For a long time, I had this belief that the worlds of web and voice would converge, unleashing upon us a whole new class of voice-web mashups. Instead all we got were some marginal ideas and rarely-used widgets. Voice, in particular, remained too difficult for web developers. In the end the two turned out to be awkward roommates, never really comfortable with each other.
Lately, VoIP insiders have started talking about taking a platform approach -– a way to add voice to web applications easily while leaving complex tasks such as peering and billing in the background. Ribbit, a Mountain View, Calif.-based startup, is the latest to join the fray, and it has perhaps the most audacious (and equally risky) strategy for bringing web and voice together.
If you strip away the hype (meaningless blather such as the company’s claim of being Silicon Valley’s first phone company), what they have done is built their own Class 5 softswitch and back-end infrastructure and married it to front-end technologies like Flash and Flex from Adobe Systems (ADBE).
Furthermore, the platform is able to take inputs from different communication tools — XMPP, Skype, Yahoo Messenger, MSN and Flash Media Server –- and make them talk to their “switch.” The platform uses SIP protocol for all voice communication.
Accordingly, Ribbit is offering API (Application Protocol Interface) access to much of our switch today, allowing third party developers to create rich integrated telephony applications without previous knowledge of telephony. Currently, the Ribbit API is optimized for Flash / Flex developers because of the pervasiveness of the technology (Flash is resident on 98% of the world’s computers). This means that Ribbit communication applications written in Flash will run without the need of a client download.
“What we have done is made voice an object that you embed into your workflow (or software),” said Ted Griggs, chief executive officer of the company. “We didn’t want to change how people did things, like communicate via Skype, and wanted to integrate the platform to work with any phone.”
Developers just have to write apps, Griggs said, and worry about the back-end telephony stuff. In exchange, Ribbit will take between 5 percent and 15 percent of the revenues generated by an application.
Briggs showed off a few applications, including one specifically targeting subscribers of Salesforce.com. Built by an independent developer, it was quite impressive. Another application worth checking out is AIRphone, developed by Joe Johnson of Knoware, that acts like an iPhone on your desktop.
The technological approach the company has taken is sound and has merit. It is the strategy and execution part that raises doubts about Ribbit and its future. For instance, I have on good authority that Adobe will be making a big splash with its VoIP plans sometime next spring, and is working furiously to put finishing touches on its offerings. This would make Adobe Ribbit’s biggest competitor. And there are others, like Lypp, who are following a similar strategy. Meanwhile Google, which owns GrandCentral, could easily roll out its own version of a voice-web platform.
The biggest challenge for Ribbit will be building and nurturing a developer ecosystem. It needs patience and a lot of money. Ribbit claims it already has about 600 developers involved in its platform, but will the $13 million the company has raised from venture capitalists be enough? I don’t know.
Nevertheless, Ribbit seems to be rising to the challenge posed by Daniel Berninger, who in a column here wrote:
The death of the telecom business remains a standard prediction, but telephone bills continue to arrive…aside from price, the telecom business remains largely unchanged by VoIP…although improved performance and falling costs usually combine to produce new applications, this does not seem to be the case for VoIP and the voice business. If the infocom sector can move beyond cheap telephone calls, it might finally represent the threat to the status quo imagined by my AT&T colleagues.