> Why not just increase the price of Apollo?

> One option many have suggested is to simply increase the price of Apollo to offset costs. The issue here is that Apollo has approximately 50,000 yearly subscribers at the moment. On average they paid $10/year many months ago, a price I chose based on operating costs I had at the time (server fees, icon design, having a part-time server engineer). Those users are owed service as they already prepaid for a year, but starting July 1st will (in the best case scenario) cost an additional $1/month each in Reddit fees. That's $50,000 in sudden monthly fee that will start incurring in 30 days.

> So you see, even if I increase the price for new subscribers, I still have those many users to contend with. If I wait until their subscription expires, slowly month after month there will be less of them. First month $50,000, second month maybe $45,000, then $40,000, etc. until everything has expired, amounting to hundreds of thousands of dollars. It would be cheaper to simply refund users.

> I hope you can recognize how that's an enormous amount of money to suddenly start incurring with 30 days notice. Even if I added 12,000 new subscribers at $5/month (an enormous feat given the short notice), after Apple's fees that would just be enough to break even.

> Going from a free API for 8 years to suddenly incurring massive costs is not something I can feasibly make work with only 30 days. That's a lot of users to migrate, plans to create, things to test, and to get through app review, and it's just not economically feasible. It's much cheaper for me to simply shut down.

We can debate if the API fees are reasonable or not - but at the end of the day, Apollo chose a model that doesn't work unless Reddit continued to favor them and apps like them. Foolish, is one word that comes to mind.

Given the popularity of Apollo, and the public outcry over the news it will be shutting down - I see zero reason Apollo couldn't switch to a monthly billing model - even if it requires refunding old subscriptions which they are already going to do.

This is a self-made disaster for Apollo, a failure to be forward thinking and control risks.

The founder started Apollo as a university project - but somehow forgot to become a real business along the way it seems.

Nobody really cares about your opinion on him or a monthly subscription; he explicitly said it's not going to happen and at the end of the day that's his decision to make, not yours.

He doesn't even need to give a reason. It's his project to manage.

It's also clear you haven't read the article, because he explicitly calls out a bunch of criticisms you have of him directly.

> Isn't this your fault for building a service reliant on someone else?

> To a certain extent, yes. However, I was assured this year by Reddit not even that long ago that no changes were planned to be made to the API Apollo uses, and I've made decisions about how to monetize my business based on what Reddit has said.

> > January 26, 2023 Reddit: "So I would expect no change, certainly not in the short to medium term. And we're talking like order of years."

> Another portion of the call:

> > January 26, 2023 Reddit: "There's not gonna be any change on it. There's no plans to, there's no plans to touch it right now in 2023. Me: "Fair enough." Reddit: "And if we do touch it, we're going to be improving it in some way."

> Your initial post in April sounded quite optimistic. Are you dumb?

> In hindsight, kinda yeah. Many of the other developers and folks I talked to were much less optimistic than I was, but I legitimately had great interactions with Reddit for many years prior to last week (they were kind, communicative, gave me heads up of changes), so when they said they were aiming to have pricing that would be fair and based in reality, I honestly believed them. That was foolish of me in hindsight, and maybe could have had a different outcome if I was more aggressive in the beginning. Sorry. /canadian

> (And to be clear, they did indeed say this. They used the word "substantive" and I wanted to make sure we had the same definition of something "having a firm basis in reality and therefore important, meaningful, or considerable")

> > Reddit: "That's exactly right. And I think, thankfully, the word is exactly the right one. It's going to have a firm basis in reality. I also just looked it up. We're going to try to be as transparent as we can."

Reading this with 15 years of corporate experience the developer was at best naive. In corporate speak Reddit is completely consistent in their actions and words. It's a crappy situation and I'm sure the developer is a great person and I agree Reddit did them dirty but also that's how these things work. You don't take dependencies on third parties without a lawyer and a contract.

> There's not gonna be any change on it.

