There are lots of mistakes Reddit made that shows they aren’t trying.
They could have given more advance notice for the API price increase. This would give apps more time to update their code to use fewer API calls. Many apps are subscription-based, so it would give them more time to update their subscription price.
The price should have been based on Reddit’s actual costs, actual revenue, and actual profits. I.e., if it costs Reddit $0.10 per user per year and their revenue per user is $0.15 per user per year from ads, then the API price should have been $0.15-$0.25 per user per year. The actual pricing shows they made it artificially high to kill the 3rd party apps. (I don’t know what the actual numbers are.)
Even if Reddit really did want to charge $5 per month for API users, the right way to do it is to start from a lower price and increase it 20%-50% per year until they get to their target price.
If a user had Reddit premium, they should have been given extra API call tokens they can give to their 3rd party app.
Typo? This Ars article claims it’s $12k per 50M calls and Christian’s original post has $0.24 per 1000 calls listed, which adds up.
The price they gave was $0.24 for 1,000 API calls. I quickly inputted this in my app, and saw that it was not far off Twitter’s outstandingly high API prices, at $12,000, and with my current usage would cost almost $2 million dollars per month, or over $20 million per year.
I think a smol problem here is that Apollo offered a “Lifetime Subscription” for a flat amount. In order to honor that Christian would be obliged to take money from elsewhere to cover those subscriptions, potentially for decades.
I’m a lifetime subscription holder and I would have, without a doubt, been fine with having my flat amount cover a few months of the new model, then switching over to the per month amount needed to keep Apollo running. But that wasn’t even offered!
He just decided that he’d had enough and pulled the plug.
The bigger problem is that Selig and other developers have an actual conscience and understood that the pricing for the new model would be way too high (in the ballpark of $20-$25 per month per user) to even cover the costs.
Not only that but he would lose around 2m per month in the rollover while ending life time and long term subscriptions bringing up the costs another 250k. Just to change it to a model that very few would pay for.
Instead he decided to cut losses, terminate the subscriptions and not owe 2m per month.
No, he wouldn’t. He’d have to start passing on the cost of the new API pricing + his fair share as dev, to the users of Apollo. Nothing about how the app currently functions would have to change, just the amount users pay to use the app under the new pricing.
He never even offered that as an option. (Not that I’ve seen or been told by him as a subscriber.) He apparently decided that doing so wasn’t worth it and pulled the plug. And to be clear, as the app developer that’s fully his right and I support his decision.
I’m old enough to know that sometimes in life you’ve just “had it,” and it’s time to walk away.
He explained that 30 days was not enough to design and roll out a whole new pricing scheme, and he would have to start paying the increased fee for all his existing lifetime subscribers, in the hopes that enough people would subscribe to the new schemes to offset the cost.
A change like that is not something you undertake lightly.
I think one of the big issues for him was monetary liability. Even if he did pass the API costs on to willing users, he still wasn’t going to be making much -if any- more money.
It’s kinda like somebody offering you $1 to get a $20 bill across the street safely - vs - someone offering you $1 to get $20,000 across the street safely.
If something went wrong with the $20, then no big deal. But if something went wrong with the $20,000, then oh shit. A dollar isn’t worth that headache.
Same thing with Apollo going completely sub based at $5 or $10 a month. If something was slightly off about his accounting or API call guesstimates or anything else, he could easily be on the hook for tens of thousands of dollars that the subscription fees didn’t cover that month or quarter or year.
There are lots of mistakes Reddit made that shows they aren’t trying.
They could have given more advance notice for the API price increase. This would give apps more time to update their code to use fewer API calls. Many apps are subscription-based, so it would give them more time to update their subscription price.
The price should have been based on Reddit’s actual costs, actual revenue, and actual profits. I.e., if it costs Reddit $0.10 per user per year and their revenue per user is $0.15 per user per year from ads, then the API price should have been $0.15-$0.25 per user per year. The actual pricing shows they made it artificially high to kill the 3rd party apps. (I don’t know what the actual numbers are.)
Even if Reddit really did want to charge $5 per month for API users, the right way to do it is to start from a lower price and increase it 20%-50% per year until they get to their target price.
If a user had Reddit premium, they should have been given extra API call tokens they can give to their 3rd party app.
I believe I read $0.24 per 500 API calls somewhere, which is insanely high.
$12 per 50,000,000 calls is the figure Selig mentions.
Oops, I left out the letter ‘k’.
Typo? This Ars article claims it’s $12k per 50M calls and Christian’s original post has $0.24 per 1000 calls listed, which adds up.
I think a smol problem here is that Apollo offered a “Lifetime Subscription” for a flat amount. In order to honor that Christian would be obliged to take money from elsewhere to cover those subscriptions, potentially for decades.
I’m a lifetime subscription holder and I would have, without a doubt, been fine with having my flat amount cover a few months of the new model, then switching over to the per month amount needed to keep Apollo running. But that wasn’t even offered!
He just decided that he’d had enough and pulled the plug.
The bigger problem is that Selig and other developers have an actual conscience and understood that the pricing for the new model would be way too high (in the ballpark of $20-$25 per month per user) to even cover the costs.
Not only that but he would lose around 2m per month in the rollover while ending life time and long term subscriptions bringing up the costs another 250k. Just to change it to a model that very few would pay for.
Instead he decided to cut losses, terminate the subscriptions and not owe 2m per month.
I don’t blame him, he’d have to start over on his project he’s worked on for years. I wouldn’t want to do it either lol
No, he wouldn’t. He’d have to start passing on the cost of the new API pricing + his fair share as dev, to the users of Apollo. Nothing about how the app currently functions would have to change, just the amount users pay to use the app under the new pricing.
He never even offered that as an option. (Not that I’ve seen or been told by him as a subscriber.) He apparently decided that doing so wasn’t worth it and pulled the plug. And to be clear, as the app developer that’s fully his right and I support his decision.
I’m old enough to know that sometimes in life you’ve just “had it,” and it’s time to walk away.
he said some of his busy users would cost $15-20/mo. That’s definitely too much for most people for something like this
He explained that 30 days was not enough to design and roll out a whole new pricing scheme, and he would have to start paying the increased fee for all his existing lifetime subscribers, in the hopes that enough people would subscribe to the new schemes to offset the cost.
A change like that is not something you undertake lightly.
I feel for him too.
I think one of the big issues for him was monetary liability. Even if he did pass the API costs on to willing users, he still wasn’t going to be making much -if any- more money.
It’s kinda like somebody offering you $1 to get a $20 bill across the street safely - vs - someone offering you $1 to get $20,000 across the street safely.
If something went wrong with the $20, then no big deal. But if something went wrong with the $20,000, then oh shit. A dollar isn’t worth that headache.
Same thing with Apollo going completely sub based at $5 or $10 a month. If something was slightly off about his accounting or API call guesstimates or anything else, he could easily be on the hook for tens of thousands of dollars that the subscription fees didn’t cover that month or quarter or year.
I would’t be able to sleep like that.
Very valid point. Thank you!