Not to be left out of the rise of social mayhem, the fine folks at the ‘plex are more active in the social spheres than many would suspect – at least not at first glance. In truth there is evidence of deep hooks and commitment that might even be going a bit too far… does the social stand for social engineering?
Last time out we looked at some of the potential and problems for Google and the quest for social signals. This time out it’s a glimpse at Google’s world of social networking, behavioral targeting and the data streams that they may be, or more likely already are, pursuing.
Now personalization and social search are great and there is certainly a future for them. What’s not so sure is how they are going to get there. While some of the technology makes me dazzle, there are underlying aspect that can make one nervous as well… As a search geek is a love/scare kinda dealy..
First let’s take a peek at some elements in play…..
Google Account
One of the cornerstones of the Google Empire is your G account, (and profile) which are tied into almost ALL of the other offerings on this list. If one were to mine various elements, there is more than enough to play with, (more recently, can U say SideWiki + FriendConnect?... if they ever get it right).
Google Open Social
Walking hand in hand with the accounts are the various implementations of GOS. As more developers bandwagon along this trail, greater data will be tied into many future social and search activities… (we’ll get back to social profiling shortly)
iGoogle personalized home page
While there are many that discount it, the iGoogle home page has been one of the faster growing services in the Google toolbelt, (and one they’ve been leaning on). The myriad of gadgets offer a variety of both implicit and explicit user feedback. We’ve recently seen Gtalk incorporated into the pages as well…And yes, they have social gadgets too!
Google reader additions
We’ve recently also seen some more social aspects being added to Google Reader including deeper sharing (and friend connection) options as well as limited voting (explicit feedback) via the ‘Like’ feature (not so far from FriendFeed’s approach). This enables profiling of user types as well as secondary explicit feedback possibilities. An ongoing social evolution.
Search Wiki
Another obvious step could be using the data from like minded users in Google’s search Wiki…. This would be of particular interest for any implementation of a social search engine when cross referenced with your contacts and other social connections. How popular this feature is remains in doubt, but it shows the desire for explicit feedback.
Side Wiki
the latest addition to the data kingdom and a semi-social one at that. Who knows, maybe they’re starting to get a feel for getting implicit AND explicit data. The latter generally being harder to get with us lazy users. Or will it go a bit stale like it’s earlier cousin Search Wiki? Only time will tell…
And why stop there? Seriously... it can be rather an endless chain (and we're not even getting into Gmail, Google Docs, Desktop etc..)
Google Talk
This one is also another fairly easy place for them to derive user profiling of ‘friends’ and other contacts in your network. Cross referenced with other applications (such as Reader, FriendConnect) they can likely get tighter user categorization profiling.
Android
Once more we have a few opportunities for them to better understand our networks and interests. Think of all the usual suspects (Gmail, Search) and add another layer of contacts to a potential social profiling program.
Google Chrome OS
Once more, should there be any reasonable saturation in market share this would avail Google a wealth of data points from actual search and social as well as application focus elements (see here for more on application focus). Any operating system is going to be fully loaded with Google applications and services.. and the potetial to even more data...
Getting the idea here?? If we look beyond these we can even consider data from the Chrome browser, Google Maps/Earth (Latitude in Google Maps) and even Google Wave. The list is extensive.... and goes on and on.
(As you well know by now) the system seeks out social connections in order to better deliver targeted ads to social network users. It also has the capacity to be a recommendation engine (of content or other users) as well. We've talked about it many times here, so I'll leave it at that.
One of the more interesting aspects is the Social Graph API. It is certainly a relevant piece - learn more;
The personalized search connections
All of this can also be helped by developing better personalization as well. There are traditionally problems finding quality signals in personalized environments due to the fact that implicit user feedback is often noisy and explicit makes users jump through too many hoops. People tend to not want to interact with their search results, as noted with many social search voting approaches that have met with less than stellar adoption (can you say; Search Wiki?).
If we look at the many signals they are starting to mine from not only our interactions, but our social ‘friends’ and (those deemed) like minded individuals, there could be stronger signals available. Imagine if you will what types of signals can be generated from this wealth of social data to provide a tighter personalization of search. Let us not forget that Google has had an interest in Personalized PageRank since ’03 (be sure to READ that last link…important stuff)
Tin foil and imagination
Another angle worth noting is from everyone’s favourite Google Tin Foil accessory purveyor; Michael Grey. One does have to wonder how all of this information and profiling might affect those in the marketplace. While from the official side of things Google states they are trying to make (your online) life better for you… could this mass of data be used against you? Does social profiling give them more weapons to deal with spammers and others deemed unworthy? It all seems to be a tough balancing act.