Nobody can make this promise, those are just words to make you feel good.

> There's no plans to, there's no plans to touch it right now in 2023.

Plans can be made quickly. Action can be taken without a plan. What is the guarantee on lead time?

> And if we do touch it, we're going to be improving it in some way.

Define improvement. Improved for who?

> It's going to have a firm basis in reality.

I have no doubt that Reddit based the API pricing on them making money on it. We can debate if they got it right.

> We're going to try to be as transparent as we can.

Try is a weasel word, this sentence is meaningless. Zero transparency can be provided and still meet the standard of being "as transparent as possible". "Try" here even gives them the opportunity to be less transparent than possible. The Glomar defense ("We can neither confirm nor deny") is "as transparent as possible" and actually meets a higher standard than Reddit promised here because the CIA didn't just "try", they successfully provided the most possible transparency (almost none).

> I'm not really sure what benefit you get out of harping on a developer you will likely never meet or interact with in a third party forum, but power to you.

Y'all need to find a better hobby. As per his own words, he also clearly realizes that in hindsight he should have been more pessimistic. But that's all moot now. The past is in the past. Pointing it out is not going to do anything.

Reddit could simply treat them reasonably and things would be fine. There's no need for favoritism, they just need to stop being actively harmful. And part of that is the fees (they're not reasonable).

It's like, I know this funded company that's doing a lot of work using intel SGX, if intel kill it, about 80 people lose the jobs and several million in VC goes up in smoke. It's insane to me that people are building businesses that can be killed by 3rd parties that they have no hope of influencing, and have no contracts with.

Another chapter of the internet drama concludes, I suppose. I wish them all the best and I'll be curious to see if reddit survives.

How is this different than any other business running (part of) their operations at a large cloud provider? Or a business having to renew their contract with the power company?

You pay for your cloud provider...

First being Reddit wouldn't have suddenly needed to stop supporting 3rd party apps for free - apps that only cost Reddit money and generate no income. If the API made Reddit money, instead of costing them money, there would be no need for a sudden change. The "suddenness" of this change is due, in no small part, to the cost of supporting basically free-loading 3rd party apps that consume resources but offer nothing in exchange.

Secondly, Apollo would have formed a business model that supported paying for API access, be it monthly subscriptions, freemium, etc. The current model of pay-once-use-forever is simply unsustainable (on an obvious level) - and the $12.99 annual subscription equally so. The Apollo model, as it was, required free API access.

Even if Apollo had been paying for API access all along, and Reddit decided to suddenly hike the prices - Apollo would have been in a better position to raise their own prices accordingly, and would have had a userbase more accepting of paying.

Paying for API access also compels the business to be more efficient in their calls. As it was, there was little-to-no incentive to operate large content caching on your own servers/services, etc. I have seen, but do not know their credibility, that Apollo was not very efficient in it's API calls and essentially hammered the API far more than was necessary. If you're not paying for it, why would you bother designing a more efficient system?

The API fees were inevitable. More are coming - be sure of that, as corporations tighten their figurative belts and look for ways to stop bleeding money.

And the "hammering" is addressed and strongly countered in the article.

Maybe it's necessary, but that seems awfully tight timing. I have not reviewed any of the rest. My understanding is they tried to emulate push notifications...

$12.99 per year was priced with majority of the compute burden foisted upon Reddit. Apollo had servers too, but apparently took a lot of steps to minimize what they did in order to keep subscription prices super low.

$12.99 doesn't pay for an the free users that might not be hammering the API quite on the same level as the paid users, but hammered it non the less.

If the polling was the bulk of the traffic they could just remove that feature.

And there's no way the compute burden is an entire dollar to do 4000 requests.

Specifically, the $12.99/year model is floating a ton of free users, in addition to paying for server resource usage. None of that $12.99 currently was allocated to paying Reddit, and we know Apollo calculated an approximate cost of $2.50 per user per month with the new API fees.

