[Warning: This post rated PG, for “Pretty Geeky.”]
Late yesterday (Wednesday, June 29) on its “Social Web Blog,” Google announced enhancements to the feature called “social search” they launched several months ago. I blogged about social search when Google first announced it, but, frankly, the topic is so out there on the edge of geekiness and includes issues that touch on many loosely connected, esoteric, and little-understood issues, it hasn’t received a great deal of attention, or, has been reduced by many in the “tech press” to being yet another in the stream of skirmishes between Google and Facebook that seem to be flowing to an inevitable full-scale multi-front war over who owns the “social web” — or some such fuzzy concept.
Simply put (or, as simply as I can), this announcement means that Google is continuing to augment search results (for registered, logged-in users) with information that may be relevant to their search because someone among their network of connections — individuals who they’ve friended or followed or connected with — has mentioned or recommended or liked something related to the search query. (As usual, Danny Sullivan does the best job of explaining the precise features involved in this new enhancement.)
The “holy grail” of search is to provide each individual the knowledge they were looking for when they entered the query. A search engine that can take into consideration who we are, what we do, who our “contacts” are, what we and they have “liked” in-common in the past should, in theory at least, have the contextual data available to understand what we mean if we’re using the word apple: the fruit or the computer. For example, if our network of contacts are apple growers, such contextual data might influence the recommended search results differently than, say, if all our contacts are people we met in line at the Apple Store waiting for the latest iPhone.
Search engine results that have been filtered through personal data (sometimes called “actions” or “jestures”) collected from us and from those who are determined to be “like us,” have been the aim of e-commerce oriented collaborative filtering “recommendation” approaches since the early days of Amazon. The recent $1 million competition called The Netflix Prize shined some public light on the incredible value of improving the “recommendation results” by just 10%.
While extensively used for e-commerce recommendations, using such data to anticipate the “content” we may be seeking is, while not a new idea*, only now is shaping up to be a true battle ground. And one of the terms being applied to this mash-up of several ideas, approaches and algorithmic attempts is called “social search.”
The stakes are huge for this race.
Well, Google’s key to domination of advertising on the web has been two-fold: A payment model that charges advertisers only when Google is successful in generating click-throughs from ads (or, pay-for-performance) rather than mere “page views” and its near-magical ability to consistently place the most relevant ads in front of individuals most likely to be interested in the product or service being advertised.
No one has ever been able to come close to performing the magic of Google’s targeting for narrow-niche oriented advertisers.
No one, that is, until Facebook.
And Facebook doesn’t even have to “guess” (or, predict via algorithms) whose eyeballs it needs to place in front of what advertising. It doesn’t have to guess because Facebook users are perhaps the most generous personal-information sharers in the history of marketing.
From the second they (we) join Facebook, users constantly share such personal information as name and address and zip code and age and gender and educational background and ethnicity and sexual preference and profession and employer and what clubs and associations and houses of worship we are in and what issues we follow and what products we buy and what causes we believe in and support. While Facebook doesn’t sell this data about specific users to advertisers, it does serve-up advertising to users based on all of that information the user has provided Facebook.
Advertisers are beginning to take note. Even local, small business advertisers — the kind of advertiser marketplace that Google has dominated during the past half-decade. Facebook, while not yet offering a web search engine that cooks in all the personal data we provide it, is already using it to allow advertisers to serve up ads that can be even more precise — and effective — than those from Google.
And while Facebook has yet to announce anything that mimics for “search advertising” the Google’s Adsense or Adwords program, such products for publishers and advertisers seem to be a logical extension of its current advertising platform.
Call it what you want: social search, collaborative search, or just search. The goal is all the same: Help people who are searching for something find the most relevant and personal and contextually relevant result or recommendation — and place the most relevant ad possible next to the result.
That’s the holy grail. That’s the war to end all wars.
That’s where Google and Facebook collide.
*Such collaboratively filtered search was one of the key aspects of the original version of SmallBusiness.com, for which we licensed certain very expensive algorithms from the no-longer existing company, Net Perceptions.