The quality of Google search has been taking a lot of heat lately. It all started with Vivek Wadhwa here, and was then picked up here, here, here, here, oh yeah and here, here, and here. The criticism has been centered on spam and how Google searches get more and more polluted by spam and content aggregating sites (Demand Media leading the pack).

A whole industry (SEO) has been built to game Google search results and Google’s engineers have been working since the early days to stay ahead of the search game and preserve the integrity of their search. So what’s happened? Have spammers and SEO gurus suddenly become smarter? Is this series of criticism a fluke as rebuked by Google? Or does Google’s own business model stand in the way of Google giving its best effort to sort the problem as suggested by Michael Arrington or KD Paine?

None of the above really…

In order to find the answer to this riddle, we have to reframe the problem: the issue Google is facing has nothing to do with spam or SEO; the fact that spam surfaces in Google searches more than it used to is symptomatic of a much more serious issue for Google.

First, let’s remember what parameters are driving a Google search (though the specifics of Google’s search algorithms are a secret as well kept as Coca Cola’s formula, there is a plethora of literature on the matter in the SEO community):

• The site’s PageRank (based on external links and how content-heavy a site is)
• External link quality (links from other sites with high PageRank)
• Presence and placement of the searched keywords in the page

Now two things have happened over the last decade with the emergence of Web 2.0:

1. The deconstruction of the web page: a web page is no longer a unit of content and there is no norm emerging yet on what the new unit is: an Amazon or Yelp review is a unit of content, so is a response in Quora, a Tweet, a comment in Disqus, a video response on YouTube, a Facebook status, a link with meta data posted in Delicious or Digg. Should I go on?
2. EVERYONE has become an author on the web and we have gone from millions of passive web users getting their information from a few thousand sources to billions of active web user-authors getting their information online from each other.

The impact of this shift on Google’s drivers for search is major:

• The PageRank of a site loses most of its meaning when the site represents a content distribution platform like Amazon, FourSquare, Twitter or Digg rather than the entity creating the content
• Stats associated (link backs, votes, etc.) to each content unit are so small that they make the system very easy to game. This issue is not just faced by Google but also Amazon, Yelp, eBay, Quora, Digg, etc. Small numbers are Heaven on Earth for spammers!

So if we can’t trust PageRank to determine the trustworthiness of a source and if link backs (or derived endorsement metrics) are available in such low numbers per content unit that not much can be done to prevent some from gaming results, what can we do?

TRAACKR is developing a new meta entity around the person authoring content rather than the web page where it gets displayed. By replacing the notion of PageRank with the one of PeopleRank* (assessing the legitimacy of the author rather than the site), the small units of content start making sense again.

Using people as a filter to high quality content is a much stronger and more sustainable paradigm than any other meta entity (whether a site, a Twitter feed, or else) and one that offers enough breadth and depth of data to make the system much harder to game.

TRAACKR’s A-List, our B2B search engine for influencer discovery, exposes our current work on this notion of PeopleRank and the tremendous value it can yield to find authoritative voices on any given topic. Finding these opinion leaders, trusted voices, and domain experts is a stepping stone towards this much bigger undertaking on which we’re working: make search cool (and useful) again.

Could Google do it? Sure. This challenge won’t be solved by tweaking Google’s algorithms though; it will require a paradigm shift away from PageRank, away from Google’s bread-and-butter business, towards this new meta entity and as we all know, ninety degree turns are not easy to maneuver when you carry with you the weight of 4 billion users in 160 countries and 40 languages…

(*) Brian Solis is among a select few visionaries who contributed to coin the term “People Rank” and the first reference I could find discussing it in the context of what we’re talking about here. Check out his post on the topic: January 2008, believe it!

Tags: Google search, Google SEO, Google spam, influencer search, pagerank, peoplerank, Vivek Wadhwa

This entry was posted on Wednesday, February 16th, 2011 at 12:47 pm and is filed under We Search, We Score, We Rant. You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.