Yesterday was a huge day on the Internets especially for those of us in search. Early in the morning Danny Sullivan published an article in which Google engineers stated that Bing was cheating and copying their search results. Most of the article seems to be fairly compelling and on the surface it makes Bing look bad. That is unless you know even just a little bit about how search engines work. Google framed it so the average person, who knows next to nothing about search engines, would believe that Bing is simply copying their results and putting them in Bing. This is a brilliant PR move especially when you release the results of your “test” on the day that Bing has their Search Summit. It’s brilliant when it comes to the PR, for the general public, but dishonest, childish, and downright lame to those who know a little about how search engines work. Hell, anyone who understands the scientific method knows that this “experiment” was faulty from the beginning. I’m sure Google expected to be celebrated for discovering what Bing was doing and that happened in some cases but much of the response was how childish and disingenuous Google was being. Most of the search industry felt the same way, Google was wrong and their test was joke.
Why The Test Was Faulty and Disingenuous
Almost immediately the Search Engine Land article was being shared and discussed. At first people were railing against Bing but when digging into the tide turned. See Google attempted to misinform the public about the true nature of the results. You see, Bing did not copy Google results. They simply used Bing toolbar data as a signal in their ranking algorithm. Google is well aware of this because they are doing the same exact thing with their toolbar data. It’s probably a fairly small element but Google set the test up in a way that would make the signal much more important. This made the test totally faulty and purposefully dishonest. Google engineers knew exactly what they were doing so they could reach the conclusion they already had in mind. So what they did is manually altered their results so that certain pages ranked for nonsense keywords that had little to no coverage in Google or Bing. They then went home and searched for these words in Google via the Bing toolbar in IE where they encountered the results they placed their manually. When they clicked on those results data was sent to Bing signaling that these pages might be relevant for the nonsense searches. They also “conveniently” used fairly strong domains in the fake test to ensure that other ranking signals were passed which also helped these pages rank. So in the end Bing did not copy results but used the fraudulent clickstream data passed through the Bing toolbar along with the other strong ranking signals passed by the pages in question to rank those pages. This is actually fairly smart and I would be totally shocked if Google was not doing the same thing. The funny thing is that Google really didn’t promote the fact that these so called copied results only happened in around 7-9% of their tests.
Industry Insiders Respond to the Google Lies
Bing responded to this fairly quickly, through their blog. As the day wore on more and more search people responded to Google and the overwhelming response was that Google was childish and totally dishonest. Brett Tabke was an entertaining Twitter read as were Michael Gray and others. Matt Cutts continued to offer his best Baghdad Bob impression while holding the water for Google’s latest PR offensive. In the end most who are familiar with the search industry felt that Google was dishonest in their methodology and simply doing this to cut down Bing while employing the same methods themselves. It was an interesting day and it will be fun to see how this plays out in the coming days but it has really lowered my respect for Google. They look whiny, childish, and completely dishonest. And with that I will leave you with some Twitter fun that was had in response to this “controversy”.