I havn't tried your service yet but definitely appreciate you doing this and wishing you success!!
I normally use skype or hangout/gvoice to make calls online, but definitely never found them that easy to use, and skype has dropped my calls before which is annoying.
Cool service! Only drawback is that I wasn't able to find a dial-pad so I couldn't navigate phone trees. Can't complain though as it's a free service that works!
Its built on someone's platform, who they are obviously paying through the nose for. Their FAQ says it costs $5 for 250 minute, which is $0.02/min. Comparatively, I sit somewhere around $0.0028/min and I am not trying to optimize my costs by doing LCR on the few tens of thousands of minutes we push. Could probably halve that if I pulled in more providers.
Uhh, yeah, and I know I'm overpaying too (though its just $20 or so). VOIP is a solved problem in the US if you avoid resellers like Twilio, Nexmo, etc and go directly to either a CLEC or a voice wholesaler.
> Who are you using for your provider?
Endstream (NPA-NXX) & Bulk Solutions (Flat rate @ $0.0045/min), but there are cheaper providers out there if you spend a few hours and set up Freeswitch to do LCR.
We bring in another CLEC for our DIDs (though we have a few dozen with em), SMS/MMS & E911 due to price & convenience, Endstream is still shaping up their offerings in the SMS/MMS arena.
For now we've got Google's reCaptcha set up. We're also going to be implementing a way for recipients of the calls to control what they do and don't receive.
After that we're going to have to watch the behavior patterns and compensate for anything that comes up.
It is one of our biggest concerns but something we believe can be conquered.
Hey all, I wanted to post my prototype here and get HN's feedback before submitting my YC application.
This project came out of the combination of my work on web telephony from the last several years and the troubles that I've had communicating with the family members that I have overseas.
Skype is amazing. But it can't beat a phone call. And if you're trying to deal with a company, a bank, or (god forbid) trying to sort out Visa issues you need to be able to make a call from anywhere.
I realized that its the 21st century and it doesn't make any sense for international calling to be so difficult or expensive. Calling should be free and available to anyone in the world and I wanted to make that happen.
I believe that through sponsors and a Tom's Shoes-style model where those that can pay for extra features subsidize the costs for those that really need the free service, we can make this work. And not only can this be a viable business model but also something that brings a huge amount of social good to the world.
International communication should be easy. And we want get there, even if its one country at a time.
So, what sets you apart from all the other WebRTC providers out there? How are you unique?
It could be offering HD Voice interop with AT&T/VZW/T-Mo, selling user verified CNAM, or anything that adds value you can't get elsewhere.
I'd just avoid trying to become a voice wholesaler, its a very competitive space filled by a bunch of hyper competitive sub-5 person businesses that are super nimble due to their size. Most are pushing a few tens of millions of minutes a month.
> Honestly just ease of use. No fees. No ads. No account required. Nothing to download or install.
So how are you going to stay viable in 3 to 5 years?
> Just go to the website, put in the phone number and call.
Ok, there are a million WebRTC betas out there right now to use for free, even AT&T offered one for the past 2 years up until they killed their WebRTC program in January.
> From what I've experienced with my relatives overseas this is something that's really needed and something they have not been able to find.
Sure, but once again to maintain viability, you need to offer something unique, telecom is a very competitive space, and to scale it you need to build key relationships. Setting up a WebRTC client is awesome, but they are a dime a dozen currently.
So first I have to grow this to some decent amount of usage where it even matters really planning out the long term. :)
But, the current plan is to charge the businesses that want to receive calls from their international customers for tracking and control.
We can allow the businesses to have branded links that auto-dial their number that they can share on their website, email, etc. With these they can also track the source of the caller, do call recording, and route the calls how they want.
For travel businesses, for example, that do business with primarily customers outside their country this is a problem that we can solve for them. Most people would never dial an international number, but they'll click a link and make a free call from their browser.
> So first I have to grow this to some decent amount of usage where it even matters really planning out the long term. :)
Eh, if you avoid Level 3 and Bandwidth.com directly, and go to a wholesaler like Endstream, you won't need any volume. Just throw $20 in your account every few months and you should be good.
> But, the current plan is to charge the businesses that want to receive calls from their international customers for tracking and control.
Mmm, you could totes do this, make sure to use Opus and pitch it as Full HD Audio (better than that tin can HD that carriers are pushing now). Grandstream supports Opus straight to their deskphones by the way, so you may wanna look at them. Their middling quality, but rock bottom price wise.
> branded links that auto-dial their number ... track the source of the caller
How do you plan to get the caller's name?
Also, travel businesses might be a viable model for this, I'd just encourage you to think outside the box, what other areas could use this kind of service?
I'll look into those providers and see what I can set up. If I can get the costs down by an order of magnitude that would obviously be amazing.
As for tracking the source of the caller, I'm not talking about their name but how they came to the website to that then generated the call. Whether thats though organic search, PPC, advertising, some random forum post, etc. It's a lot easier to get exact referral information when the call is a web event rather than an actual phone call.
So I can understand this. You're sort of offering an international 1800 number? Except instead of using a phone number prefix, you're using links and WebRTC? It does sound pretty cool.
Hey, trome, you know a lot about this space judging from your comments. I'm really interested to know who some of the other WebRTC providers you have in mind are. I can't find anything on Google, I'm afraid.
I just googled Call USA from browser, and successfully called myself with no registration or auth from both of the below providers. Both use WebRTC to access your mic & handle signaling and call media. There are a few dozen more providers that have WebRTC demos set up that you can use like the ones below, WebRTC.org used to have a list of businesses that supported WebRTC but I can't seem to find it now.
So I don't want to "bash the competition" but these sites are exactly the problem I ran into when I tried to find a way to do this before building it myself.
On the first one my calls just fail for no reason. And on the second one my call is limited to 5 minutes and it looks like they only allow a certain number of calls per day.
That's the issue. There's a ton of "demos" of this online but nobody that actually takes it seriously or makes any effort to really refine the experience as a core of their product. That's exactly what I want to do.
> On the first one my calls just fail for no reason. And on the second one my call is limited to 5 minutes and it looks like they only allow a certain number of calls per day.
Have you considered that maybe there is a good reason that your competitors have these restrictions? For example, because scammers might abuse your service by pretending to be the IRS?
Oh yeah, definitely. In fact, I believe that dealing with fraud will be the #1 concern for this business aside from getting traction in the first place. Thats why even my prototype is protected using Google's reCaptcha.
However, I also believe that this is something that can be handled while also providing a good experience for real users.
I've run other web telephony businesses for several years now and those get hit by scammers at least once a month, but I've learned how to automatically block them out. I'm sure there's a way to do that here as well.
What browser are you using? Both work for me in Firefox ESR 45.7.0.
Without money to encourage you to spend time on it, how are you going to refine the experience? How are you going to refine the experience? Debugging browser incompatibility issues with non-Evergreen browsers sounds like a nightmare to me, and is part of what holds back WebRTC today.
I normally use skype or hangout/gvoice to make calls online, but definitely never found them that easy to use, and skype has dropped my calls before which is annoying.