It's an interesting idea, and nice to see, but honestly I'm not sure if the convenience of it all working 'out of the box' justifies the cost. My company, for example, set up almost all the functionality they offer in about an hour of Python scripting (using Tropo.)
Eventually, we spent about an extra hour creating an actual support rotator that allows you to toggle your number on/off when you don't want to deal with answering phone calls.
It was entirely free to set up and develop. There's no pressure to transition to a paying account, either. When you do decide to transition, it's much cheaper than any of the TellFi plans, any way you look at it.
Additionally, Tropo has great support - I've wandered into their IRC at 4:00 AM and gotten nearly instant feedback, every time.
So, basically, if you're decently technically proficient, I don't see the value add of TellFi.
You could use the same argument for doing everything yourself though.
"Tropo is nice, but with a few days of programming Asterisk, I got all the functionality you have out of Tropo and I'm not tied down to their pricing plans".
If you hosted your own VoIP pbx, you could just negotiate your own SIP trunking service for much cheaper.
I'm not actually recommending this route, as Asterisk sucks and Tropo is pretty nicely done, but I thought I'd give you another perspective.
Asterisk doesn't suck! I setup a pbx for our own small biz many years ago - I think it was asterisk@home edition or something way back then. I upgraded a couple years into it and that thing ran, without a problem, for the duration of our business - 6 years. The only time that box went down was when we had power outages.
The business was phone intensive, handling hundreds of inbound and outbound calls a day. I love asterisk. Sure, the config files aren't the best to work with, but it wasn't that hard. I put it up on an old P4 machine and just let it run. Amazing really. The FreePBX app (that I upgraded to) made it really easy to do everything a PBX should - and this was YEARS ago.
I then worked with asterisk again a few years ago when doing some consulting for a call center. This time, the front end was vicidial. Again, asterisk was rock solid.
Not to be defensive - but it just doesn't suck. It was way ahead of its time in my opinion. It WORKED! I am reasonably confident it is the underlying infrastructure that twilio uses too.
Ok, to be honest, Asterisk doesn't suck. It's used by thousands of companies and not all of them are miserable with it. However, hundreds of inbound calls a day is not what I call intensive. Hundreds a minute and suddenly Asterisk begs for mercy. It doesn't thread to multiple CPUs. Hell, the architecture was so awful one of the main programmers quit and wrote FREESwitch.
Honestly, it probably performs worse on computers these days as GHZ has been going down to make room for multiple cores. For a small business with plenty of time, not too many calls, and not a ton of money, it's not a terrible solution.
For a small business, 100's of calls a day is above average. But, to your point, the call center I implemented it at had a configuration that did many thousands of calls a day going OUT of it. Perhaps inbound would be a problem (I honestly don't know), but we had 4 pri's hooked up to it and tapped them out 8 hours a day. For anyone who doesn't know the terminology, this is over 90 simultaneous calls. This involved much more than just placing calls - all the queues, auto dials and answering machine detection was going on at the same time. The server was a proper server - but I believe less than 2k (this was 2 years ago). I should note that Vicidial did have some mods to the core asterisk code to account for some issues - but these are things the average small business isn't going to encounter.
Anyway, I'm not trying to downplay the real announcement here - I think it's a very valuable type of service and definitely something a fair number of businesses need. I just had to defend Asterisk! :-)
While I agree "100's" of calls is not intensive at all, asterisk has it's place just like every thing else. It's all about how you set it up. Our asterisk servers rarely crash, and handle more then a 100 or 200 calls per minute with ease.
Well, I guess it comes down to personality and level of what you're willing to do yourself to save some money/save yourself some time.
In my case, some simple scripting provided all the functionality this app provides. It probably cost an extra ~20 minutes of my time over if I had gone with TellFi. I chose python, since I'm most comfortable it, but I could have done it with a variety of languages. Their support is so good, and the scripting is so simple, that it really only requires a minimal understanding to get it all working.
I obviously could have gone with Asterisk, and saved even more money. It would be "a few days" of work, however, not about an hour. I'm just saying that, to me, it's not worth that much work to save such a small amount of money/month to start that far back in the chain, whereas it is worth an hour of my time to save some money and potentially allow myself more control/functionality in the future.
I'm not saying that TellFi is useless—I'm just saying that to me, the value proposition over something like Tropo or Twilio is not clear.
Many companies that require simple yet flexible and powerful telephony may not have any coding skills. TellFi is offering an idiot-proof solution. Thats invaluable.
Even if you have amazing coding power on your team (or personally), it comes down to opportunity cost. As a coder, it might be pretty cool to feel that you hacked together your own phone system, but then if you want to change the routing system or welcome message or something, do you have to hack some more? And when does that end?
Maybe that time can be better spend elsewhere (i.e. improving on something that only you can do - maybe something that improves your product).
I personally think TellFi looks really cool, and the built in analytics is super cool too. If I lived in the states (they only provide US and Canadian numbers) I'd definitely give it a whirl.
Sounds a lot like what Twilio is offering with their OpenVBX. I am not sure how much statistics tracking is available with Twilio but you can host the service (I don't recall if they provide some hosting plan).
Additionally Twilio OpenVBX is opensource.
Eventually, we spent about an extra hour creating an actual support rotator that allows you to toggle your number on/off when you don't want to deal with answering phone calls.
It was entirely free to set up and develop. There's no pressure to transition to a paying account, either. When you do decide to transition, it's much cheaper than any of the TellFi plans, any way you look at it.
Additionally, Tropo has great support - I've wandered into their IRC at 4:00 AM and gotten nearly instant feedback, every time.
So, basically, if you're decently technically proficient, I don't see the value add of TellFi.