ETel - Why Let Web Developers Have All the Fun?

Etel_001 Trevor Baca summed up the spirit of the Emerging Telephony (ETel) conference when he said "Web developers have all the fun.  Why can’t the phone be as easy?"  Indeed, the conference was all about employing Web 2.0 concepts such as Web services, mashups, and open source in order to enable new ways to use the telephone.  ETel brought together more than 250 practitioners, mostly from small companies, for three days of presentations, discussions, and exhibits at the San Francisco Airport Marriott last week.  The big legacy carriers were conspicuously absent, with the exception of British Telecom and France Telecom.  The handset manufacturers were represented by Nokia and Motorola, and the big Web properties by Yahoo and Microsoft.

If Telephony 1.0 was The Phone Company providing a monolithic experience encompassing everything from directories, to handsets and the network connecting them, Telephony 2.0 is a more diverse set of applications and the underlying services.  These applications encompass the traditional phone call, but calls are more likely to start with a name or a web link rather than a phone number, can be routed in any number of innovative ways over a variety of networks, and can be delivered to a wide range of devices.

A taxonomy of Phone 2.0 features:

  • Addressing.  In the 1.0 world, you dialed a number you knew from memory or looked it up in a directory.  Mobile phones popularized the shortcut of dialing straight from the directory listing, whether in the handset itself or through an extra-cost directory assistance service.  In the 2.0 world, the phone number is just a low-level address and may even be hidden to preserve the privacy of the called party.  The caller can start with an identifier such as an iname (e.g. =cherot) or from a web site such as a social network or discussion forum.  An application can enable the caller to go directly from a request such as "find someone who can answer my question" to ringing a phone of an expert without requiring the intervening steps of looking up a list of recipients, finding the associated phone number, and dialing the phone.
  • Call Initiation can take place by "dialing" a number at a traditional handset, but it can also be performed by an application that may reside in that same handset (e.g. TalkPlus), in a web page (Jaduka), in a personal computer (Skype), or anywhere in the phone network. 
  • Filtering.  In the old days, everyone had a secretary, or wished they had one, to decide which calls to put through and which ones for which to take a message.  Then came answering machines, voice mail, and caller-ID as primitive filtering mechanisms.  Now applications such as GrandCentral can filter calls based on calling number, called number, time of day, and the schedule of the recipient.  Calls can be put through to the intended caller, deflected to voice mail, or even connected to a fake "out of order" tone - the latter reputed to be a means to get removed from automated telemarketing calls.
  • Routing.  The same parameters that can be used for filtering can also be used to select one or more addresses (phone numbers) to which the call should be delivered.  Some services allow the call to be extended to several numbers at once, with the call connected to the first one that answers.
  • Call presentation.  Just ringing the phone is so last millennium.  Near the end of the 20th century ringing bells were replaced by computerized beeps, ringtones of popular songs, and in a retro moment, the digitized sound of a mechanical bell.  Those sounds are still with us, but the call may delivered to a PC before or in lieu of being delivered to a phone next to it on the desk.  In addition to making noise, the PC can display a pop-up window or toast containing any number of relevant facts, such as the caller ID, caller's name and (in the case of calls initiated from a web application) what web page the caller is looking at.  While the call is being set up, the caller can be informed of the progress of the call by ringback tones, recorded messages, and (in the case of web-initiated calls) visual displays.
  • In-call services.  The basic phone call is still a voice connection between two parties, but that call can include any number of callers, be recorded, and be carried over channels with considerably higher fidelity and/or lower cost than the traditional phone network.  The call may also include other collaborative activities such as showing a presentation or sharing a screen, e.g. Convoq's ASAP.
  • Post-call services.  Phone companies have long generated billing records which can be used to generate reports and databases, but these databases can also be used to enable follow-up activities such as returning the phone call.

