Tuesday, March 17, 2009

Google Voice: Indexing Real Life Into Search?

This post is a collaboration with our friends and colleagues at eCairn up in San Francisco - Laurent Pfertzel and Dominique Lahaix.

Google's recent move to become the gateway for "shaping the share of voice" (Grand Central) brings about some possibilities that will not only challenge the conventions established around search, but introduce some progressive new forms of thinking. For one thing, most content isn't found through typical search indexing; it has involved into "nanosearch" or "microsearch" functions - just look at how utilities like Twitter are dominating and populating fields of content. For another, we are about to contend with the first real iterations of the semantic web (artificial intelligence). Yet ironically (and somewhat counterintuitive to this), what we must really now consider is that our lives as a whole - our real lives - are now being indexed into search, and in more ways than we can imagine.

As we all know, Google made a fortune with its now famous "we organize the world information" mantra. Facebook, and to a large extent LinkedIn, in combination with other social networks and utilities, are about to do the same with their "we organize the world of people". Granted, the actual methods for monetization will continue to be a sticking point, but if there isn't proof in the pudding, Marketing Vox ran an interesting data-point on how things are rapidly shifting.

What does this mean for the Google pagerank algorithm? For nearly a decade, it's been the invisible cog that the Google machine uses to decide the most relevant information for a search. But the definition of "most relevant" has changed, and continues to change. It's becoming less content-driven as people engage in social media, and more about exchange links, videos, music and so on. By Tweeting a link or adding it to their Facebook page, people tell their connections what's relevant. Further, new search solutions are coming in with platforms that pose a serious threat to Google's monopoly in the world of relevance. As Nick Arnett said, "Twitter is a people-driven, massively parallel headline organizer".

As a very anecdotal data-point, when eCairn published the top 150 social media marketing blogs into its monitoring platform, to no surprise of the group, most of the traffic came from Twitter, Del.icio.us, FriendFeed and StumbleUpon, and far less from other blogs' inbound links or even Google search. Here's the detailed summary provided by Dominique Lahaix:

- One top blogger saw our list and tweeted it

- Twitter stated to bring a lot of traffic, initially individual tweets, then Twitter search

- Del.icio.us came second, as the news spread to more and more bookmarked people

- Almost at the same time we got StumbleUpon traffic

- Early the next day, people started making "derivatives" from our list (OPMLs) and it made the front page of ReadWriteWeb

- We got almost no traffic from search engines (max 100) although we're on page one (#6) for a Google search on "social media marketing blogs", #2 for "top social media blogs" and #8 for social media blogs

So here's where we're going with all of this: if Google doesn't want to see itself outpaced in the race to organize our internet life, perhaps they've purposely, and very wisely, chosen to invest beyond our current social infrastructure. The company, which spawns and supports a number of rising verticals through highly effective open source technology, seems to be looking at a much bigger prize: everyday experiences, and those that are not necessarily designated to, or definable by, offline or online functions. Hence we circle back to Grand Central, it's new voice recognition platform that will likely upend Skype's 400 million-strong captive user base and change the telecom game...among many other things.

So what are we really looking at here? To start, three factors come to mind:

1. Pieces of our "phone" conversations - at least those don't initially violate IP or privacy issues - will flood search queries.

2. These pieces will overtake content indices (text, video, podcasts, etc.) by the sheer volume of conversations.

3. Behavioral analytics (via AI) will take shape in the form of "active capture" versus "active reach".

To clarify, the nature of these conversations will lend themselves to a delivery mechanism in which topical elements or points of commonality previously established through "normal" search parameters will now be verified though real conversation. Further, there are a slew of revenue opportunities linked to micro-targeted environments, and it's certainly not inconceivable that that in this very same way, you will be able to connect people with specific ad content and/or messaging at a precise moment within a physical environment.

Think about it: you and a friend or colleague (or someone you've never met before) are having a chat about cheese at a conference, and all this chat is being broadcast and monitored through the Grand Central platform. When sub-topics like "curdling" or "melting" are mentioned, these exchanges are indexed into search. When queries are run - both on the consumer side and ad network side, for example - you are then strategically fed specific micro-messages on kiosks and digital banners/displays within the environment. You might even opt in to short-code for that cool looking curdling instrument you saw on one banner, and through your PDA, you've made a purchase in a matter of seconds.

What this presents to us is the notion that everyday exchanges will become these fairly stealth (i.e. non-intrusive) engagement points that feed the larger search machine, and, provide a reciprocal framework for accessing the things we want, when we want them. And this is only the beginning...

1 comment:

Gunther Sonnenfeld said...

Sarah - thank you for the kind comment, it is greatly appreciated and always nice to know that someone derives value from reading the material you share.

-Gunther