• messagelost

    I’m not clear on why taking Google’s results and a lot of other search engine results and redisplaying them without attribution somehow resolves the morality of the matter. Also, considering Google is at the very least a plurality (if not outright majority) of the world’s searches, a “search signal” is, substantially, a Google signal.

    The most confusing part, however, is after spending several paragraphs downplaying the importance of the signal, he won’t even consider removing it. If it’s such a “weak” signal that it would only effect 10% of synthetic queries, why not remove it if it can result in displaying a competitor’s results as your own?

  • http://fjpoblam fjpoblam

    Still, Matt Cutts’s article seems pretty convincing: My thoughts on this week’s debate http://bit.ly/hxl0Hk

  • http://www.2buildbacklinks.com Mark Collier

    Hi Danny

    Thanks for the great post and I totally agree with all your comments. Bing are merely tracking clicks and making a connection between Page A and Page B and in this case Page A is a Google search result and in fact this Bing algorithmic factor isn’t targeted at Google.

    What I wonder is why Google didn’t recognize this, or more likely why did they continue to publish on this topic if they did?

    I also wrote on this topic @ SEOMoz’s YouMoz: http://www.seomoz.org/ugc/the-bing-sting-facts-why-bing-arent-copying-google-11924

  • pyro

    While Bing’s first result for “bombilete” isn’t a copy of Google’s first result for “bombilete”, it is a copy of google’s first result for “bombilate” and most users will follow the spell correction.

    Google seems to understand the spell correction and Bing doesn’t and yet Bing still returns the page Google is going to send a large number of users to. Still looks like copying to me.

  • http://www.resourceshelf.com gdprice

    We all know that page estimates are estimates. That said, when you mentioned the number of results from Google and Bing for the term pontneddfechan I was struck by the large disparity.

    I went to Google and Bing and checked using pontneddfechan and bombilete. Of course, these are also estimates but after running them several times on several browsers they appear to be, for the most part, consistent.

    Search Term: pontneddfechan

    Google Total Estimate: 72,600
    Actual: 570 Results (with filter off: 576)
    Browser: Chrome

    Bing Total Estimate: 5530
    Estimate: Around 410
    Browser: Chrome

    Search Term: bombilete

    Google Total: 46
    Results: 25
    Browser: Chrome

    Bing Total: 11
    Results: 11
    Browser: Chrome

    So, from what I found both Google and Bing are relatively close in terms of total pages.

  • http://www.2buildbacklinks.com Mark Collier

    @pyro
    While the result ends up the same in some cases they are not copying because they are looking at the click data.

    For example if you went from a page about bombilete, for example a Yahoo Answers page to another page. Bing makes a connection between the two pages thus suggesting the page that you clicked onto is about bombilete in some shape or form.

    The reason it looks like copying Google is because in this case the original page is a search engine result page and not a normal information based webpage. Thus the illusion of copying is created.

    This Bing algorithmic factor is not targeted at Google and only holds a small weighting (that’s why the impact is only seen on very long tail searches).

    Hope that clears it up

    Mark Collier

  • http://about.me/teusje teusje

    took some screen shots earlier this week:
    http://teusje.wordpress.com/2011/02/02/google-bing/

  • Chandrashekar

    @messagelost:
    Google DOES have 64 % of market share. Ultimately it’s not google results that dictate the relevancy of a site. It’s people’s choices.

    For Example: I might serach for Santa Claus. Out of the 10 results shown i only find 5th one relavant. That sort of data is what decides the rank of a page. Bing cannot remove that signal cause it would mean losing 64% of user behavior.

    @fjpoblam
    He did leave out some key facts:
    1. Over 100 “crazy” seraches were tried out. About 90 of them failed. The ones shown are the remaining 10 ones.
    2. Did google try to replicate the behavior in other search sites ? Was there a control group in the experiment.
    3. What does google do with the data from google toolbar and google analytics. Does Chrome collect such data ?

    No official response from Google so far on these pertinent questions

  • pyro

    @Mark Collier

    If it was such a small factor bing would have dropped it 4 days ago. It’s gotta be a pretty strong signal to put up with this much bad press. I’d imagine the spammers are going to manipulate going forward so even if it used to be a really strong signal it’s about to get much weaker.

    How hard was it for Google to figure out the spell correction/find 10 decent results for long and medium tail queries? How hard was it for the user to click on the top result? While you can claim you’re just “analyzing the click stream” Google has a lot more to do with those results being clicked on that the users do. That’s why everyone thinks it’s copying.

    It’s really easy for Bing to say going forward we’ll ignore all clicks of google.com properties and make a good faith effort to ignore obvious partners. If this is such a tiny signal this wouldn’t even hurt relevance. Instead the Bingers are twisting themselves inside out coming up with justifications for why they’ll continue to do this (which is essentially copying) all while maintaining that it’s somehow not copying.

    Either it’s a weak signal and they’d have already dropped it or it’s a strong signal and they simply copying google.

  • http://www.2buildbacklinks.com Mark Collier

    @pyro

    It is clear that Google must not use this signal as they don’t seem to have thought about it while doing this research. Therefore it is safe to assume that Bing have a competitive advantage over Google, small or large in terms of a unique, logical and quality ranking factor, so why would they remove the factor and their competitive advantage?

