not manage all that and still be in the cloud. It's called don't use AWS or Azure.

> they are going to need an infrastructure person

No.

I run multi-site Ceph+Nomad clusters with NixOS on Hetzner for our startup and maintaining those takes less than 5% of my time.

By using great tools and understanding them well you can do it with little manpower. I learned all those tools in around 3 months total -- so around as much as getting a basic understanding of AWS IAM ;-)

The only thing you don't get with that from your list is auto-scaling. But the with Hetzner the price difference vs AWS is 10x for storage, 20x for compute, and 10000x for traffic, so we just over-provision a little. And my 5% time /includes/ manual upscaling.

Yes, I am oncall 24/7 to manage that infra, but I'd be as well when using hosted cloud services. Yes, fixing a Ceph issue, or handling Hashicorp Consul not handling an out-of-disk situation correctly is more complicated than waiting for S3 go come back from its outage, but the savings are massive. Testing whether your backup restore works is something you need to do equally with hosted services.

So it is definitely possible to self-manage everything, for 5% of one engineer.

> By using great tools and understanding them well you can do it with little manpower.

“and understanding them well” is doing a lot of legwork there. From a standing start how does a startup that has the skills & experience to make the product but not necessarily manage the infrastructure get to the point of understanding the tools well, or even knowing which tools are best to learn to the point of understanding well?

> So it is definitely possible to self-manage everything, for 5% of one engineer.

I can accept that as true, if you have the right person/people, and they are willing (particularly the on-call part).

I'm in a similar situation; what resources did you find helpful for learning NixOS? Tho I could skip that for now and stick containerized, in which case I just need Nomad..but I'm not certain on picking it over K8s in any case. Just knowing I'm gonna have to deal with this soon and you seem to have it figured out enough!

I found NixOps when searching for an alternative to Ansible that is actually declarative and not just a "bash in yaml" runner. Our Ansible deployments took > 10 minutes and were not "congruent" (well explained in [1]): Removing the Ansible line that installed nginx did not uninstall nginx, so the state on all servers diverged over time and we had no clue what was runing where. Docker was also very slow because changing something early in a Dockerfile leads to lots of re-building, because again it's just bash scripts with snapshotting.

I thought "surely somebody must have invented a better system for this" and NixOps was exactly that. Deploying config changes always took a few seconds with that, instead of 10 minutes.

> what resources did you find helpful for learning NixOS?

This was already in 2017 so documentation was worse than it is today.

On a flight I read the Nix, NixOS, Nixpkgs manuals top to bottom. I also read some of the nix-pills, but didn't like that they went so deep into the weeds of packaging when my primary interest at the time was OS configuration management. In retrospect, I should have read those also front to end to save some time later when packaging our own software and some specific dependencies became more important for us. I also read various blog posts, examples, and asked some questions in the IRC channel (now Matrix), where there were some people that simply knew every detail and were willing to spend hours sharing their knowledge (thanks cleverca22!).

I also read key NixOS logic source code, such as the `switch-to-configuration` script that switches between 2 declarative configs (like many, I do not like that this is written in Perl, and I'm sure it will eventually be switched).

A thing I did wrong was to learn too late how to write my own NixOS modules; I wrote our own systems as "plain nix functions" but they would have been better as NixOS modules, because those allow overriding parts of the config from outside, and make code more composable (see also https://news.ycombinator.com/item?id=41355203).

I spent 2 months prototyping all our infra in NixOps and learned by doing.

I also learned specifically where the gaps are: NixOS generally handles what's running on a single machine (with systemd units), and with e.g. NixOps you can access the global config of other machines (to render e.g. a Wireguard config file where you need to put in all machines to connect to, so {all machines IPs} \ {own IP}). It does not handle active cross-machine coordination, e.g. if some GlusterFS or Ceph tutorial says "first run this command on this machine, then afterwards that command on that other machine", or "run this command on any machine, but only run it once". So I learned Consul as a distributed lock service to coordinate (mutex) commands across machines. Luckily, the amount of software that needs "installation by human operator running commands" is continuously going down, declarative config becomes more of a norm.

With NixOS, a good thing is that while it is reasonably complex, it is simple enough that you can understand it fully, that is, for any given behaviour you _know_ where in the nixpkgs code it is. I recommend to use that approach (spend a few months to understand it fully), because it makes you massively more productive.

I also believe that this is a big benefit of NixOS vs e.g. containers on Kubernetes: Kubernetes is big and complicated, with likely more lines of code than anybody could read, and the mechanisms are more involved (for example, you need to know a lot of iptables to know how a request is routed eventually to your application code). NixOS is simpler (packaging software and rendering systemd units); it uses a more radically different fundament but in turn advanced features on top of it are straightforward (multiple versions of libraries on the same machine, knowing for every binary exactly which source code built it, running _only_ what's declared, automatic transparent build caching, spawning VMs that mimic your physical servers). NixOS provides less than cluster orchestrators like Nomad and Kubernetes (e.g. no multi-machine rolling deploys with automatic rollbacks), but one person can keep it all in their head, and it is very good at building things that run in cluster orchestrators. (Disclosure: I know much more about NixOS than Kubernetes; maybe Kubernetes disagree with me and think that a single person can understand Kubernetes source entirely to get the fast directed debugging I claim is possible with NixOS.)

Often, you also don't need a cluster orchestrator. Our Ceph runs straight on NixOS on Hetzner dedicated machines, it does not run in our Nomad. We use Nomad to schedule our application-specific jobs onto our machines -- that is, we use the cluster orchestrator for their original design goal (ball-packing CPU + memory jobs across machines), and do not use the cluster orchestrator as a "code packaging and deployment tool", which is what much of current Docker+Kubernetes is used for. We find that Nix is simpler and better for the latter.

Starting from NixOps, we Nixified all our our tooling (e.g. build our Haskell / C++ / Python / TypeScript with Nix), fixed things in nixpkgs in our submodule and made lots of upstream PRs for it (I'm currently at ~300 nixpkgs commits). NixOS works extra well if you upstream stuff your company needs, because it will reduce your maintenance burden and make other industrial users' life easier too. Especially recommended is to upstream NixOS VM tests for services you rely on; for example, I contributed the Consul multi-machine VM test [2], which automatically runs for any version upgrade to Consul in nixpkgs so nobody will break our infra that way.

Hope this helps!

[1]: https://flyingcircus.io/en/about-us/blog-news/details-view/t...

[2]: https://github.com/NixOS/nixpkgs/blob/72936c3bf6272f05922812...

Keep in mind that this is a hobby project that is currently bleeding money. No more money is going to be lost if they lose their DB, don't have backups, go down for a week, etc. So a lot of the things you mention aren't really relevant to this case.

What would you prefer, this website eventually shutting down because the donations barely cover hosting costs and there's nobody to maintain it, or the website occasionally going down but otherwise actually being profitable enough that the founder can continue maintaining it on a part-time basis and keeping the site alive?

$7500/mo is less than half of one headcount --- if you could get all hosting for free. This is a sideshow. With these numbers, their viability is not determined by hosting costs.

The problem here is that we all have opinions about hosting, but not so many useful opinions about business models, so hosting feedback is what this person is going to get.

$7500/mo easily hires two great 10 yr experience developers in most parts of the world.

Maybe I'm out of touch with how things have changed over the last year or 2, but 2 years ago you'd struggle to find 1 great dev with 10 YoE for that price ($90K/yr)

Even if things have changed now, I can't imagine you'd find a great dev with 10 YoE for less than $80K/yr, and that's hiring globally (with the time zone issues that come with it). You can probably get 1 OK dev and 1 bad dev with 10 YoE for that price though, but you'd usually be better off just hiring one great dev than any other combination.

It's not $90k/yr, it's something substantially less, because the fully loaded cost of an employee is much higher than their nominal salary.

I'm assuming you mean more?

I just meant, you'd be paying these supposed engineers less than $90,000/yr.

Oh you're right, the salary the employee would be getting in this hypothetical would be less than the fully loaded cost.