Making all of this possible is a dramatic change in the way phone applications are developed and deployed.  What formerly required a substantial investment in specialized telephony equipment and circuits can now be performed with standard servers connected over the public Internet.  Companies such as Global Crossing and Level 3  will accept calls over the Internet in SIP and deliver them for pennies to any phone in the world.  They will also provision local or toll-free numbers anywhere and deliver them to you the same way.  You can process these calls with any number of commercially available software packages or go open-source with Asterisk.  If you don't want to operate your own servers, you can send VoiceXML and CCXML to Voxeo or SOAP to Jaduka.  The last bastion of this revolution is the mobile phone network, especially in the USA where the handset business is tightly controlled by the mobile operators.  While Apple is trying to upend that model with the iPhone, it has yet to embrace any commitment to openness of the platform.  Meanwhile there are efforts to product an open-source, Linux-based mobile phone: the Qtopia Greenphone and OpenMoko.

As the motto of the ETel conference states:

Opportunity doesn't always knock.
Sometimes it calls.


Frankencode

As I was enthusing to my friend and former business partner Doug Levin about the quantity of open source telephony software I saw at ETel this week, he remarked about the amount of open source that was Frankencode - software of dubious provenance and unknown reliability.  Doug should know.  His company, Black Duck Software, is kind of an automated policeman to track an organization's software assets.  I guess asking him about software is like asking a cop about human nature - he sees all sides of it.

That said, there is a sizable movement behind Asterisk and the company behind it, Digium, attracted $13.8 million from Matrix Partners last summer.  And there are a number of open source handset projects, including the Qtopia Greenphone and OpenMoko.  However, as Alex Russell pointed out at the conference, the wireless operators view such efforts the way the characters of Shelley's novel viewed Dr. Frankenstein's creation.

Emotional Adoption Curve

Emotional_adoption_curveOne of the most interesting and useful presentations at ETel was not about telephony at all, except in the most general sense.  Jeff Bonforte of Yahoo had to abandon his original presentation at the behest of his lawyers and instead gave an excellent discourse on understanding the emotions behind consumer adoption of new products.  In an update of Geoffrey Moore's Crossing the Chasm, Bonforte called attention to the emotions behind the behavior of each segment of the adoption curve and gives them new names:

  • Lovers
  • Irrational
  • Efficient
  • Laughers
  • Comfortable

The Lovers (Innovators in Moore's lexicon) are the techies who buy the product because they find the technology intrinsically interesting.  They can send misleading signals to the seller because their emotional response to technology is the opposite to that of the larger population.  They look on solving tough technical problems as fun.  This 3% of the market is likely to acquire product X through the "X Club."

The Irrational (Early Adopters) feel the same emotions are the general population, but feel them with more intensity.  These are often negative emotions such as anger, fright, or loneliness.  The strength of these feelings can lead to buying behavior that is not economically rational, such as installing Skype because one hates the phone company,  The good news is that once the technology improves, ordinary people who feel the more moderate versions of the same emotions will also be motivated to buy.  There are plenty of targets for these emotions, such as taxes, banks, telcos, health care providers, government services, airlines, lawyers, Microsoft, the opposite sex, and "The Man" so there is plenty of opportunity.  This 22% is likely to purchase product X at The X and Y Store.

The Efficient (Early Majority) will purchase when the technology becomes practical.  This 25% goes to Best Buy.

The Laughers (Late Majority) are Yahoo's core constituency.  This 35% goes to Costco.

The Comfortable (Laggards) are the 15% who go to Walgreen's or Safeway because the Costco parking lot is too confusing.

ETel Launchpad

Om Malik and Surj Patel kicked off the ETel conference with a startup Launchpad, six minute demos of new products from 7 companies:

  • GrandCentral - single number reach + unified messaging
  • OpenFire (formerly Wildfire) from Jive Software - XMPP Instant Messaging server that does VoIP
  • Cellcrypt - client for mobile phones to push encrypted voice mail
  • Peerant - Ruby P2P app to connect visitors on web sites to agents
  • mySay - phone number to leave and retrieve group voice messages
  • Flat Planet Phone Company - Platform for SMB voice services - $199/year for reseller partition
  • MIG33 from Project Goth Ltd - VoIP over mobile data

The winner is to be decided by sending a text message to Mozes  - at this writing GrandCentral was ahead, followed by MIG33.

My Photo

Other Places to Find Me

Tracking