Comment by V__
2 years ago
Signal had 40 million active users in 2021 [1]. With 14 million in infra cost, that comes to .35 per user/year. Total expenses are about 33 million, so about .825 per user/year. All in all that seems very reasonable.
2 years ago
Signal had 40 million active users in 2021 [1]. With 14 million in infra cost, that comes to .35 per user/year. Total expenses are about 33 million, so about .825 per user/year. All in all that seems very reasonable.
Mastodon org + Mastodon.social also have costs of 0.6 EUR/year, though they have two orders of magnitude less users [1]. This is really what most social media costs. These rates are even payable by many in poorer countries.
[1] https://news.ycombinator.com/item?id=38117385
IIRC WhatsApp used to charge $1 per year
https://venturebeat.com/mobile/whatsapp-subscription/
With how much Mastodon.social tends to fall over when Twitter does something stupid (again), their rates are probably a bit too low for a more robust service like Signal.
Signal also intentionally doesn't store too much data, long term data costs will slowly grow over the years. I imagine for a bigger platform, costs can grow to multiples of the rates for Signal and smaller Mastodon servers.
€10 per year should be more than enough for most users, though, and it should be quite affordable for most countries.
Signal also fell over flat when the whatsapp outage happened a couple years ago. It's just difficult to handle spikes in demand.
Yeah, the issue is more that there is substantial friction in paying any amount of money, especially in poorer countries with no access to e.g. banking or payment cards. I'm sure no one here, and few people even in comparatively poorer countries, would object if Signal/their messenger of choice cost 0.60$ per year to use. The problem is that making the service have a ~1$/yr price tag (as WhatsApp once had) is itself a barrier to a huge portion of the target audience.
In Pakistan at least, sometimes you can donate to charity etc by texting a special number [1]. That subtracts some fixed amount from your prepaid mobile balance (which the vast majority of people use) or adds to your postpaid bill. I imagine its possible for some business to charge customers this way as well.
Then again, instant C2C and C2B digital payments using mobile phones is growing extremely fast in most of the global south.
[1] https://www.app.com.pk/national/pta-introduces-9999-sms-code...
1 reply →
It’s beginning to sound like the 1 EUR/year that at some point WhatsApp wanted to charge and it seemed reasonable to me at the time. Signal is even better and even more so justified.
They used to "require" a subscription of 1$/year but it was not enforced. If you missed the deadline, nothing happened. It was basically the WinRAR model but for an online service.
That may have been an A/B testing of sorts then, because I was booted right away.
1 reply →
This is kind of the number I was looking for -- "Cover your own costs: $1/year. Cover yourself and five other people: $5/year." I feel like something pointing out that the costs are around $1/year on signing up, maybe with a reminder once a year, would get most people self-funding pretty quickly.
Reminds me of ... WhatsApp :D
(Originally WhatsApp charged $1/year.)
And I was SOOO happy when I heard WhatsApp's business model: Finally, I'M THE CUSTOMER! I gladly signed up for the "free year" and started getting other people to sign up for it... only to have it bought by FB, and never charged my $1 yearly fee. :-(
Then I tried to get people to use Telegram, but hey never implemented encryption by default, instead implementing things like chatrooms with millions of people... then I signed up for Signal, but waited to see what would happen -- and they started doing some weird crypto thing. Thankfully that all seems to have not been an issue, so I might actually start recommending Signal.
1 reply →
I'd be happy to pay $1/year for signal, and I'd pay $2/year if it were decoupled from my phone number.
If you pay Signal $1/year, they'll realistically see about 60-70 cents of that – and that's only considering payment processor fees.
Now add the cost of providing support (it's a paid product now!), payment handling on their end (in a privacy-preserving way, which excludes most common payment methods), and top it off with the immense damage to the network effect by excluding all the users that can't or simply don't want to pay $1/year...
Donations seem like the much better option here.