So the person claiming you can get 2 "great" devs with 10 YoE for 90K/yr is effectively saying there are 2 "great" devs with 10 YoE somewhere in the world who are willing to work for like $35K/yr (or something in that realm less than 45K/yr).

I don't think the market has gotten that bad, but again, I'm not really paying attention to the job market right now.

Right, I would just bring this back to, the fact that we're even entertaining this discussion illustrates how these hosting costs have basically nothing to do with this person's economic problem. With 2MM active users, if $90k/yr is breaking the bank, you haven't figured it out yet.

That's not to say it can't be figured out! But people telling them that they should reduce hosting costs are doing a disservice.

You are out of touch with most parts of the world. Average income worldwide is about $10K a year. So 90K it about 9 times average income and in many parts of the world hires the top 10% of developers of that country.

But It will probably hire less then half of a US developer that then pisses all the other money away on overpriced AWS Servers, so that is right.

Average income is heavily skewed by unskilled workers who constitute the overwhelming majority of the workforce, as well as workers with less than a high school degree which constitute the overwhelming majority of the workforce.

I actually agree with you that 90K gets you top-10% of the developers in many countries, but you were saying that 45K would get you a great developer with 10 YoE.

Devs with 10 YoE are already probably less than 20% of developers, and there are devs with 10 YoE who aren't "great" developers. Perhaps this is my own western near-sightedness talking, but I'm under the impression that great developers may constitute fewer than 5% of developers in some of these countries in question (I assume due to a combination lower English proficiency which is important for working with Western employers, different cultural norms which cause many developers from some countries to adopt anti-patterns for working effectively with teams, and more limited access to reliable internet and better equipment, which could result in lengthened feedback loops during their learning process).

But my primary argument wasn't that you'd be unable to get a great dev with 10YoE, it was that you'd be unable to get two great devs with 10 YoE for 90K/yr, or 45K each, or as tpacek pointed out, more like 35K/yr take-home.

I don't think this has much to do with what I said, and is rather a grievance comment about how much cheaper programmers are abroad from the US. But, if you really believe this, you should be outcompeting a lot of US tech companies with this strat.

That is easy to turn around, if the company would do smarter and cheaper hosting we would ask why they didn't spend more time into finding a viable business model.

It sounds like a lot, but even if they dropped server expenses to zero they'd only have $90k/year to spend.

That is not much, considering they need to cover all aspects of running a nonprofit; they need enough technical staff to maintain a 24/7 on call rota while letting people occasionally go on holiday; and they need to pay healthcare and pensions and suchlike on top of salaries.

And it's not like they can offer a meagre salary now but promise vast wealth in the future, like a startup might.

That's only around $10/hour, which doesn't actually buy one a whole lot of servers/databases/bandwidth/monitoring/logs/etc.

>> which doesn't actually buy one a whole lot of servers/databases/bandwidth/monitoring/logs/etc.

It buys you 150 machines for a month 12 core 24GB VPS with unlimited traffic on a 1Gbps link

See my comment below.

And then you get to stand up your own databases, load balancers, monitoring, logging, etc. For which you need a development team with significant operations experience to do correctly - who will surely cost you more than $90k/year

I get it, AWS looks expensive, but a bunch of their foundational services are real force-multipliers if you don't have the cash to build out entire operational teams.

>> if you don't have the cash to build out entire operational teams

Its just not true that AWS doesn't need expensive experts to get stuff done - it really does.

Anyone who is half decent on the command line in Linux can get all those servers installed and running without the spaghetti complexity of AWS.

The cloud as a magical place of simplicity and ease of use and infinite scalability in every direction - I think its the opposite of that - AWS is a nightmarish tangle of complexity and hard to configure, understand, relate and maintain systems.

Its MUCH easier just to load up a single powerful machine with everything you need. I'm not saying that works for all workloads but a single machine or a few machines can take you an awful long way.

> Its MUCH easier just to load up a single powerful machine with everything you need. I'm not saying that works for all workloads but a single machine or a few machines can take you an awful long way.

For the core service I tend to favour monoliths too, but I would say you are vastly underestimating the halo of other crap needed to operationalise a real website/SaaS.

Where is your load balancer? Your database redundancy? Where are backups stored? Where are you streaming your logs for long-term retention? Where are you handling metrics/alarming?

Bare metal is great, but you have to build a ton of shit to actually ship product.

> Where is your load balancer? Your database redundancy? Where are backups stored? Where are you streaming your logs for long-term retention? Where are you handling metrics/alarming?

What we lose sight of, is that those things aren't as important as we, as SREs would like to think. When you're a corporation of one person, trying to stay afloat, you can just rely on a single big box and spend your time dealing with all the other problems first. Make sure you have an escape hatch so you can scale up if need be, but don't overengineer for a problem you won't run in to.

Have managed all that in the past with bare metal and more (you forgot configuring routers, installing OS, managing upgrades, dealing with hardware swaps, etc). Its soooo much more sane to deal with that than AWS configuration, actually relatively easy for someone half competent. Luckily we can just employ people to mess around full time with AWS.

> Your database redundancy? Where are backups stored? Where are you streaming your logs for long-term retention? Where are you handling metrics/alarming?

Who cares? At this point it's a hobby project that makes zero profit and is bleeding money. No more money is going to be lost if they lose the DB tomorrow. No more money is going to be lost if they go down for an hour or a day or a week (in fact, they might _save_ money if they don't get more AWS charges during the outage).

They have nothing to lose, and about 6k/month to gain by moving to cost-effective hosting, which could actually make this a decent side-project.

> And then you get to stand up your own databases, load balancers, monitoring, logging, etc.

You get all if not most of this on DigitalOcean, Linode, Upcloud, Scaleway, etc all of a LOT cheaper.

> but a bunch of their foundational services are real force-multipliers if you don't have the cash to build out entire operational teams.

No, it's not. As above and for a lot of things AWS' complexity and silly factor can make it even worse. In GCP I can setup a dual region bucket. As simple as that. In AWS I need to setup 2 buckets, a replication role, bucket policies, lifecycle policies and a lot more just to get the same. Force multiplier? As in make it slower? EKS takes longer than the default Terraform timeout to provision. The list goes on...

I think a lot of folks make their lives unnecessarily complicated by trying to do things on AWS in an explicilty non-AWS way (I inherited a startup codebase last year that did this to themselves in spades).

Why go to the trouble of running Kubernetes on top of AWS, when ECS does roughly the same job at a fraction of the complexity?

Why use Terraform when CloudFormation maps better to the underlying primitives?

Less vendor lockin.

My devex team has developed helm charts we can use that automatically detect EKS, AKS or gcp K8s and configure the parts of an app to work with each environment, but the end users of the helm chart don’t really have to care.

Vendor lockin. Knowledge transfer.

And I disagree that AWS is less complex. Managing services across AWS is complex, K8s is as well but I would rather manage K8S on bare instances.

> Why go to the trouble of running Kubernetes on top of AWS, when ECS does roughly the same job at a fraction of the complexity?

Because it doesn't. ECS has its own complexities - perhaps as a result of EC2 fleets / autoscaling groups and more. Suddenly you need launch templates and it goes on. Have you tried updating ECS via CLI? It's largely confusing.

> Why use Terraform when CloudFormation maps better to the underlying primitives?

If only. It has gotten better, but historically CloudFormation has a lot of missing features and still do. Cloudformation can get stuck for hours and you'd just have to wait. The cross region support is terrible. Not to say Terraform doesn't have its quirks, but it's definitely not "worse".

The reliability/uptime guarantees of the cloud providers are dubious, but in this case I don't think they even need to be discussed: this product makes no profit. No money is going to be lost if the thing goes down, because it already doesn't make profit. In fact, just keeping the thing up is making them lose money, so short of completely shutting it down, moving it to more cost-effective hosting would at least mean they can keep it going for longer on their donations.

We're paying ~$1k/month for all our webservers (which is a dozen ECS instances). They're handling about 3,000 requests per second (but that does sometimes massively spike to tens of thousands if not more).

