[asterisk-biz] Google's voice product [OT]

[John Todd]

On Mar 17, 2009, at 1:42 PM, David Gibbons wrote:

> I think we're overcomplicating things here.
>
> All they have to do is use their voice recognition software (you  
> know, the one that we all train for free when we call 800-goog-411)  
> to do basic transcription, and low and behold they have marketable  
> data. Just like advertisers want to know what we do online, they  
> will want to know what we walk about on the phone once that data is  
> available.
>
> Advertisers would pay an arm and a leg to know every time someone  
> mentioned their product on the phone and in what context they  
> mentioned it...
>
> --Dave
> [snip]

I would agree on this point, and think it is core to the whole Google  
move into voice.

Google is first and foremost a search company.  They provide  
ubiquitous access to knowledge, and allow certain parties to influence  
the results in a manner that favorably presents the products of those  
parties based on context of the search.  ("influence" means "present  
ads along side of" in this setting.)

Currently Google is missing out on several huge areas in which to  
assert their influence of knowledge.  Print media is one, but the lack  
of interactivity there means it's a dead end (though they are busily  
scanning books, though perhaps nobody cares to read any longer.)  TV  
is another, but they've not done so well there, either, because they  
were asserting just their placement model and not the fundamental  
attraction of the content itself, and also they have no ability to  
ensure good placement even with what they've got.  (YouTube is the  
counter to this, where they've marshaled the content and done very  
well with ad placements.)   Personal music devices is one place they  
haven't been (see below).  Person-to-person conversation is the one  
that they're notably lacking as well.  I suspect GoogleVoice is the  
first tentative step into the person-to-person stream of data.

If they could insert themselves into the audio stream of  
conversations, they would gather an impressive volume of data that  
could be used even in an anonymous fashion.  While there may be  
resistance to a whispering voice on my calls pitching adjustable  
mortgages, I suspect that there will be less objection to "aggregate"  
intercepts if the exchange is free voice communication or other  
feature set that is fashionable.  Personal identification may not even  
be required to make this data valuable - it is almost as useful to a  
marketer for them to know that 28 year old women in San Francisco with  
an income level of more than $100,000 and who is a commuter and who  
happens to have mentioned that she thinks that the new Lexus model is  
a bit too expensive this year.

Android is the next-best thing to being inserted in the communications  
channel, as it provides ubiquity of access.  I don't think they'll  
replace Apple for a while, but I suspect that Android-based devices  
will eventually take over the market because of the open  
capabilities.  I also expect that Android on mobile phones is step one  
of a several-segment process, of which Google may only be partially  
aware.  The extension of Android into consumer-grade reality  
augmentation devices (audio, layered heads-up-displays) is where the  
value becomes staggering for advertising and knowledge influence  
potential.   But first, people need to want to carry "version 1"  
Android devices around and get used to trading privacy for  
convenience. Wait 5-10 years, and we'll see what's happening on that  
front.  We won't see this type of interception for a few years, at  
best - the technology isn't scaleable enough yet, nor are people  
comfortable with the concept... yet.

What I hope continues to be the case is that there is the ability to  
opt out of interception and parsing, as Google has done with some of  
their other products.  However, most people simply don't care about  
that issue as long as there is some prize you hold out for them.   
Google currently has a decent prize - very good search capabilities.   
However, I've seen people willing to trade their privacy for far, far  
less and I expect that Google will continue to find a good revenue  
margin between those who are willing to pay for influence and those  
that are willing to be influenced.  I don't lament Google that  
business - it's clearly what people want.

In summary: Google has many reasons to want to be in the voice  
business, and none of them have anything to do with minutes or call  
revenues.

PS: It is inevitable, though perhaps in a slow fashion, that Google  
Voice will envelop services delivered via Android as well as  
communications channels delivered via GoogleTalk.

PPS: This will all eventually come down to a very, very nasty battle  
over spectrum since that ultimately dictates who gets to control the  
influence on the ears/eyes of the user.  The first salvo has been  
fired with the white space auctions of last year.

PPPS: Sorry about the Asterisk diversion here.  I just happen to have  
done a lot of thinking on these topics.  This is actually relevant to  
Asterisk, I believe. Asterisk will continue to be the toolkit for  
integrating businesses into Google voice services in many instances -  
I anxiously await a more open API for Google Voice.  (i.e.: analog  
interfaces, IT integration, fax, IVR mappings, Skype, call queues,  
etc. are probably best done in Asterisk even if Google starts to look  
more like an alternate last-mile or last-foot data delivery channel.)

JT

---
John Todd                       email:jtodd at digium.com
Digium, Inc. | Asterisk Open Source Community Director
445 Jan Davis Drive NW -  Huntsville AL 35806  -   USA
direct: +1-256-428-6083         http://www.digium.com/