home about contact internet marketing book twitter business book archives subscribe

Google Turns to Crowdsourcing to Fix Its Search Algorithm


The release of Google’s Personal Blacklist might just be a white flag. The new Chrome extension allows searchers to block specific websites from their search results. Each removal also sends a message to Google informing it about sites that users believe should be pushed down the list of results.

It’s a crowd-sourced solution to the problem of Google spam submitted by content farms like Associated Content, EzineArticles and Demand Media’s string of websites — and it might just be effective. Certainly for users of the extension, the eradication of short $15 posts rich in keywords and inbound links but poor in quality information  makes for a refreshing improvement to the search experience. It increases the chances that they’ll actually find what they’re looking for, that they’ll find it faster, and that they’ll feel less frustration as they pore over the results. As those blocks feed back to Google, other searchers will come to benefit too.

The solution might be real and effective, but Google was built on the foundation of a smart algorithm that sorted automatically the Internet’s billions of Web pages, turning them into a useful resource.

When the Library of Congress asks Flickr users to help tag its image collection, that’s an understandable request from an organization that looks after the nation’s cultural treasures on citizens’ behalf. There’s nothing wrong with the non-profit asking those citizens to lend a hand. When SETI calls on computer users around the world to scan the skies for messages from other worlds, that’s a plea from a cash-strapped but exciting venture for the planet’s inhabitants to join in listening for signals addressed to them.

If Google Can’t Write Effective Algorithms…

But when a corporation turns to its customers to provide a solution that its highly-paid developers haven’t been able to solve alone, that’s an admittance of a fundamental weakness in the company’s main product. If Google can’t create an algorithm to bash LiveStrong, what can it do?

Not provide a solution to advertisers. Despite a range of products that run from satellite mapping to 3D sketching, 97 percent of Google’s revenues still come from just one source: advertising. And yet its AdWords advertising system — its main advertising platform — remains extremely frustrating for advertisers.

The complexity of identifying keywords, building targeted landing pages and, most importantly, coping with AdWords’ Quality Score, an opaque measurement that determines the success of an AdWords campaign, means that buying PPC advertising slots on Google now demands specialized training and expert knowledge. Search for a book on Amazon about using AdWords and you’ll be offered a bewildering choice of more than 780 publications.

And they’ll all recommend painstaking processes of trial and error to determine the most effective copy, the best search phrases and optimal landing pages.

If you’re going to invest that much time and effort into buying traffic, you might as well buy a book on SEO and get the visitors for free.

Or turn to Facebook. Compared to Google, Mark Zuckerberg’s company is a minnow. While Google declared profits of more than $8.5 billion from revenues of $29.3 billion in 2010, Goldman Sachs has estimated Facebook’s profits for the first nine months of last year at $355 million on revenues of $1.2 billion. (Other analysts have put the figure higher and have predicted that Facebook’s revenues will reach $4 billion this year.)

But Facebook was founded in 2004 when Google was already six years old — middle-aged in Internet terms. While the social media site’s rate of revenue growth has been slower than that of Google (in 2004, the search company was already reporting revenues of almost $3.2 billion), it does now face serious competition in the online advertising market. One valuation by SharesPost, a secondary trading market, makes Facebook the second most valuable property on the Web, trailing only Google.

If Facebook’s advertising system, which allows advertisers to target individuals instead of search phrases, is found to be easier to use and more effective than AdWords, Google’s problems will extend beyond its battle with content farms.

Turning off Google TV

Even outside its core industries Google is struggling. Its foray into television has turned viewers off, and its social media platform Buzz seems to have faded away after failing to take on Twitter and being outraved by Quora.

Perhaps the only positive sign for Google has been the growth of Android, the mobile operating system the company bought in 2005. The platform is now said to be the second most popular mobile OS in the United States, having overtaken Apple’s iOS and trailing only Blackberry’s RIM.

Android’s popularity has been helped by the fact that it’s offered on such a wide range of different devices, including the latest tablets such as Motorola’s Xoom. Apple might be in third place but it’s also on only one phone, a product that brings in revenue for Apple with each sale. Google, however, earns nothing when a manufacturer chooses to use Android instead of Windows Phone 7.

That fragmentation comes with a cost too. Rovio Media, has had to release several Android versions of its Angry Birds game to cope with the different releases used on handsets. Only 36 percent of Android phones are running the latest edition.

For Google, none of these problems may matter yet. Despite the rise of Bing, the company still dominates the search market. Despite the rise of Facebook, Google is still streets ahead in terms of size, revenues and product range. And despite Android’s unpopularity with developers, its spread across mobile platforms means that it’s able to continue serving ads to Internet users wherever they may be and however they’re surfing the Web. Google may no longer enjoy the 100 percent year-on-year growth it saw in its early days but net income still increased by a third from 2009 to 2010 even as the rest of the economy stumbled.

Google might be getting a lot of things wrong — far more than it used to — but for now at least, its bottom line looks right. The question though is whether a nimbler upstart can now do to it what it did to Lycos and Yahoo! — and how long it will take for them to figure out a way to crowdsource a lower ranking for technology search results from 2004.



One Comment

  1. philippe zara Says:

    how have things played out since this article was written a year ago?

Leave a Comment