This $2.50 fee seems to align costs with what we can reasonably expect Reddit to earn per user on their platform. Reddit prices Premium membership at $5.99 monthly, which among other benefits removes all ads. $5.99 likely indicates a $2-3 profit when all ads are removed but user engagement remains constant.

The typical mildly engaged Reddit users probably easily spends an hour a day looking at ads via the app - so while the math may be fuzzy, it seems like Reddit possibly based their API pricing off something like this.

At a minimum, that is $2.50 x 12 = $30 of cost annually per user. This means all users need to pay Apollo $30 a year to break even on just Reddit fees, or some subset of users needs to pay a lot more than $30 a year to float a bunch of free users. Apollo has other expenses too (labor, servers, etc.).

Even if the API fees were reduced 80% down to $0.50 per user per month, that's still $6 annually per user - and Apollo has a lot of free users.

All of this is to say, the $12.99/year membership for Apollo was never going to work with any API fees.

I expect that the reddit premium price is far above the ad revenue from a mildly engaged user.

These things are different though, we must admit. Imgur used to be much bigger, but today hardly pulls the same audience volume that Reddit or Twitter do, for instance. I don't think Imgur could get away with API pricing like Twitter is, and Reddit will soon be.

> I expect that the reddit premium price is far above the ad revenue from a mildly engaged user.

Given the number of ads you see while browsing via the official app, I would peg it around $2 per user per month (which mostly aligns with my previous statements and estimates, as well as with API pricing), for someone that uses Reddit for 1-2 hours daily. I suspect, without evidence and can be wrong, that 1-2 hours daily is typical for a mildly engaged Reddit user.

Apollo doesn't need to drop free users - they need to get revenue from them. If not directly, then via ads. Although that seems antithetical to what Apollo was trying to create... but reality has come knocking.

He probably could, but not in 30 days.

> but somehow forgot to become a real business along the way it seems.

Or chose not to.

Reddit will begin charging for access to its API (nytimes.com) 303 points by alexrustic 51 days ago | 339 comments --- https://news.ycombinator.com/item?id=35617763

https://old.reddit.com/r/apolloapp/comments/12ram0f/had_a_fe... was posted April 19th.

https://www.redditinc.com/blog/2023apiupdates

> ANNOUNCEMENTS Staff • April 18, 2023

> ...

> To ensure developers have the tools and information they need to continue to use Reddit safely, protect our users’ privacy and security, and adhere to local regulations, we’re making updates to the ways some can access the Reddit Data API:

> We are introducing a new premium access point for third parties who require additional capabilities, higher usage limits, and broader usage rights.

The prices themselves were announced May 30th. I guess if you’re feeling generous, that would be 32 days notice.

He had notice there would be a change and was explicitly told:

> The information they did provide however was: we will be moving to a paid API as it's not tenable for Reddit to pay for third-party apps indefinitely (understandable, agreed), so they're looking to do equitable pricing based in reality. They mentioned that they were not looking to be like Twitter, which has API pricing so high it was publicly ridiculed.

There is absolutely no actionable information there, and everything they said indicated that it wouldn’t be an unreasonable change.

30 days is ample notice to Apollo's users.

Apollo has had since April to figure out how to make a monthly subscription work, on a technical level. Now... having done nothing smart in the meantime, are left with very little time to make the changes. That is 100% on Apollo.

You shift the blame like there’s no tomorrow. At this point you either work for them, or are getting paid exorbitant amounts of money to defend them. That’s the only reasonable explanation for why you’d be pushing the blame so hard.

Even if he had a full system set up, 32 days is an unreasonably short amount of time to make any sort of material change to your terms, let alone raising the cost exorbitantly.

Hilariously, your first comment wishes people would pledge to pay for Apollo. How we got from there to… this bullshit is beyond me. At least it only took a few hours to show your true colors.

Your "32 days" timeline you keep going on about is false no matter what way you look at it.