We're paying ~$1k/month for all our tooling servers (so the 150 different test runners, representers, analyzers that are used to check peoples code. There's >=1 of those running every second). Bare in mind, we're running student's code in over 70 languages here. Each is a docker container (often many gb large) - so we pay for HDD too.

The biggest actual cost is the database at $2k/month. We have about 600 queries per second, and around 10MB per second (spiking to 47MB per second) of read throughput. It's an autoscaling database, but AWS determines that it's at the level it needs to be, and if I turn that down, performance suffers (I've tried).

Beyond that, all the other individual services are ~$300/m, so quite small amounts, but for things we rely on (e.g. caching servers, a shared filesystem amongst all those servers, and other things).

$1.2k on tax is also fun.

Thank you so much for sharing your costs! I was very curious to see what the real expenses for AWS services look like, as I've always assumed AWS is massively overpriced.

From the numbers you've provided, it seems like your total cost is around $5.5k, so I assume the remaining $2k is attributed to traffic.

Everything looks quite reasonable, except for the database and traffic costs. I've run MySQL servers handling 150k reads/sec and 50k updates/sec with no issues, even on very cheap machines (around €30 per month). Years ago, we were serving over 100 million pages (of heavy content) per month, and we didn’t even bother looking at traffic statistics because, here in Germany, it’s hard to hit the traffic limits that most hosting providers impose.

That being said, AWS is less expensive than I initially thought. At the same time, I’m confident you could reduce your hosting bill by up to $2k without even leaving AWS by setting up your own database server. Moving away from AWS entirely might be challenging, as managing a fleet of about 30 servers would likely take one or two days of work per week (I'm managing a dozen mostly idling servers and I work one day per month on them). When your hosting bill reaches $30k, I'm very sure it would be cheaper to hire someone (hint, hint ;-)), that moves everything to dedicated servers and manages them.

A part of me is tempted to say 'maybe some cost reduction in cloud bill is possible' but for the scale you operate at and cost you already have, I feel like refactoring the revenue model is the greater strategic 'bang for your buck'

Not to be that guy, but I'm pretty sure that these volumes (talking about ECS and database here) could be handled quite well by a few dedicated $100/mo servers, provided the code isn't hugely unoptimized. So I'm sure you could save (very conservatively) half your budget when going on-prem. That said, it probably won't help you much overall, I would imagine.

I don't want to armchair ops your decisions, but maybe do consider moving to some dedicated Hetzner servers, at least for the runners. You can probably reduce that cost tenfold relatively simply.

AWS will charge for network traffic, both in and out. Would need to calculate what is cheaper.

The runners take some code and return text, right? That shouldn't be too much traffic, hopefully.

Move everything to Hetzner (or the like) at double the capacity that AWS has on average and you still spend about 10% of what you spend on AWS.

That is how stupid costly AWS is and 99% of people using AWS are wasting money like there is now tomorrow.

Problem is Hetzner has such stringent user verification its tough to get hosting on there.

You don't even need to use Hetzner, I'm actually quite surprised they are spending 2k on DB. For the env isolation i understand they need ec2 but...

it shouldn't cost anywhere that amount, for instance I have 3 million users on a $40/month digital ocean VPS with just Django + Postgres and I have far more reads and writes.

for env isolation you could spin up a $5/month VPS and shut it down when idle

Sure Hetzner ist just one example. Many provides sell at about 5 to 10% of AWS prices.

Eh? It's not "tough" to get hosting there... It's super easy if you use legit details and live in a country they are allowed to do business with.

> 99% of people using AWS are wasting money like there is now tomorrow

Why do you think that is? Are that many people that dumb or are there other reasons they use managed cloud?

I've mentioned this before on HN, but we had a single postgres machine primary read/write server that did 4k QPS 24/7. It had DDR ram as storage on a PCIe card, of course, but this was before "SSD" was a thing. It was for a site that hosted portfolios of images, for both people in the images and people who took the images, and such. The front end data (the images and text) was, iirc, 3TB. Sometimes we'd need a server in a new location, so a locked metal briefcase was carried from the DC where the front-end data lived, to our offices, where one of our "IT" people would then carry it on to the new location and offload it to those servers in that location.

Anyhow that database server was probably ~$35,000 all in. That's 5 months of your current AWS spend. One of the things i did during that time was take a 2 generation newer server, a $35,000 1u Dell with 512GB of ram, and mirrored the postgres database into tmpfs and enabled replication, then we set that machine as primary. The new machine didn't break a sweat. So much so that one of the things me and the (really very awesome and nice; Hi, Chuck, if you're out there!) DBA did was set postgres to use no more than 640KB of memory, then ran the entire site, with 4k QPS, on that postgres instance with 640KB of memory (not counting the 280GB of tmpfs storage, of course!), just to prove it would work. It did - although some of the bookeeping queries (not sure what they're called) were taking a very long time, and would have had to be refactored to use less temporary memory, and such.

anyhow my point is, there are people out there that can do things cheaper, or faster, or more efficiently than whatever you got goin on right now. Your statistics on "per second" usage and the like don't sound too demanding. If you could squirrel away $500/month for a few months, and you ask around for someone that can rack metal and has peering, there are people (including me) who could get you co-lo in <16U[0] with redundancy, where your only monthly infra charges would be the co-lo fees.

[0] Old, extremely beefy, but large servers are generally 4U, but dirt cheap for what you get. Ex: 80 thread, 512GB RAM, 8 SAS bay HP server, $800 shipped. And i bought those 6 years ago. However: 5950x, 128GB RAM, 24 SATA port can be had for <$2000 (i'm guessing based on what i paid a few years ago), and that's roughly equivalent in power (kernel compile takes 3 seconds longer on the 5950x but it uses 1/4th the power at the wall). The reason i tagged on <16U is because at most you're gunna need 4x4U, two "front end" and two "back end" machines, with duties split and everything redundant in the rack. I haven't looked in a while to see what's available on ebay as far as more density, but for sure 16U or less!

The issue becomes: how do you find someone who knows how to do all that, that is willing to work for next to nothing because they believe in the ngo/nfp? Maybe there's a tech forum that people like that read, who knows.

good luck, and thank you for doing things to help other people. I hope it all works out in the end.

email in profile.

RAM as storage for DB? So data loss on reboot? Sounds like very specific use case.

https://en.wikipedia.org/wiki/Fusion-io

maybe i misspoke - at one point there was battery backed DDR ram they used in PCIe, but by the time i came around they were using Fusion IO PCIe devices, which i guess were NAND flash, not DDR. or, alternatively, that is how it was explained during onboarding - "it's like DDR on a PCIe card, so the iops are 1000x that of SAS 10k drives"

unless you're talking about our experiment of tmpfs - then yeah, the use case was "genewitch heard bill gates say 640k should be enough for anyone; here's a super beefy machine to test that theory; theory tested." We didn't run the site live on that machine for more than 10 minutes or so, we switched it back to the fusion-io backed server immediately. It was a proof of concept about one of the things we could do with these new servers - read replicas with the DB in tmpfs for extreme speed and no IO blocking.

I was curious, so I just checked on one of our customers (I work at a small MSP in the UK) by way of comparison and we have on chugging along happily at more than double those numbers on a $288 Linode dedicated CPU instance. And we're only on that size for ease of disk space handling as the database is several hundred GB. CPU is basically at zero, it's the disk IO that actually gets you on some of these busier databases (from my experience).

RDS is extremely expensive. All managed databases are.

That said, it's a trade off of convenience and being in the AWS bubble, and weighing up the pros/cons of separating out services. Data Transfer is another thing to consider too of course. Sticking your database elsewhere might cost more in egress traffic communicating with it from your other AWS infrastructure. If you're all in on other AWS services, sometimes the RDS price is just worth it when it comes to the total price. Sound like this might be the case for your setup.

I hope you do manage to work things out. The service you have is great.

PS - Side note on RDS sizing. You might already know but sometimes it's worth increasing the storage size on gp3 type storage above 400GB (if you haven't already) as you get 12,000 IOPS baseline against 500MiB/s throughput[0] when you have that much storage. That's 4 times the below 400GB baseline performance but you only pay for the additional storage cost. It can make a difference if you're IOPS constrained or trying to deal with bursty traffic but want to use the smallest instance size possible otherwise to save costs.

