Has VoIP succeeded?

Luca Filigheddu provides 10 reasons why VoIP has succeeded and then Aswath essentially says VoIP has failed.

I tend to side with Aswath, but in fact they are both right. It depends on how you define “succeed.” In many areas, VoIP has made a great deal of progress since, say 2000, when nobody really heard of it, and only ubber-geeks were really using it.

Luca cites some of the key reasons for the progress VoIP has made. However, it’s way too early to declare “success” (or “mission accomplished”). On the consumer side, there are still really only two major VoIP themes that people know about: Vonage-style VoIP and Skype-style VoIP. There are lesser known products/projects going in other directions, like PhoneGnome for instance, but they are simply still too low on the radar to factor yet. One could perhaps argue that so-called “web-activated telephony” is a third category, but I’m not sure it counts yet either.

I wrote that VoIP was boring and floundering two years ago in two posts here and here and in fact, not much has changed since.

One thing I said back then I think still rings very true today:

Among the impact players in today’s VoIP, Skype is the clear winner in terms of taking a new approach, being bold. And it turned out to be worth a couple of billion dollars to do that. I’d argue however, that all the existing major players defining VoIP circa 2007 are holding VoIP progress back, rather than moving us forward. They are the Compuserves and Genies of VoIP.

The Vonage-like players are holding VoIP progress back by being phone companies, by taking a 130 year old frame of thinking and cramming it on top of the Internet. They are phone companies that use the Internet for their “mud” to bury their pipes into. Above that “mud” layer, they have done nothing significant to change telephony or, more importantly, improve upon human communications. They take a perfectly good operating model, the Internet, and a lot of powerful VoIP-enabled capabilities (like SIP) and they dumb-them down to turn them into a POTS phone of 1975. And they spend $2B telling people this is the right way to do things, so there are now hundreds, if not thousands, of copycats, selling the industry and the public that same story. Now, it might as well be fact. In the mean time, what they’re selling is snake oil. It’s not like the Internet at all. There is no interoperability – other than the old PSTN. Contrast Internet email. With email, who is the post office? Nobody. With VoIP, why do we need a “phone company” – the real answer is we don’t. But we’re stuck with that thinking for who knows how many years, or decades now because a lot of money has been spent selling it. It’s kind of like how, in the US, people think cell phones are the carrier and vice versa – people don’t even think about it; it’s just a given that you get your cell phone from a wireless provider and that that’s just what you do. There are no network effects in a Vonage model. I could go on and on, but the point is the Vonage-like VoIP model takes us nowhere. The end result is I send my check to different address. I’m no closer to a genuine “Internet” Voice capability.

Skype is holding things back because it is not interoperable. Luca names SIP, the Internet standard for VoIP, as one of the reasons VoIP has succeeded and that’s certainly true at large. In fact, without SIP, Skype might have had some serious trouble developing their SkypeOut and SkypeIn services. But Skype doesn’t expose SIP to end users and isn’t interoperable with SIP. Skype users can’t place calls to or receive calls from SIP-based services.

So neither of the major VoIP trends, the “successes”, in VoIP to date take us any closer to an Internet-like direction for voice. They take us closer to the days before the Internet when we used closed, proprietary “online services”. The VoIP services that have had the most success today barely deserve the “IP” in their name. They take a lot from the Internet, but they don’t give anything back. GizmoProject is probably the closest thing to a step in the right direction in VoIP to date, but even that service is less “open” than it could be.

Taking the list Jeff Pulver recently mentioned that he says comes from his VON 2003 notes, here’s what he thought the industry needed most, back in 2003:

– A royalty free, single variable bit rate adaptive codec from dialup to broadband.
– Low cost, low end IP edge devices.
– Service Providers to deploy an architecture that distributes service execution.
– Take advantage of the great QoS available today on the Internet’s backbone.
– IPv6 needs to be supported and rolled out.
– The Purple Minutes battle cry has yet to be truly heard.
– Drop the IP in front of IP Communications since we really represent the future of the Communications Industry.

The fact is none of the above have really happened, in any significant ways. You could argue some of the underlying technology is there (say Speex for a codec) but it’s not available in practice – it’s not in any of the hardware one can actually buy. And what’s more, it’s not practical to use (and isn’t part of the standard). As I’ve mentioned in other posts, service providers seem barely able to implement G711 in a way that permits interoperability. If you start throwing in anything else, fuggetaboutit. That’s the reality of the state of this industry today. The same could be said for pretty much Jeff’s entire 2003 list.

So has VoIP succeed? Well, there have been successes, and progress, which we shouldn’t ignore, but it has a long way to go.