Welcome to my blog and your digital tech sword!

Thursday, February 03, 2011

Bing Vs. Google: Who's the real villain?

Last evening I came across Danny Sullivan’s recent post “Google: Bing is Cheating, Copying Our Search Results” on searchengineland. Soon after, Google stepped out of the closet with a post on their official blog "Microsoft's Bing uses Google search results - and denies it." Interestingly, it was published right before the Farsight 2011 conference between Matt Cutts (Head of Webspam, Google), Harry Shum (Bing Corporate VP) and Rich Skrenta (CEO, Blekko).

To be honest, I was shocked after reading them. Being an extreme Microsoft fanatic, I was a bit disappointed too. I asked myself "Is this really true? Is Bing actually copying Google's search results?" I was not convinced so I decided to do a little background check.  This was Microsoft's response to Google - "Setting the record straight."

Today, as the facts unfolded after a little analysis and reading some brilliantly written articles like "20 Google Engineers Prove Bing’s Privacy Policy", the truth was out of Google's secret black box. The table turned upside down. I'll now sum up all my findings in a short and an organized way:

My first claim:
  • Microsoft does not copy Google's search result.

To prove my point, I have this little picture that schematically demonstrates how online searches work.


Case 1: In a usual scenario, when the user enters a search query, a search engine uses signals to determine a particular page's relevance to the search term. Signals like - Is the search term in the URL? Does the page title contain the keyword, Does the page contain the keyword, a page's meta tag information, number of times the keyword appears in the page, etc. This happens on the search engine's own database created while crawling the web.

Apart from that it also relies on click-through data on external sites like facebook, twitter, YouTube, etc. For example a video that was highly popular and making rounds on facebook will be ranked higher in the 'Most Viewed' section on YouTube. In this case, YouTube tracks click-through on facebook and uses this information to rank the videos on it's own site. Similarly, both Bing and Google rely on click-through on external sites (including each other) as a source of signal for additional information.

Case 2: In the isolated scenario that Google created in the sting operation (using  honeypot technique) was searching for gibberish non-existent words on both Bing as well as Google. In this case what happened was since the search term was non-existent, Bing was unable to find anything relevant in it’s own database. It did not find any related information on facebook, twitter or any other websites either. So the only source of signal here was Google. Bing did not hack into Google's database or scanned the Google search result page. It only used information obtained from user click-through data on Google, hence the resemblance in the search results.

Not to mention that Bing only showed similar search results in 9 out of the 100 tests (yes, 9/100) Google engineers conducted. Mathematically speaking, even when Google was the only source of signal - Bing only showed similarities in 9/100 cases. Now imagine when common terms are searched for and Bing has over a million external websites where it can look out for data, how much would Bing rely on Google? I'm sure you can do the Math.

Secondly, Bing uses click-through via "Suggested Sites" and the "Bing Toolbar" which are manual user opt-ins. They are not enabled by default. Also, Bing's Privacy Policy clearly states that keywords/URL will be anonymously be sent to Microsoft in order to give the user a better, personalized search experience. Is this wrong? No.

My second claim:
  • Google copied Microsoft.

  1. Bing was the first to introduce the Background wallpaper, remember? Oh that, everybody does.
  1. Microsoft invented the "Page Summary" that pops up on the right to the search results. Google got inspired and came up with "Page Previews".


  1. Remember the old Google image search? Apparently that got inspired from Bing too and introduced the 'column on the left to the images' in the search results page, the 'all results on one-page' concept and the 'preview image thumbnail' when you hover your mouse over it. See the below images for details:


     
  2. Bing was the first to introduce 'Twitter real-time' search. After a couple of months, Google followed too. (read here - http://news.ebrandz.com/microsoft/2010/3237-bing-rolls-out-new-real-time-twitter-search-feeds.html and http://www.pcworld.com/businesscenter/article/174131/realtime_search_google_and_bing_rivalry_intensifies_on_facebook_and_twitter.html)
     
  3. I'm sure they're trying their best to copy the Windows OS too and develop it (as Chrome OS), but seems like copying an OS is way difficult than copying minor UI tweaks on a search engine.