[0]https://docs.aws.amazon.com/AmazonRDS/latest/UserGuide/CHAP_...

Cloud providers have gotten really good at selling scale anxiety. Very easy to reach $10k/mo when building on a modern web stack using managed services. Kafka alone could get you there once you factor in multiple environments and add-ons.

Maybe this will be helpful to the author (partially already mentioned):

- Nonprofit business model does not equal "everything is free for the user forever", I'm guessing you already know that, but the wording on why you don't believe in nonprofit business models explicitly mentioned keeping everything free as the reason. You can earn revenue from users in a nonprofit business.

- You have a big audience with good engagement for the segment, there are multiple ways to make money without abandoning the core mission (job boards, screencast upsells for advanced courses, premium content, whatever else, look at how Remoteok.com makes money, copy-paste as the founder is super open on his process)

- Being a for-profit business and fundraising, will temporarily solve your issue of having funds to run a business. It will not solve the issue of not knowing how to/being afraid to charge users or other parties for the value they get out of your product. You could already be solving this problem today, and you have a 2million audience pipeline built in to solve that issue.

I'm not dismissing the challenge of some business segments being extremely difficult to make money in despite the value being meaningful, I work in healthcare, so I know, but since your new business will effectively be in the same segment, do focus on the revenue aspect much sooner and much than you think you'll need to, because you already know what happens if you don't.

And big respect for what you've built in a super crowded space, you obviously have the product and user empathy chops needed, wishing you the best of luck on nailing the business chops!

Nice post! Nice site! I just added 70 links from PLDB to your site.

Some thoughts before I donate:

> I've lost faith in the nonprofit business model

Great! You will find a way to generate revenue streams. This is a good pivot! I would recommend going fully open source and public domain as you add a profit model. IMO public domain for-profit companies are superior to (C) non-profit companies. Profits don't make you evil, closed source and copyright does.

> every PR in a live Exercism repository should get a review before it can be merged

Why? I recommend just merging in. Quickly fixing/undoing problematic commits.

> $7,500 in monthly donations. That covers our server cost

OMG!!! Stop! This money should go to you 3, not cloud providers.

Server costs should be $0. This should be a local first, entirely client side site. Maybe 1 $50 a month droplet if you need some read/write user stuff. With less than 100K active users in a day, 1 server should be more than enough.

Finally, ditch the (c) symbol at bottom. Go public domain!

Here's my user test and authoring of this comment: https://www.loom.com/share/763e480438c4481ba0aa056c6ef0cbbb?...

Where the fuck are the VCs looking at serious hackers with millions of users who haven’t gotten the business model dialed in yet?

Y-Combinator says “build something people want”. Two million sounds like a lot of people.

I’ve only glanced over this so I could be missing some critical fact, but superficially it sounds like pmarca or pg should write a check before the rest of us stop believing in the startup lottery.

Well, exercism is a "not-for-profit organisation registered in the UK" according to the footer of their website.

Edit: It appears that their "not-for-profit" message in their footer is a bit misleading, as explained by those who replied to my comment.

OpenAI is a 501c nonprofit. Doesn’t stop Sam taking big time meetings more often than he checks his email.

Part of OpenAI is a nonprofit, not in its entirety.

Matt Levine uses a diagram to allude to the complexity. And that was before Larry Summers and the National Security Administration got involved.

A better term for OpenAI would be “rogue state”.

https://www.bloomberg.com/opinion/articles/2023-11-20/who-co...

They're not officially a non-profit, they're a limited company. Obviously they can still operate as such - that's down to the board how they distribute profits via dividends, so they might choose to do so.

A true non-profit company in the UK would be a registered Community Interest Company, or a registered Charity.

Hello. This isn't actually correct. Exercism is a Company Limited by Guarantee. It has no shareholders. It can't pay dividends.

Community Interest Companies do have sharedholders and can pay dividends. CIC's are definitely not charities in the scope of the word that you're using it.

They've not set up the structures for it to be not-for-profit here, they're just an ordinary registered company.

Typically you'd form a Charitable Trust or Charitable Incorporated Organisation, and Exercism isn't listed as either on the Charity Comission website.

I've replied to this further down. Exercism is a nonprofit - we have no shareholders. We've had many large organisations donate to us (including Mozilla and Google) who have vetted all this.

> Where the fuck are the VCs looking at serious hackers with millions of users who haven’t gotten the business model dialed in yet?

Busy investing in AI?

These guys should "pivot to AI". Should be able to get a full B for an Ai company with 2 million users.

Fuck I’ll train a LLaMA tune for them on Saturday(s) gratis.

So Exercism AI?

Sell the code submissions as training data.

There's nothing special about a VC. It's just someone with money to allocate. Go ask them to give you a big chunk of the company for $25k if you think this is just a few steps from being a blow-up success. Being able to differentially detect when a company is successful is a pretty good skill. If you think you've found a diamond in the rough, you should go in.

This absurd hagiography about efficient markets allocating capital will be viewed by future historians like some kind of flat earth theory. Eugene Fama himself is walking back the idea and he won a Nobel Prize for asserting it.

Markets are what people do when they’re not having sex. No sense in fighting markets.

Cartels and market failures are what happens when the holders of capital aren’t terrified of Lina Khan or Maximilian Robespierre for longer than 12 seconds.

The Battery Club is for kingmakers who decide things together for their mutual benefit behind closed doors. They learned their lesson after Don’t Poach Gate: it’s not in writing anymore.

Only real metric I see is 800 monthly donors and $7500 from them ($10 per person) - that's what that business is worth, those are really people that want (they pay so that's what matters).

Having 1600 people sign in daily and never logging back again - creating 2M zombie accounts is I suppose critical fact that you are missing.

If it’s two million churned out dead accounts, non-replaced, and 800 actual people who care then the author of the post was posting in bad faith and I’d retract my remark.

To get from 800 real people to two million user id rows one would be running some kind of spam scam, and if that’s the case I apologize.

I don't think it is case of spam scam.

I do think they really can have legit sign ups, but people "wanting to learn code" usually fizzle out quick. I have my experience as a dev when friends/family/colleagues ask for guidance, I provide them with links resources and some mentoring and after 2-3 days they don't follow up ever again.

But not using metric like daily active users, returning users, might indicate that they want to inflate their worth or they convinced themselves to believe in the wrong metric.

Oh I don’t doubt for a minute that there’s some gap here. But the key people are serious enough to list rigorous code review as a core value in the next breath after “I just stopped paying everyone including myself.”

If the millions of users thing is even a little real, that’s a galactically less risky play than friggin LangChain for Pets or whatever.

I hope I’m being paranoid but I get the sense that someone pissed off someone important.

Well, the standard quote leaves out "and will pay for"

> The one area we have had some promising success is in advertising on the site. But the effort it takes to find advertisers and manage them, and my general desire not to flood Exercism with adverts, has meant that I feel this isn't a very sustainable strategy.

To me that begs the obvious question... what about trying some automated advertising like Google Ads? Probably not the most theoretically optimum use of their ad space, but minimal effort is involved.

so basically letting google be their business???

Better than shutting down entirely I think.

That probably requires investment in good content moderation and monitoring the account. There are people out there who like to report anything that violates T&C or create such violations themselves and then report them. Either for the kick or could be done by competition.

What I am trying to say, there is no free lunch.

I haven't used this site but I did run a service a while ago that had 100K free users. All the talk here about not using AWS is bunk. Not needle moving. OP needs to bite the bullet and charge their customers. At least some of them, for something. Heck some of their traffic probably comes from bots. Bill them. Make a white label version and try to sell that to big customers. Actually the other way round: try to sell it then make it if they bite.

Sounds like a very useful service. I just signed up. Did you consider to inform new users that the service is financed by donations? You seem to focus on "always free" which is nice but I guess most people are also ok to pay some money if your service is useful to them. I guess this might drive some more donations if people knew that you need money. Also a banner like Wikipedia (maybe a bit less annoying) would work.