And let’s take it a bit further… much like past musing with behavioural targeting that they’re ever-so-interested in we must ask ourselves; where is the line between social/behavioural targeting and modification?
If we implicitly trust a company to ‘give us what we want’, will we be able to see when they are feeding us ‘what they want’ or is in their (and their advertisers) best interest not ours? These are but a few of the questions yet to be answered (or even asked) as we move towards a brave new world of behavioural targeting and personalization with a social layer.
“I find this to be incredibly significant. Google essentially says that Relevance will no longer be defined as “what the user wants“, but as “what we think might change the user’s mind, so as to get them to want something other than what they specifically asked for“. If the user types (for example) for [Reebok] and an Adidas ad appears because Google has allowed Adidas to keyword-target [Reebok] queries, then the user may indeed click (users do sometimes click on results that are not relevant.) There will be cash flow. But under no stretch of imagination would I call the Adidas ad Relevant to what the user actually asked for.”
As if our computers weren't good enough?
Now let’s hold up for a moment – Google is merely the example. It is the one to write about to get folks attention. In reality Microsoft has a ton of interest in behavioural targeting as does Yahoo! too ( and marketing firms around the globe). We could as easily profile them (as they have us) to show some areas of interest, if not concern.
What really drove me were other vertical systems they are also putting in place. Microsoft and Google are fighting it out for data collection and behavioural targeting from television set top boxes… (Google and MS have patents actually). Essentially they want to take the behavioural targeting to new aspects of our lives and soon we’ll be locked in from phone to computer to TV. Something we shall also look at soon….
For now watch Eric Schmidt in this video. The clip will start at the relevant stuff; 10:53 to 11:40 – He talks about set top box ad targeting.
Behavioural targeting is a small step away from behavioural modification
Where is the line? At what point does it begin have an adverse affect on our very evolution as people? Not sure, but it does eat at me sometimes… We haven’t even touched mobile (and geo) related avenues. I am often fascinated while at the same time often strangely concerned about what it all implies to our future (and my children after me).
Great post Dave. You've always been great at picking up on Google's behavioural and personalization habits and I always love to hear your insight as to how they could use this data.
I'm still torn on what I think about all of this. In a lot of ways I would love to see Google use all my data to server up the goods I really want, but I definitely see the potential to offer up the goods that Google thinks I should want.
Well that's always the thing. People say to me "Hey, nothing wrong with me getting more of what I WANT" - problem is that after a while you begin to 'trust' the media sources, that they know you.
But now what happens if they start to slip in the odd item you may not really be needing/wanting.... Will you notice? Or will you subtly be transformed? That's the real problem, is that it has massive possible negative repercussions - depends how it's used.
We shall see... but one thing is for certain, there's a target on our backs....hehe :Psychotic:
Google has been pretty good in that they keep their information fairly transparent, have strong personally identifiable information policies, and make a lot of their info publicly available.
The problem, as Dave points out, comes when they gain your trust then slip little bits away here and there. And with Google's new trends dipping further and further into Doubleclick territory its a little troubling.
And with MS racing with them for the best data networks... yeah, "troublesome"...
It really is insane to think of all the implications-both good and bad- patterns might emerge that end up saving us all (for instance how the pandemics are spotted through search results long before the WHO gets them) instead of the doomsday scenario's that are so easy to conjure up ala matrix ;-)But wow, I can't believe what is possible right. now.
Comments
I'm still torn on what I think about all of this. In a lot of ways I would love to see Google use all my data to server up the goods I really want, but I definitely see the potential to offer up the goods that Google thinks I should want.
Curious to see what the future holds.
But now what happens if they start to slip in the odd item you may not really be needing/wanting.... Will you notice? Or will you subtly be transformed? That's the real problem, is that it has massive possible negative repercussions - depends how it's used.
We shall see... but one thing is for certain, there's a target on our backs....hehe
:Psychotic:
The problem, as Dave points out, comes when they gain your trust then slip little bits away here and there. And with Google's new trends dipping further and further into Doubleclick territory its a little troubling.
And with MS racing with them for the best data networks... yeah, "troublesome"...
http://abovethecrowd.com/2009/10/29/google-redefines-disruption-the-%E2%80%9Cless-than-free%E2%80%9D-business-model/
In response to the news of Google releasing this service for free Garmin stock fell 16%. and TomTom fell 21%.
It will be interesting to see what turn by turn navigatory advertising will feel like.
abovethecrowd.com/2009/10/29/google-redefines-disruption-the-%E2%80%9Cless-than-free%E2%80%9D-business-model/
RSS feed for comments to this post