You can also charge for a 10 year minimum and get to a higher retained %
You don't need to provide support, even much more expensive consumer services live without a proper one, so being explicit about the fact that you only pay for infrastructure could suffice
Not sure why payment privacy has to be so strict for everyone
The network effect damage is real, but maybe it could be limited with donations :)
6 replies →
Thanks for over-analyzing my comment. $1/year, $2/year, $5/year, is all insignificant in the wide array of things I pay for. Sure, I'd pay $10/year for Signal as it is today if they really needed me to. And I never said to make payment mandatory. You're just way over analyzing a simple comment.
I'd pay substantially more for Signal if I could bot accounts.
I'd like a signal daemon on all my servers for alerting which could message me via Signal. This is worth a monthly fee to me.
I know people running small businesses who would really like to have a business Signal account: an ability to send Signal messages as a business identity without tying it to some specific phone number. This would be worth a subscription even if they had to get their customers to install Signal.
Signal need to figure out what product they sell that's going to fund the privacy objective: because there's plenty and they're worth having.
If you want one for just personal use; this works well: https://github.com/AsamK/signal-cli
Just sign up with a Twilio number (using voice call) and you can make your own bot.
2 replies →
I'd pay much more than $2 if they offered account identifiers other than phone numbers. Trying to get a burner SIM or DID while still staying anonymous is getting increasingly difficult.
But I think it's pretty clear by now that this is a feature for FVEY IC, not a bug. FFS, they burned development resources on stickers, but abjectly refuse to offer alternative account identifiers. The standard apologist response is, "but phone numbers make adoption easier". Sure, but nobody is asking to replace the identifiers, or even to make them nondefault. We're just asking for the option. It could be hidden behind a developer mode for all I care, but it should be there.
The fact that they abjectly refuse to do it is enough to tell you about what their true motivations likely are.
Agreed, at this point I don't believe the "privacy" aspect of Signal's sales sheet means anything. Most that I know use it primarily because they can have clients on all platforms, including desktop.
> We're just asking for the option
Indeed, the Wire messenger is done like this - it offers phone number, but has an option to not use them and only rely on the usernames (although I think you need to register in the web browser for that)
Based on App Store downloads on both platforms, they are well over 200M at this point.
A lot of people, myself included, have it installed but never use it after they dropped SMS support.
Only a tiny fraction of my contacts use Signal, and most of those are also on Whatsapp, Telegram, Discord, and others.
Signal offers essentially nothing to me.
The sms decision made signal go from THE messaging app on my phone to an app I only use with a very small subset of my contacts. It is infuriating that they didn't allow users to retain that functionality when it costs them nothing, and they could have disabled it by default.
3 replies →
Except real privacy?
17 replies →
My lawyer stopped using signal due to the sms support being dropped. It became too much of a hassle and wasn't worth it.
Many of my family also dropped Signal.
It is now really only used by the hyper-privacy conscious.
I really don't get why people are still using SMS. Is data really that expensive?
WhatsApp in 2013 spent 148 M$/y with 400 MAU, or about 0.375 $/user-year. That's remarkably similar!
https://en.wikipedia.org/w/index.php?title=WhatsApp&oldid=11...
(Small difference is that WhatsApp had a profitability of –93 %.)
Whatsapp got pretty big at 1 eur/year (iOS) and 1 eur for lifetime (Android) here in the netherlands.
I do fear they'll loose most tech un-savvy users because they don't know how to pay (safely).
That doesn't mean they were actually profitable at those rates though. They could have been in growth hacking mode with venture backing.
They were well-known for not doing that, though.
2 replies →
I wonder how many people paid the $5 for WhatsApp back in the day. It gave you nothing but you were able to do it. I think I did.
Whatsapp was asking for $1/year [1]
[1] https://www.theguardian.com/technology/2016/jan/18/whatsapp-...
I've been using WhatsApp when the nominal $1/year fee was still around, but somehow never ended up being actually charged, and I don't know anyone that did.