Hello. You'll get an email after a day telling you about it, and you'll get a banner after you've submitted a couple of exercises :)

My heart is in pain from reading this post - as everybody's who had a startup & encountered the "poor fit" argument. I particularly feel for you regarding the problem of being more generic than needed when people prefer smaller scope (happened to me a few times).

I don't have a solution for you, and you probably tried a lot of prudent things. I assume you have already contacted people that buy into educational software (e.g. Kevin O'Leary of 'Shark Tank' fame)?

I wish you some kind of miracle that turns things around; sometimes once conversation can change everything. Surely this user base deserves that the site finds a way to sustain.

What does the "poor fit" argument refer to?

> For the last few months, I've been working on a new educational product teaching coding fundamentals that I'm going to launch in 2025. 96% of people who try to learn coding give up - which I find unacceptable, so I'm aiming to change that.

Feels like you've been kickstarting a competitor/pivot on the side.

With those stats, even with just 4% of people wanting to learn, and just a fraction of them actually paying, that would be dozens of thousands of paying customers. Even at 1USD per month, it seems a relatively sustainable business.

Yeah, I think he is confusing "non-profit" with "non-payment". You can have a non-profit company with paying customers. I think he need to take a basic economics class.

I think the tone of this comment is unnecessarily condemning (apologies if I am reading it incorrectly) – I can totally understand why someone who has worked for years on a non-profit business which has created a lot of value for people (I think exercism is pretty good software) might want to be able to at least pay themself.

Non-profit doesn't mean you didn't get to pay yourself. It just means that any revenue that exceeds expenses must be reinvested into the business.

That sounds like Amazon to me.

Except amazon shares would be worth much less if there was a promise of no dividends ever

Thanks, yeah. New product will hopefully mean I can keep funding and growing exercism. They're very synergistic and will support each other.

Charging people $1 on sign-up would probably lower that 96% metric

I don’t know if Jeremy will see this, but:

At my org it’s hard to get justification for donations.

However if this was something like ~100$ a year or whatever per dev, I’d be paying it today for my team.

I remember reading an article about a non-profit discovering a work around for this problem.

They created a subscription plan that had no real value except maybe adding some trivial "feature" like a gold star next to their name (I don't remember the details). This made it possible for the companies to receive a standard invoice that had a greater chance of approval.

Yeah, can't justify donation to corporate, need some reason to pay for stuff.

Hello :)

Appreciate the comment. If you have thoughts on that, I'd love to hear them here: https://forum.exercism.org/t/exercism-teams-coming-soon/1266...

out of money, final payroll, no alternative but to go into zombie mode, yet for the last few months the co-founder has been working on his next product which will launch in 2025.

maybe everything that could have been tried was tried but this comes off like some of the effort spent working on the next company could have been spent keeping this one alive

They said they will be using the proceeds from that for-profile company to pay for a team member's salary and keep Exercism running as much as they can. So this next company could be considered to be an effort to keep this old one alive!

It sucks being out of money and then you have to fend off the "why aren't you more dedicated" line of questioning when you had to find an alternative to keep the company alive

Oooh contingent on the next unproven idea being economically viable.

Amusing. I walk away from all business partners that think this way

They hand waive away “who will pay what”

> but this comes off like ...

You don't know more about the company than them.

Most likely they saw this coming months ahead of time, and needed to figure out how to pay their own rents after actual doomsday and needed to pre-emptively start figuring out what's next.

Founders who don't start with a pre-existing financial cushion have very, very little job security if they underpay themselves and their businesses shut down without an exit. Also, founders tend to be generalists, and most big companies don't need more generalists, so it's not easy to get hired, doubly so if you have wasted a lot of your brain pitching to investors and having coffee chats with clients. The harsh reality is, the more you are forced to chat with investors and customers instead of working heads-down, the more you lose your hard technical abilities that other people would hire you for.

As a founder of a shutting down company, if you want to pay your rent on time, you need to do one of 3 things: (a) pay higher salaries prior to shutdown to give you some time to figure yourself out (controversial), (b) start studying to be a specialist well ahead of shutdown and get on the interview treadmill, (c) start working on your next thing well ahead of shutdown.

If you do none of those things, things can get really dangerous to your personal finances.

Exactly. Thank you. Also, I'm using the new thing to keep exercism funded. But even if I wasn't, your post would be entirely true!

That's too bad.

I used to contribute to Exercism a lot (I'm still technically a maintainer but haven't meaningfully helped in a while), I think it was the first OSS project I contributed a large amount of effort to. I was more or less the sole maintainer for C++ for a while and also helped with the C track.

It's definitely a good service and OSS project, I recommend it a lot. Fundamentally though people don't want to pay for things that they can get for free, or they're used to getting for free.

If people actually want OSS projects to survive, you should be giving a little bit of recurring money. I set aside $XX/month to distribute to projects and I don't even work anymore.

I have a theory that developers are a difficult demographic to market to and monetize off of.

Its bc we are skeptical and frugal.

Why not charge 5000 users money instead of servicing 2M users for free?

It seems to me if you can’t get 1 in 500 of your users to give you $10 per year then maybe your product isn’t actually valuable or useful and should just shut down.

I say this without knowing anything whatsoever about this product or its utility; it just seems like a general thing. If you have this many signups and you can’t charge even a tiny tiny fraction of them, it means the signup number is irrelevant.

They're charging 4000 users. The paywall is ethical.

They say in the article they tried to force 1000 random users to convert to paying account and none of them did

Can you point out where they said this?

Can't you just add in Wikipedia-style donation request popup/banners for everyone in general? I reckon plenty of people will be willing to spare some change even multiple times when they log on, just like they do with Wikipedia.

If anything Wikipedia as an example of how unsuccessful a voluntary donation model is.

0.2% of people who use wikipedia donate to it. Unless you are doing tens of millions of unique views a month, it's probably not gonna be sustainable.

People really hate paying for things. Even things they use and love.

the only way is to lower your operating costs but you are correct ppl on the internet simply do not like paying for stuff even donating is a tough ask.

even getting people to click on affiliate links that won't cost them anything I find is tough as most people remove the referral link or purchase on another device probably out of privacy concerns.

Also it's a subtraction from the legitimacy bank account. It's well known that Wikipedia doesn't need those donations.

Endless telethons for public television in America annoy 100 people for every one that donates. Despite that you see ads on public TV for the Archer Daniels Midland corporation all the time even though, allegedly, there is no advertising on public TV.

I don't know enough about Wikipedia, though I have heard this point made before.

I do know a lot about noncommercial broadcasting in the US though, and your comments appear superficial.

A sponsorship message is different from a call-to-action advertisement. The former is permitted, the latter is not. They are not different enough, in my opinion, but they are discernible.

Public TV and radio station finances are public information. Audience fundraisers are a meaningful portion of their operating budget (~25-50%), though it varies by market. As are program underwriting (up to 50% in strong markets), and often CPB funds contribute a few points (this is highly politicized and if you overindex on it, your news sources are likely polluted).

> It's well known that Wikipedia doesn't need those donations.

I disagree. Wikipedia is similar to Exercism on that count. A tiny fraction of expenses are hosting the encyclopedia. It's also necessary to pay people to program the software underlying MediaWiki.

People aren't bothering to edit anymore either. Search engine summaries, voice assistants, and now large-language models repackage information from Wikipedia in a way that discourages people from going to the website itself. Most low-hanging fruit is gone, and the standard of quality has increased.

There's a whole project dedicated to helping undergraduate students edit Wikipedia as a part of their coursework now. It's a great learning opportunity because a Wikipedia editor will check your sources more than the average TA. The flipside is that an undergraduate degree in the liberal arts isn't necessarily enough to write at a Wikipedia level, which is a problem.

These external factors are decimating the volunteer community and the visitor population. Spending on outreach, staffing, and grants to local affiliates is much costlier than having people wander onto Wikipedia and contribute, but that's no longer the era we live in.

This is the problem with ed-techs. Nobody wants to pay. Those that do (Universities, schools) are also competing with the platforms to some degree.

It's hard to build a business model around education because the product succeeding removes the user's need for it in the first place.

