Chris Kranky

Recent Posts


All the talk about telephony API’s

Chris KoehnckeChris Koehncke

gearsIn the past weeks there has been increasing noise about cloud based telephony API’s with several big telco types weighing in. Adding wind to the sail has been the buzz surrounding WebRTC and AT&T’s kinda sorta announcement about something they’re doing here but not really (a typical AT&T announcement). Talking to friends in the NYC financial community, they’re asking me about WebRTC and where the big upsides will be.  So I started to search.

If you follow this API stream back to the headwaters, you end up at Florida’s Voxeo who has been providing the tools necessary for all sorts of whacked out telephony applications since 1999. It hasn’t all been cherries and nuts for Voxeo, they’ve raised a ton of cash early on (~$48m)  from top tier VC’s, burnt thru it and then struggled along and raised another $9m in 2009 from local investors. Voxeo also struggled as to whether they wanted to sell telephony development systems (mostly a one time purchase) or be a service provider. Right, wrong or indifferent – no one has acquired Voxeo.

Having said that, Voxeo has probably the best set of telephony tools on the market and having crawled across the floor of broken VoIP glass, they’re highly experienced at what they do and offer. So I like them. Now do I like them financially? The issue Voxeo finds is customer’s willingness to pay. Once any customer gets big enough, they start wondering whether there is a cheaper way to do whatever they’re doing which tends to steer them to building it themselves bastardizing one of the several open source telephony engines. Thus Voxeo is constantly competing with a “we can do this without you” customer attitude  The fact that Voxeo has a huge assortment of tools is meaningless if I have a specific application (e.g. you don’t need a sink wrench if you’re not taking a sink out).

Voxeo’s stuff isn’t for the novice and recognizing this along came another company, Twilio. Twilio had a clear business model, we’re a cloud provider. They kept their interface simple and spent more time trying to appeal to the normal web developer who accidentally wanted to do something with telephony. I’m not going to yap about their telephony API because its damn cheap if you don’t use it and damn expensive if you do. My guess is telephony isn’t a big % of their revenues.

Twilio did solve a huge problem. Namely, how to send an SMS easily. If I’m a programmer sending an SMS is really hard. Sure I could go to Mobile365 (assuming I knew about them) and some dorky sales guy would talk to me and want me to spend $1000 a month and have a 24 month contract and and and. What Twilio did was make it simple. Plunk down your credit card, write to their simple interface and send SMS’s for a low price on a per drink basis. Brilliant.

Both these companies have solved real problems, neither has shown that it’s a huge business, both have shown that it’s a fairly difficult business to get into and fully understand. In fact, I’d say that their each category killers. Good for them.

Now what business AT&T, Telefonica, Deutsche Telecom, Telstra (all have announced some type of API program) have in this business segment I have no clue.  British Telecom purchased the then “hot” telephony API company Ribbit for an astonishing $105m (cash please) and we’ve never heard from Ribbit again (good for the founders sipping a cool one in the Caribbean now I assume).

With all the talk about telephony API’s, I heard scant little about real life deployments. Who of us actually like reaching a machine or having a machine call us. Why do I still have to navigate endless menu’s to reach a hapless agent? I’m not sold yet that this is a huge untapped market.