It's possible that they were only enforcing it in some regions, though.
2 replies →
The price changed a few times but they definitely had a lifetime thing once.
All pricing was entirely optional
Here's one reference to a different price (can't find lifetime except for people complaining that Facebook didn't honor it on original ToS)
https://www.wired.com/2011/11/whatsapp-messenger-app/
I have an old receipt in my Google Pay for whatsapp at a whopping 99 cents :)
I'm paying what works out to about 15 cents per "booking" in my app due to API fees. Maybe more,.. and I'm just now realizing we'll probably be losing money if people used their accounts to their limits. Like 500 bookings would cost me at least $75 but we charge about 50. Anyway $1/year is great
Definitely reasonable but the ultra privacy-conscious/paranoid can't easily donate or pay privately.
Sure, but privacy isn't black or white. A donation to signal does not compromise the content of your messaging.
So what you've leaked is the information that you have an interest in private conversations. This might be a problem in some countries, but I think it's fair to ask folks in affluent countries with working (sorta) democracies to shoulder that burden. I.e. you don't donate if there's elevated threat to your safety, there are enough people who aren't under elevated threat.
There's also the possibility of using a donation mixer like Silent Donor, though I'd evaluate that very carefully. (There's a record of the transfer in, and the mixer needs to keep temporary records for transferring out. There's also the question how you verify the mixer doesn't skim.)
Some donation mixers accept crypto currency, so for maximum paranoia, I suppose crypto->crypto mixer->donation mixer->charity might be workable. Or hand cash to a friend who donates in your stead.
As always, the best path is to set aside paranoia and build a threat model instead to see what the actual risks are.
There's never enough talk like this and I'm not sure why. It's always about the threat model. In this respect I always like to think of it in terms of probability. Probabilities and likelihoods aren't just about capturing randomness like quantum fluctuations or rolling dice, they are fundamentally about capturing uncertainty. Your threat model is your conditions and you can only calculate likelihoods as you don't know everything. There are no guarantees of privacy or security. This is why I always hated the conversations around when Signal was discussing deleting messages and people were saying that it's useless because someone could have saved the message before you deleted them. But this is also standard practice in industry because they understand the probabilistic framework and that there's a good chance that you delete before they save. Framing privacy and security as binary/deterministic options doesn't just do a poor but "good enough approximation" of these but actually leads you to make decisions that would decrease your privacy and security!
It's like brute forcing, we just want something where we'd be surprised if someone could accomplish it within the lifetime of the universe though technically it is possible for them to get it on the very first try if they are very very lucky. Which is an extreme understatement. It's far more likely that you could walk up to a random door, put the wrong key in, have the door's lock fall out of place, and open it to find a bear, a methhead, and a Rabbi sitting around a table drinking tea, playing cards, and the Rabbi has a full house. I'll take my odds on 256 bit encryption.
They take checks by mail. You definitely can do a cashier's check and I'm sure they'd take the "cash in an envelope" method that places like Mullvad do too. Looks like they also support crypto, and that includes Zcash. So I don't think this is a great excuse. The only "can't easily donate" aspect is going to also be tied with the "can't easily get a cashier's check or find an anonymous person to sell me bitcoin for cash" kinda issues, and when you're operating at that level I'm not sure anything is "easy." (but that's not that hard usually)
https://support.signal.org/hc/en-us/articles/360031949872-Do...
How is a check in any way private? Your name is on it.
7 replies →
There are clever ways around that. I use posteo as my mailprovider. They have a system where you can pay anonymously: https://posteo.de/en/site/payment
Signal requires a real phone number to open an account, you are not anonymous to Signal.
I can pop into almost any phone shop around here and walk out with a free SIM card, which I can top up for cash.
Phone numbers can be obtained anonymously in many countries. I have several anonymous Signal accounts, each with their own anonymous phone number.
3 replies →
Very reasonable with only 40 million users?! It's shockingly expensive.