Kinda like a dating app, where maximizing usage means not getting you hitched (too quickly).

One of my theses as an edtech founder is that the education can't be the core business model long term.

Switch hosting model. Assuming you run expensive cloud services for $7500.

Get some inexpensive bare metal servers and you maybe can save 80% cost.

Their model may be one of the few that would truly be quite challenging to pull off with bare metal—running sandboxed code for 73 languages is hard to do economically with any setup, but being able to lean on cloud services for provisioning would make it much more achievable.

ah I see that might be a bit more tricky. I was wondering why the costs were so high.

For example, I serve roughly 3 million users on a $40/month VPS. I only run Postgres + Django, if I use Supabase and all these other fancy cloud SaaS I wouldn't be able to sustain it as the costs would increase 10~20x fold.

(supabase ceo)

our postgres offering costs roughly equivalent of RDS so if you use it with Django I don’t think you will come close to 20-40x a VPS, ymmv, depending on your workload

We have an Auth service, but since you use Django you wouldn’t need it - sounds like you’d only need postgres (like a lot of our customers)

I don't think you get the point. RDS is expensive, just because you having pricing parity with it doesn't mean that you are going to be at the same pricing level as cheap hosting.

And thats fine, you have features on top that can be worth paying for certain customers and expenses of your own to pay. Those of us that have been around in this business for 30 years know how expensive PAAS and IAAS can be and how easy it is to roll your own solutions and when not to do that.

True, but I wonder what the cost split is. I'd expect the actual "run code from users" bit to be a minor line item, with the bulk being the actual app servers, databases and egress bandwidth.

They can very well keep the REPL/"run code from users" bit on AWS, but serve their actual website, application and database from fixed-price bare-metal servers.

> I think it's fair to say that at this stage I've lost faith in the nonprofit business model working in a way that allows Exercism to reach any of its potential. Keeping something free for everyone relies on either the user being the product, or on significant donations, and without either, it's very hard to grow.

Non-profit =/= the product is free. The vast majority of hospitals and universities are non-profit, and their product/service sure as hell isn't free.

Nonprofits operate under a non-distribution constraint, meaning any surplus revenues must be reinvested into the organization's mission rather than distributed to private ownership.

What they've lost faith in is running a business with a free product.

This are the LinkedIn comments on this post: https://www.linkedin.com/posts/exercism_activity-72374803032...

They shouldn't have advertised 'free forever' on the website. OMG. In that area, people don't even think about the next day.

One of the things I learned is that you should never offer even one tiny ounce of added value which the average user is not capable of perceiving.

Literally the entire idea of profit rests on arbitraging between perception and reality. You don't want to pay for surplus value that is beyond your customer's perceptive capabilities.

Like many others I had never heard of this service. I looked and I am very very impressed. The way topics are laid out... someone put a lot of thought into this.

I would love for this site to continue and thrive.

I also agree with some others here that the hosting at AWS is probably best left to organizations that have a lot of spare cash. It is way overpriced. I self host a fairly busy site. It is mostly hands-off. I have never really fully understood the AWS premium.

Hello. Blog post author and Exercism co-founder here.

Thanks for all the comments, especially those kind supportive ones. I've replied to a few, but I thought it would making a couple of general points for context.

1. I feel something that's missed is that if Exercism just continues to help tens of thousands of people per month and has no-one working full-time on it in the future, that's not the end of the world. Our donations cover server costs (thank you, donors!) so there's not an existential threat. And our wonderful volunteers keep our content growing and up to date (we've added a dozen new languages already this year). For most of Exercism's existence, it has had no paid staff. Me and Erik having to become volunteers again is ok - it just means things will improve slower.

The blog post was written as an update to our community, and our community has been on this journey with me over the last few year. They've seen us go from zero money, to raising over $1M in donations and hiring people, to not having follow-up donations and having to let those people go. They have a lot of context that's maybe not seen by people new to the story. The post is sad, because Erik, who's well loved is leaving, and because the last couple of years haven't financially gone in the direction we'd have liked it to. But the post is not me complaining.

Exercism hasn't been financially successful during the last 18 months, but it's fulfilled the core part of its mission by helping hundreds of thousands of people. And it will continue to help people moving forward.

2. Our aim is to get people to learn a programming language and then to get on with their lives. It's actively not to monetise people. That's important, because it changes the model a little. We don't expect people to stick around month-after-month doing stuff (one person said that having 70k MAU is terrible - but you can also see it as totally expected - you sign up, learn, leave - that's the point). Which means having some recurring Premium offering doesn't massively work. Nearly all the people who donate are also contributors, power-learners (learning multiple languages), or people who feel aligned to a world where something like Exercism exists, and want to ensure it continues to do so.

So the only real end-user model is to charge an upfront fee for using Exercism, which actively goes against the mission and doesn't honour the commitment under which people have contributed.

3. To all the other monetisation ideas, thank you for them. I appreciate you all wanting to help. We've considered and experimented with lots of them over the years. We've not found one that fits well yet but maybe/hopefully that'll change in the future. I have areas I will continue to peruse and am hopeful about.

4. There are numerous threads on server costs, and while I appreciate the opinions there, saving some cost there wouldn't massively shift the needle. Our donations cover our server costs. Hiring good developers and other people costs a lot more money than the ~$3k/month we'd save. I'd also say that there's some general under-appreciation of the complexity of running student's code in 73 programming language sandboxes in real-time. Exercism is a lot more complex than a webserver/database model - we have over 400 GitHub repos. I'm certain some ops gurus would be able to do a better job at building a server infrastructure than I've done for less money, but the one I've built works with basically zero maintenance and not having to be on call 24/7/365 is something I need/value.

As I said in my blog post, I'm going to take a couple of weeks out now, and come back hopefully re-energised to keep building Exercism, and the new project alongside it, which hopefully will help keep Exercism funded. Thank you again for all the support, ideas and well-wishes.

---

Changelog: - Added "For most of Exercism's existence..." to the end of that paragraph.

Sorry for the additional unwanted advice, but have you considered a pay-what-you-want pricing for the upfront fee?

You can set the minimum to zero and a default above your target, and perhaps ask only after a short free trial period. I heard that it often yields better results than we would expect.

hello @ihid - would love to support Exercism through the Insiders program as your site is great resource for developer education! Is there a way you can provide PPP (Purhasing Power Parity) pricing for people outside of the U.S. and Europe? For example if you use this PPP calculator https://ppp.jackmcdade.com/ and compute PPP pricing for the Philippines, $10/month feels more like $20.88. If you want to recalibrate it, the PPP Pricing would be P4.79. Thanks in advance!

> My plan is to give beginners a rock-solid base, then funnel them into Exercism

Shouldn't you be doing this the other way?

If you have 2 million users, then just make something that you can sell to them.

Whenever you have a bunch of customers paying you $X there is a subset of those customers that will pay you $(X+Y).

You don't have to "make them the product" in the sense that you would be selling their eyeballs, but you're teaching people to code, surely some of them will want to get a job as a coder afterwards.

What about creating a paid placement program where you give recruiters at companies hungry for tech talent access to the coders that perform the best on some tests? The coders can voluntarily take the tests, so they know their info is going to recruiters.

You could also offer certifications as a service to companies looking to test proficiency of their existing staff or create a subscription model for use by companies who want to upskill their existing staff.

With 2m users surely you have people using it from all over the world. Companies might pay a premium to hire the best up and coming talent on Upwork because generally their hourly rates are not as high as experienced devs on the platform, but you can tell based on the code they're writing that, despite their relative inexperience, they're an outlier in terms of coding skill.

You could get affiliate revenue or commissions by selling other, paid certifications such as for SAP/ABAP, Cisco, Microsoft and the like.

It just seems like you can take this as the platform product and upsell, rather than trying to create another product that you sell to people who aren't already using Exercism.

I just don't agree with this leetcode industry. That, recruiters and the whole tech hiring industry that has spun off from a process originally intended to get to know and gain confidence in a potential employee. It is just I have doubts in the educational help these platforms really deliver given that there are already so many resources to learn these days say compared to 20y ago.