Apollo has known since April API pricing was coming, and there was no way their existing "pay once, use it forever" or even the $12.99 per year model was going to be sustainable - particularly given how many free users they float monthly.

Not knowing the exact fee in advance is completely irrelevant. The infrastructure needed to be built to support a monthly subscription model, freemium models, or whatever they needed to do to pivot and remain in business.

100% of the reason people use Apollo is because of the data and community Reddit has built. Apollo basically leached off that data, and made a profit while doing so. Any sound business operator would understand and plan for risks that endanger their business. Apollo failed to take action in a timely manner, and failed to mitigate risks to it's business.

Now they're being asked to pay for the API access that makes Apollo's business possible - and even if the API fees were 1/4 of the proposed amount, Apollo was going to need to change their model. That is the point you, and many others seem to be missing - Apollo was not going to survive paying anything for API access - Apollo had no plan.

I cannot make it any more clear - Apollo failed to plan for, and mitigate risks and failed to take corrective actions in time to save their business.

Apollo was run like a hobby side-project - not a business.

Not everything has to be run like a business, you understand that right? You’re acting like this is a big when it’s a feature.

This is a project run by a single person. And has been, since its creation. On purpose. You clearly don’t like that for whatever idiotic reason and your entire premise hinges on the fact that you think every single project in the world needs to be run like a perfect megacorp. Guess what? The world doesn’t work like that.

The rest of your BS has already been debunked. Multiple times. By multiple people. You restating it continues to be just that, regardless how many times you insist on it.

If you want to work for Reddit so badly, try applying for a job. I’m sure they’ll appreciate your obvious shilling for no discernible reason.

Sure, that is reasonable. But then the victim card cannot be played when the business fails to act accordingly, no?

> The rest of your BS has already been debunked

Besides your insistence on alternative facts, I do not see anything "debunking" anything I have written. Again, just because you don't agree does not make it not true.

> If you want to work for Reddit so badly, try applying for a job.

Generally, you will be better off attacking someone's argument instead of their motivations.

What business?

It’s not a fucking business. Stop insisting it is. We literally just went over this.

> Besides your insistence on alternative facts, I do not see anything "debunking" anything I have written. Again, just because you don't agree does not make it not true.

The fact that half of your comments have been flagged and downvoted to death, and that multiple people have pointed out your straight up lies should tell you plenty about who is dealing in “alternative facts”.

> Generally, you will be better off attacking someone's argument instead of their motivations.

Yeah that hasn’t seemed to work with you for the past half day, so I don’t understand why you would expect people to keep trying when you clearly don’t bother with things like “taking things to heart”, “having a discussion in good faith”… or really “reading”.

Next time, stick to actual facts and don’t pretend everything is a business that needs to be maximally efficient.

Apollo was/is a business, handles hundreds of thousands of dollars annually, and has a huge customer base. We cannot make up definitions and then insist others agree.

> We literally just went over this.

You mean, you did. Sorry to not play into your fantasy universe - out here in the real world, shouting something louder doesn't make it true.

> The fact that half of your comments have been flagged and downvoted to death

You make the mistake of placing stock in whatever button people click on a website - usually based on emotion, but not always. Either way, Internet Points are not real...

You may not like the facts - but that's why they are facts. They do not change even if you hate them. Calm down, re-read the original posts, and then we can discuss things more civilly.

https://apolloapp.io/pro-ultra/

Apparently Apollo has paid subscriptions as a way to finance it.

He would have to turn off the functionality if there is no subscription in order to avoid spending too much for API calls.

Currently this subscription is described as:

> second is Apollo Ultra, the highest tier, which includes everything in Apollo Pro plus additional features. Apollo Ultra is a subscription offering of $1.49 USD per month (or $12.99/year, or a lifetime unlock option is offered in the app) and is a subscription due to options within it having ongoing monthly costs to me, the developer.

