Over the weekend, Blake Ross, Facebook's product director and co-founder of Firefox, worked with Facebook engineers Tom Occhino and Marshall Roch to demonstrate how evil Google's newly launched Search plus Your World (SPYW) feature really is, and created a "proof of concept" showing how it should really work. His team got some help from Twitter engineers, Myspace engineers, and consulted other social networks as well to really make sure the message hits home: SPYW should surface results from all social networks, not just Google+.
By leveraging Google's own algorithms, the group built a bookmarklet called "don't be evil" (a jab at Google's informal motto) and released it on a new website named Focus on the User. Ross sent out a message on Twitter about the initiative this morning: "Google results are more relevant when they include all social networks rather than just Google+. Check it out: focusontheuser.org."
Google says it requires permission from other social networks like Facebook and Twitter to deeply crawl their websites in order to provide a consistent experience with SPYW. Since Facebook and Twitter don't give Google permission to their users' private content, SPYW mainly features Google+ content. The team's goal is to show Google is lying because the search giant already indexes all public information on social networks, and there's no reason why it can't use that data as well.
To use the new "proof of concept" Google, you can install the bookmarklet yourself for Chrome, Firefox, and Safari (Internet Explorer and Opera users are out of luck). Go to the website, click on the blue "Try a More Relevant Google" button, and drag the "don't be evil" button up to your bookmarks bar at the top of your browser. If you don't have a bookmarks bar in your browser, look in the settings and enable it first.
Once the bookmarklet is installed (you’ll see it appear on your bookmarks bar), all you have to do is click on it to see completely different Google search results after you submit your query. The new results include links to CrunchBase (not exactly a social network, but okay), Facebook, Flickr, Foursquare, FriendFeed, GitHub, Google+, LinkedIn, MySpace, Quora, Stack Overflow, Tumblr, Twitter, instead of just Google+. Once you've visited google.com and clicked on the bookmarklet once, it will continue to work for subsequent queries until you click away from google.com. Next time you visit google.com, however, you'll have to click the bookmarklet again.
So, how does it work? If Google's search engine decides that it's relevant to surface a Google+ page in response to a query where Google+ content is hardcoded, the tool searches Google for the name of the Google+ page and identifies the social profiles within the first ten pages of Google's search results (top 100 results). The ones Google ranks highest, regardless of what social network they are from, replace the previous results that would only be from Google+.
To be clear: the tool not only reorders the search engine results, but also the results of the promotional Google+ boxes on the right side of the results, as well as the autocomplete results that feature Google+ accounts when you type into the search box. In Google language these three are known as: People & Pages results, Google+ Sitelinks, and Google+ Suggestions In Autocomplete.
Here's the Focus on the User's official description:
How much better would social search be if Google surfaced results from all across the web? The results speak for themselves. We created a tool that uses Google's own relevance measure—the ranking of their organic search results—to determine what social content should appear in the areas where Google+ results are currently hardcoded. All of the information in this demo comes from Google itself, and all of the ranking decisions are made by Google's own algorithms. No other services or APIs are accessed.
The last part is key: Google argues that it would include results from Facebook and Twitter if it was given access to all the data from the social networks. Last week, Google executive chairman Eric Schmidt said: "The core question is, 'Would we be willing to [include Facebook and Twitter], and the answer is, 'We'll be happy to talk to them about it'."
As you can see in the video above, this little coding project clearly shows that Google doesn't really need the data it is asking for in order to make SPYW work fairly. The bookmarklet never accesses any server or API outside of google.com. The information being provided has already been indexed and ranked by Google.
The bookmarklet even shows profile pictures, which are pulled directly from Google. The tool enters the address of the new social result into Google's Rich Snippets Testing Tool and uses the information that Google provides. In other words, Google is only showing images for Google+ results, even though it has images from many other services.
By the way, the tool might not work for you because SPYW hasn't been rolled out to everyone yet. Google has pushed out the new social features on google.com to a percentage of U.S. users. If you don't see any of the Google+ social results on Google yet, you do not have the feature yet, so you can't use the "don't be evil" tool.
Google is trying to use SPYW as a bargaining chip. I'm sure the search giant would love to have more data from the aforementioned companies, especially your Facebook social graph (your Friends Lists, who your friends are friends with, and so on), which is what the last battle between Facebook and Google was about. Facebook and friends are fighting back.
Facebook declined to comment on this article.
- Why Facebook is not part of Google's SPYW
- Facebook, Twitter react to Google's SPYW
- Google to Facebook: "We are delighted to be underestimated"
- Zuckerberg: Google, Microsoft collect data "behind your back"
- Google: Facebook is becoming "a closed walled garden"
- Google: Facebook is blocking Google+ invite links