We're not really part of the "leetcode industry". People have used Exercism to learn new programming languages for a decade. We've deliberately not wanted to become a place of competition or a recruitment space. We've always focused on helping people learn based on their intrinsic motivations.

Sorry you are right, what I posted isn't fair. I should have read more into it.

No problem. Thanks for acknowledging that! :)

Exercism is a great site and I'll be sad if it goes away entirely. It's a really wonderful service and community.

I used it [throughout 2023](https://xavd.id/blog/post/12-languages-in-12-months/) to tour new languages and have been on a deep dive of Rust this year in their [48in24 program](https://exercism.org/challenges/48in24). The UI and tooling is great, there are real humans available to mentor and everything is open source and volunteer driven. They make learning fun and that's no small feat.

Of course, things like this cost money to run, especially when there's so much going on behind the scenes. Sad that the current iteration hasn't worked out financially, but I have every faith in the team that there are good things ahead.

I've completed Exercism courses and they are great.

One thing that was obvious to me is that if you want to focus on a specific topic the $15/month subscription is too cheap, you can pay, take the course, cancel because you are done with the course, all within the month.

The other thing that is challenging is that the majority of programmer is either too cheap or "too broke", $15 USD is a meaningful amount of money in other countries with weaker currency, and in countries with strong currency I've met people paying thousands for a university diploma, doing a well paid internship that "could not afford exercism". That's though, I hope I never have to build a business for software developers.

All of that shared experience comes from 3-4 years ago, so prices, models might have changed since that, but if they go under it's sad, their UI and their UX was so clean, fun, interactive, pleasant, etc, it was a joy to study using their material.

Also internet is flooded with similar learning websites. First time I've heard about exercism though. Looks cool.

Completed the whole c++ and PHP paths there. It is by no means a way to learn those languages in my opinion.

Then what was the value for you?

I really liked the gamification aspect of the site and I couldn't leave the course unfinished!

Should offer some paid structured courses for $200, and offer a 90% discount for existing users.

Without revenue any company will fail regardless of popularity. A firm may acquire the company for the customer/lead list, but I wouldn't count on that kind of valuation matching the work your team put into the projects. =3

The biggest issue I had with the tech boom and the current tech culture is when they replaced "profit" with "people" as a metric of success. To me they took the phrase "people over profit" and turned it into "people are profit".

It is time investors businesses and started looking at the bottom line again, even non-profits. I mean can you imagine if we based the success of a store with how many people walked in even if they did not buy anything?

Even non-profits need to make money, so their business model was either wrong or they do not have a product worth paying for. On top of that we are probably in economic stagflation so we will be seeing more stories like this as the months go on.

So you need to charge, but don’t want to lie about the 100% free part even though you can change that

So you need to upcharge. It is 100% free… to start

and then there is a subscription that is paid to finish the language course, shown in monthly cost but charged annually by default

more languages? charge more

You can sell the account data to recruitment firms. Everyone hates this but you know that your users are the product, your users know that they are supposed to be the product. Keep a cut with the recruiters so they have to pay you a little commission when candidates land a job

I'm not sure whether to be appalled by your suggestions or impressed by your blatancy

wait till you see educative.io‘s pricing then

85% off on the annual plan, just that it’s $200, but at least you’re not paying $99/month anymore

the interesting thing about the not “not-for-profit” model is that all you have to do is just undercut educative.io by a tiny bit

What I did once in the very early 2000s was to white label a free platform so that other people could license the software from us for $$$ to pay for our free flsash game platform in 2001.

Archchair business strategist, snippet from the homepage: "Exercism is fun, effective and 100% free, forever"

Business model: "I've lost faith in the nonprofit business model"

Also noting changing in the IT market, tech layoffs, AI revolution, less demand for tech workers.

Hope it will work out for you!

Kudos for being passionate and providing stuff for free.

A lot of people will probably tell you to add a paywall, not sure it's breaking the "non-profit business model" if you keep the price really low. But at the same time, even a low price would deter some users (myself included) and it would also devalue your content :/

Thanks :)

Is there a way to get funding through libraries? Kind of like how it's possible to read subscription journals and newspapers for "free" in the library. The library must be paying for the subscription right? I think there is a strong values alignment with public libraries and exercism.

Exercise seems like a goldmine for coding LLMs:The same challenges solved again and again using different languages and user provided feedback. Much better for analysis and training than random GitHub repositories.

Have you tried selling access to that data to the usual suspects ?

Given that you are not making money on this at all, it would be wise to consider charging a small amount. If the number of users drops because of this, you still won't have lost any money. But those who choose to stay will help make this sustainable.

A shame. All the random garbage that comes out of things like ycombinator, promoted by 'visionaries' like graham, thiel and other bottom feeders, and something truly useful to the world like exercism is starving for funding :|

Exercism looks really cool! Have you thought of reaching out to philanthropists, particularly in tech? Just a wild guess but I believe there would be at least 1 wealthy individual out there willing to donate money for this cause.

It does not help that many expect learning material to be free online. This ends up capping the quality of learning exp as the orgs can’t re-invest earnings to make a product better.

I don't understand. I don't even know what exercism is.

You created a site that has 2 million users and don't know how to monetize? Just charge people $5 a month to use. It will solve all your problems.

I was donating, but my skill level is not good enough to successfully use the site. It's not really a place for beginners though, so once I learn more I will probably sign back up

Wow this is really surprising. Exercism is simply the best put together language training program I've come across. I hope they find a way out of these headwinds.

Adsense alone would more than cover this.

Of the 2M users, guessing the MAU is much lower, which would render the ad model non viable.

Still always sucks to see solid products rendered unsupportable

> Adsense alone would more than cover this.

God please no. This is already dreadfully normalized and the web is utterly drowning in low-effort link spam, dick pills, cryptocurrency and hot Russian singles. I can't think of anything sadder than taking this lovely product and letting Google piss all over it.

> I can't think of anything sadder

I can

Seven years ago, excercism helped me get through the children's cancer ward with my mind intact when my daughter got treated for lymphoma. Thank you

This a political/economic system failure. How do we design a system that allows for entities like exercism to be funded/supported?

Can I buy in as a share holder and monetize this?

First time I've heard about them. I hope this helps them get a bigger donation pool and support from similarly minded organizations

:( exercism has been great. I've been subscribed to the insiders program for more than a year now, I wish I could help more.

Can you perhaps issue certificates which we can get after completing the course? USD 50 - USD 100 isn't that bad.

How can there be NO money in 2 million users?

I would be more interested in active users. I have had personal projects with 100s of registered users yet I was the only one who actually used it. So 2M users might translate to somewhere between 1000 and 10000 unique users per month. Considering the niche; many of them young people with very little discretionary spending money.

We have about 70,000 MAU.

People tend to spend an intense few weeks learning a language, then disappear for a year. Then come back and learn another language. A large part of the 2M is still "alive" (I don't know exactly what number that would be, but I'd think around 600k) but Exercism isn't the sort of product you use day-in-day-out.

At 1.500 signups per day (45K a month) 70,000 MAU are horrible retention. Usually you need about 10 times MAU to signups. Unless you can make that happen you have nothing that people actually use and not just try out.

It takes people about a month to work through a language track, learn the language, and get on with their life. So I wouldn't expect to have more than 45k MAU. 70k MAU is because people LIKE the site and come back to learn another language.

We're not trying to be some business that sucks people into using our site every day and monetise them. We're aiming to teach people effectively and efficiently and then get them moving on with their lives.

We could make the site less efficient, or change it's purpose, but that's not the mission of the organisation.

If there is nobody using your service for more than half a month (at that point they would mostly already be counted for 2 months) almost nobody will ever pay. What is your day 7, day 30 and/or day90 retention?

If is low all you can do is lower costs, which should also be straightforward.

They have 800 monthly donors totalling $7,500 per month. That is some long tail but I doubt the 2 million are active users.

> Right now, we have about 800 monthly donors and about $7,500 in monthly donations. That covers our server costs pretty much exactly.

For only 2M registered (not even monthly) users? Tell me you are using AWS without telling me you are using AWS.