The approach would be to do a subscription price change for Apollo Ultra from $1.49/month to $10/month (or whatever the math works out to be - I'm using $10 as an example).

About subscription price changes - https://support.apple.com/en-us/HT213252

> Developers who offer subscriptions can increase the price of a subscription without interrupting service only under certain specific conditions. If the increase does not exceed approximately USD $5 and 50% of the subscription price, or USD $50 and 50% for annual subscriptions, and where permitted by law, developers may change the price without interrupting service. Developers may do this no more than once per year.

Everyone who is subscribed would get an alert and no payment would go through.

However, the code is already there to work as a subscription app.

---

I would also draw attention to https://www.reddit.com/r/redditdev/comments/141mjij/lets_tal...

> My last thing I wanted to address, and I might be burying the lede a bit here, is some of misleading, or downright inaccurate and untruthful claims that the admins have made in regards to these changes..

> > Apollo could reduce their cost by 3.5x if they were as efficient as these other 3P apps.

> So I have not dug into Apollo specifically as I didn't have an iOS rooted device handy. BUT, my guess as to the "increased calls" is due to them more frequently checking if a user has messages, and/or less caching of comment sections and more re-pulling them for the latest on navigation. Could Apollo not check for messages as frequently? Sure.. Reddit is Fun used to check for messages on any refresh it seems, and they sometime somewhat recently seem to have changed that and for game day threads which I frequently use it for, I often miss responses to my comments for a very long time because it seems to only do it now every so often.

> ...

> Edit: it was brought to my attention that Apollo does push notifications for messages even when you aren't using the app. This is almost certainly the main discrepancy between it and other apps API usage. And it could have been a back end change then related to the polling for those notifications that caused a reduction in API calls

Some of the architecture for that is described https://www.reddit.com/r/apolloapp/comments/9l3ema/apollo_13...

> For some quick math, Apollo has well over 100K active users. The server polls Reddit approximately every 6 seconds, so that's 10 requests per minute per user, or 600 requests per hour per user (assuming they only have one account and one device). At 100,000+ users, that's in the realm of 60 million requests per hour that my server would have to handle, not to mention parsing the results, coordinating tokens, etc. I really can't do that for nothing, so the plan was to offer push notifications with a small fee associated to cover these ongoing server costs.

> I understand the logic in not charging for basic system features such as camera usage, but push notifications require a server in order to function, and servers aren't free (in fact they get costly quick). I also offer a completely free system that does not use a server so those who don't want to have to pay can have their device function as the server and use local notifications (which are slightly delayed as it uses Background Fetch and using the device uses more battery), but remote notifications necessitate a server.

As it is designed, Apollo is very heavy on active polling for notifications.

This is described by Christian here - https://youtu.be/Ypwgu1BpaO0?t=1729 that it is polling reddit once every 10 seconds for each user of Apollo constantly.

Spite is how humans enforce social norms. In this case, they lied to him, slandered him with easily falsifiable things they even admitted he didn't say.

The relationship is broken.

Shutting down a popular business because things got temporarily inconvenient is immature at best.

HN, alone, is filled with people willing to pay $5 a month. He doesn't need to find new subscribers, he already has them! He's just not asking them for what he needs to continue operating.

That's entirely on Apollo...

Monetisation of social networks only works via ads/tracking.

You're just being an unconstructive critic.

And you have no evidence it's temporary.

Right - and in 10 years everyone will still be using Reddit, Apollo will just be a distant memory, and nobody will give any thought to the API pricing model.

Just reality...

The 30 day window was not enough time to rectify that and would cost him 50k in the first month to cover the diff. The author suggested he needed at least 3 months to implement changes and switch at least some portion of yearly subscribers to a higher price.

If they came out with a reasonable set of conditions to do API access, people would be a lot less upset. But they didn't. And those API fees are guaranteed to go only up, up, and further up.

Time to get out of reddit when the gettin's good.