 
Conclusion?

Google's generalization of a small isolated case where only in 9/100 cases Bing showed similarities to Google's result and calling that "Bing copies Google's search results" is absolutely ridiculous. This is the third time Google has cried foul over something which just doesn't make sense.

Recommended Read:
Coming Up Next on the Blog:
Google's Wonderful Blunders.

8 comments:

  1. Interesting and informative read. ;-)

    ReplyDelete
  2. Quite a lot of stuff to point out here:

    1. First of all, your diagram is spot on and is the only way Bing could've gotten the same search results. The fact still does remain that they are, in effect, copying from Google. By collecting user data, they inadvertently collected data directly from Google's search results. It's just not so blatant as it sounds.

    2. Under 'Google copied Microsoft', I would refute claims 4 (as Twitter offered it's data to MS and Google had to collect it) and 5 (it's not a copy. It's a whole different concept). But those points apart, you're saying how Google copies MS in general, and in that case, you have to include how MS copied Google in general. Which again you'd find many examples for.

    All in all, both are companies with very competent engineers and I would call the copying match a stalemate which will continue to occur in developments' name. The only reason I like Google more is because it is at the forefront of the 'richer web with open standards innovation' campaign.

    ReplyDelete
  3. @Suhail - Thank you pointing it out. Talking about the 'richer web with open standards innovation', Microsoft was the first to come out with HTML5 compatible web browser: IE9 with hardware-assisted graphics. Secondly, Microsoft is still taking the stance for Open Standards by supporting the H.264 format whereas Google as clearly refused to do so. More here:

    http://www.zdnet.com/blog/bott/by-dropping-h264-is-google-avoiding-a-trap-or-walking-into-one/2867

    http://www.slashgear.com/microsoft-releases-h-264-plugin-for-chrome-ie9-webm-video-via-plugin-02129993/

    Please try to elaborate what you meant by 'richer web with open standards innovation'? :)

    ReplyDelete
  4. IE = an ancient browser. Where's the CSS3 support in IE8 as compared to all other browsers? #justsaying

    It might be coming with IE9 but other browsers already have it.
    Please also remember the beta experiments of Google on Youtube way before M$ did anything.
    Oh yes, I pity those 47% for now.

    P.S Fanboy, btw you are on Blogspot ('G' Arena).

    ReplyDelete
  5. @Anonymous:

    Why are we even talking about IE8? IE9 not only has CSS3 and SVG 1.1 (2nd edition) and HTML5 but also has hardware accelerated graphics!

    I really did not understand your second statement.

    Yes, I'm a fanboy. And I'm on Blogspot (another arena acquired by 'G'). By that I mean, yes, blogger is good - but it's not Google who deserves the credit.

    ReplyDelete
  6. IE9 isn't the first HTML5 compatible web browser. Chrome has been incorporating HTML5 features with each release having added features from long before MS. No browser is truly HTML5 compatible yet as HTML5 is still a work in progress.
    And H.264 isn't open.
    Apart from this, I also meant webmaster tools and SEO enhancements like the 'review' or 'recipe' snippet that helps search engines return more relevant results.
    The problem with MS is you never know what's going on behind the scenes (same with Google, but to a lesser extent)

    ReplyDelete
  7. With each Release , respective companies comes up with better updates , and it is a competition , but eventually , everything gets integrated. Google's entire strategy seems to be taking a toll on them , because , if MS had or may have copied data , clearly they're not that stupid to do so , not anticipating anyone to find out. Come on.. accidental , maybe , intentional , not a chance!

    remember..Microsoft provided the platform for everyone to learn from and evolve , so essentially , everyone has copied something from MS , sometime , inevitably.

    ReplyDelete