Let's think about their heaviest workload, which is likely running user-submitted code server-side to test those exercise submissions. Say it's really bad performance-wise and costs 1s of single-core CPU time per submission. 45m submission gets us 520 days of CPU time.

I'm going to go out on a limb and claim their entire site could run off a $50/month dedicated server, and it would be unimpressed.

Just add a premium tier or partner with coding bootcamps. Don't give up so easily.

I wonder why this would not be a fit for GitHub and Microsoft...seems like it could complement GitHub...

Do a paid job boar within the website itself Or paid mentorships or something like that

why not charge for access? There is plenty of people making money with subscriptions and courses in this space. You just need to gate the content. You have an audience already, bait and switch. Money will come.

Why is this free? That's the problem here. They'll easily get 10-100x their donation amount if they make this freemium. They don't even have to degrade the core experience.

Life takes energy. You need to make some of it back to survive. Asking for money isn't a dirty thing.

it's that any profit made must be reinvested into the organization to further its mission, rather than being distributed to shareholders or owners. Non-profit organizations can and often do generate revenue that exceeds their expenses, in other words... profit. The key distinction is how that surplus is used. In a non-profit:

- Profits are reinvested to support the organization's goals and activities. - There are no owners or shareholders who receive dividend payments. - Any surplus funds are typically used to expand services, improve operations, or build financial reserves for future needs.

Non-profits still need to be financially secure.

$7500/mo in donations is also already really, really good. It seems like they could easily turn this into a viable business with some small changes.

It's free because all the competition is free and switching costs are zero. They claim that trying to upsell people causes them to quit.

Surprised that a Pluralsight type company wouldn't acquire.

Crazy idea: charge $1

Charing $1 is harder than it sounds. Many commercial transactions have a minimum "service fee" that would eat into that $1.

But $12/yr. is probably doable.

exercise-monetize your users! even it’s a pay-what-you-want with a minimum of say $10 or $20usd per year.

your service is loved.

Maybe pull an OpenAI and go for profit?

Sorry for being upfront honest and transparent:

If you have that amount of users and you cant even pay 3(!) people, then your business model is massivly fci*g wrongly configured, Jesus!

You have that much traffic and you are unable to monetiz - i guess, this is because the "founder" is also the lead developer and the product genius.

You need some business guys to tune this!

Jesus, if this product fails, it will be a great example in whatever Stanford Business 101 course :-))

> We've tried a lot of things to change that. We've spoken to hundreds of funders and companies,

The goal isn’t to get investors! It is to get money from customers. A lot of startups seem to get this wrong.

This is apparently a nonprofit, not a startup. They weren't looking for investors, they were looking for donors.

Ah yes you’re right. Thanks.

Tried to donate in crypto, failed. Bad UX. May I have just address and chain to send to?

Can you integrate commercially with a jobs board? Or create a jobs website?

Founder mode can raise 1B with plain html page, but himself and investors always shy from tossing few mil at useful content sites it gonna later be scraping. It's not sad, it's really ... disgusting.

sad truth of a no profit model

>Rust does not use a garbage collector, using advanced static analysis to provide deterministic drops instead.

Huh?

capitalism doesn't really play well with digital assets, services, nor anything digital really.

we need a lot more Byung-Chul-Han somehow

Just donated a small amount. I think this is a nice service, and can think of a couple of friends who struggle with programming (professionally!) but are too embarrassed to ask and progress, and could make use of this service.

I hear the pain-points about the donation-driven model not rising to the task, but I do wonder how prominent that particular call to action is on the site. On one hand you don't want to bombard your users with annoying jimmy-wales-style boo-hoo donation banners. On the other hand you don't want your users to be completely oblivious to the fact that their favourite service is about to be shut down if everybody thinks that "it's ok, they probably have enough money / I wouldn't make a difference anyway".

Have you considered having a discreet yet prominent progress bar on the website, showing your monthly donation needs/goals? This could act as a good prompt for users to consider donating. One suggestion would be to accompany this by an information button, which when clicked takes you to a page explaining the different costs that need to be covered, where your money goes when you donate, and a detailed breakdown of what is currently still required / missing. Optics-wise, I would also try to find a way to differentiate between "large" donors and donation "trickles". E.g. you could color code on that progress bar different "tiers" of donation (or perhaps display this separately as a pie-chart in the breakdown information page). The point of this is to show to people who might be tempted to donate "small" amounts but are sitting on the fence because they don't think it'll make a difference, that it "can" actually make a cumulative difference, and push that kind of user past that threshold and donate "something".

Alternatively / additionally (you may do this already), if and when you see that a user has already extracted value from your service, such as, e.g., when they have submitted their 100th exercise, you could show a "congratulations" banner that says that you hope they're finding the service useful, and politely asks them if they could consider donating (oftentimes worded as "buy us a beer or two"). If they decide to donate, then after their donation, you could also have a tickbox at the bottom of this banner to ask in a respectful manner (i.e. not a dark-pattern-worded / confirmshaming one) if they would like to be reminded again after some time / number of exercises, or if they simply prefer to make a one-off donation and that's fine.

Finally, depending on your resources, if people "do" buy you a "beer", it would also be a nice touch to thank them for it, in an as personalised a manner as possible. Better yet, you could give them some sort of 'reward' / 'recognition' on the site; e.g. I think something silly such as revealing a hidden feature that allows a user to then tick a box to add a silly 'holding a beer' overlay to their avatar icon (in a similar way to how stackoverflow gives 'hats' close to christmas time), would actually go a long way to motivate people, after seeing these beer mugs on other people's avatars. And it's "silly" enough that people might donate to get this just for the fun and silliness of it.

The donation model is an odd one. Obviously, not everyone donates, but I think given some prompting and lack of friction in doing so, there's a significant number of people who would be inclined to donate out of a sense of gratitude, but who don't do so if they haven't been given a reasonable opportunity / visibility to do so, or some sort of confirmation that their donation doesn't just go to greedy shareholders on top of some sort of pre-existing dodgy revenue stream, but makes an actual difference to real people running the service.

The title has a typo - it's Exercism, not Exercise.

Thanks! Fixed now.

Submitted title was "Exercise has 2M users but no money in the bank". Solution is not to fix the typo but to follow the site guidelines: "Please use the original title, unless it is misleading or linkbait". It's not necessary to add the company name since the domain is displayed next to the title.

At some point they messed up their login with google captcha stuff. From the moment I discovered this new state of affairs, I have avoided exercism. A shame, that they ruined it.

From the article: > “I've lost faith in the nonprofit business model”.

There’s your problem right there. Investors want profit. If you want investors then you have to convince them you’ll deliver ROI.

Altruism is not a business model. There is nothing wrong with altruism but it is not compatible (except in short term or insignificant ways) with most of the business world, which is profit driven.

I think it's fair to say we need a new model, just saying but all of that is way bigger than me or you or Exercism.

Non profit doesn't mean you survive on donations. Non profit can still charge money to keep the lights on. If you have 2M users, charging $0.50/year for use can still get you $1M to run your business.

Edit: I think the comment was taken quite literally. The point I was trying to make was that the user base exists, monetization for profit is definitely hard, but monetization for keeping the lights on can work. You could try a flat fee for all users, you could do seasonal exclusive contents to support the platform, you could try selling merch, you could literally put a banner annually like Wikipedia does etc. The point being, if you have 2M users and you provide enough value, there are ways to make it work. Donations unfortunately don’t work, because it’s not top of mind for people and people need a little pushing.

this isn't at all realistic; a tiny fraction of free users are going to be willing to pay for your product.

Updated the comment with clarifications.

As a side point I was curious how low you can charge given processing fees, and you'll have to charge $1 to get $0.50 cents roughly (checked Stripe/Paypal for my currency CAD).

Updated the comment with clarification

He wrote:

“I think it's fair to say that at this stage I've lost faith in the nonprofit business model working in a way that allows Exercism to reach any of its potential.”

Removing the context that makes it clear that a quote is only applying to a very limited situation is a bit dishonest.

Reply on news.ycombinator.com