  • http://thecodemonk.com Aaron Smith

    So, Google accuses Bing of copying information, that they can’t explain how it does it, or that why it only worked in 7% of their tests… Bing explains why, which is watching the clickstream data (which google toolbar does, tests have proven this in the last few days), yet no one sees the problem with all the spam in google searches (which google makes money off of), and no one sees a problem with google scraping travel info, reviews, etc, for their maps? Google copies all that data and makes it look like it’s coming from them until you click on it… They are essentially doing more than what bing does as far as copying, yet no one is calling them out on that?

    Pot, meet kettle.

  • pyro

    @Mark Collier

    That’s sort of the point. For Bing, copying Google is a competitive advantage. Not so for Google. I’m sure both analyze toolbar click streams but Google already knows what people do on Google.com and probably doesn’t derive much extra benefit from toolbar click streams. Bing on the other hand seems to desperately need to cherry pick Google results or they’d have dropped the practice (at least when the clicks deal with Google properties).

    They can’t have it both ways – it’s either a weak signal and not worth this PR mess or a strong signal which means it’s roughly equivalent to copying Google (albeit in an obfuscated way.)

  • http://onvento.com Onvento

    One basic problem i see with google’s case is that they are making their conclusions based on arcane search phrases. That is not where the bulk of searches happen.

    for frequently used search phrases Iam very sure Bing & Googles results will be dissimilair.

  • http://www.2buildbacklinks.com Mark Collier

    @pyro

    Its not what Bing know about what people do on Google.com. What they do want to know is as much about the user as possible.

    Therefore yes, Bing needs to track actions on Google but also every other website in the World Wide Web.

  • http://searchengineland.com/ Danny Sullivan

    Mark, Google understand that Bing is making a connection between Page A and Page B. Google argues that this connection is, in some cases, causing Bing to effectively copy its results. Bing disagrees.

    Pryo, Bing may be auto-correcting the spelling for bombilete, just as Google itself sometimes does.

    The point is that in the first case Google presented, Bing was showing the right answer for the misspelled search term because, Google argues, it could see the right answer due to Google’s auto-correction.

    In this second case, Google doesn’t auto-correct. A user would only see the “right” answer on Google if they manually clicked on the spelling correction. Doing that would generate a new search — for the correct spelling — which Bing would potentially see. In turn, that would largely only help Bing with answers for the correct spelling, not the incorrect one.

  • http://www.2buildbacklinks.com Mark Collier

    @Danny

    Thanks for clearing that up, but naturally in some cases where that connection is all Bing have on a keyword/query then it will lead to duplication of results. But surly that isn’t true for often searched queries where the information on potential results is abundant?

  • http://www.michael-martinez.com/ Michael Martinez

    \The point is that in the first case Google presented, Bing was showing the right answer for the misspelled search term because, Google argues, it could see the right answer due to Google’s auto-correction.\

    So when Google shows that Bing replicates 1,000 search listings out of 1,000 search listings for greater than 90% of their test queries, you’ll have a story to write about.

    Meanwhile, Google looks stupid for using inadequate data to prove outrageous accusations against a rival that had already publicly disclosed it was following user clicks for part of its discovery process.

    I’m not sure what Google’s excuse for looking stupid is, but what is the SEO community’s excuse for dragging this out?

  • stealthrobot

    I wonder what people would be saying if the experiment had been a little different. What if Google had created honeypot pages with hundreds of results instead of just one? And what if their engineers then proceeded to click on the 50th result instead of the first? And let’s say for the sake of argument, that previously Bing had ranked that result on their engine in the 9th spot.

    Would Google argue that Bing is stealing their results if they then promoted that result to 7th or 8th?

    I could see the argument for Bing “stealing” results from Google if they actually captured the relevance ranking Google assigns to a result, but the claim is that they don’t. What Bing claims is that they care about user relevance, specifically, what url a user navigates to after performing a search. It shouldn’t matter if that is the first result or the hundredth; whether it is on Google, Bing, Amazon, or any other site that provides search.

    “Copying” and “stealing” are hyperbole. Bing is just trying to learn what pages are relevant to users.

  • webmihir

    I asked Matt cutts the following question on his blog with no response yet:

    ———-
    Matt,

    Your claim that the IE suggestion dialog does not make it clear that it is sending information to MSFT is good. But then, the Google.com page does the same – the fact that my search history is being anonymously collected is hidden deep in the Google ToS somewhere – why is that not on front page either?
    ———-

  • http://wordpress.org/extend/plugins/seo-rank-reporter/ David Scoville

    Danny, I appreciate that you’re trying to be a true journalist by presenting both sides of the story from both Google’s and Bing’s perspective.

    That said… I’m a web developer. I hate IE–and this supposed scandal really makes me angry at Bing.

    Bing should have never used Google’s results in its algorithm. That’s like Bing trying to beat competition by copying the competition. Usually, in this free market economy, you beat the competition by innovating–and this is not really innovation (even if you want to call it “capturing user data”). In the end, we, the end users, don’t get a better search engine–just a different logo.

  • http://searchengineland.com/ Danny Sullivan

    Mark, no, in cases where Bing has more than a pure signal coming only from Google, it seems very unlikely to duplicate Google’s results. Bing said that in the story above.

    Michael, any time a major competitor takes a step of calling another major competitor a “cheater,” you have a story to write about. That alone is news. For Google to do it is pretty unprecedented.

    But in terms of the number of queries, Google claims this is causing Bing to effectively cherry-pick the hardest things they solve in search, the stuff where they feel they really stand out. Bing, of course, says it’s just something that rarely happens. That they both are so opposed in how they view this is one reason for the conflict.

    I think it’s dragging out because we still have some unanswered questions we’d like answered. As part of that, we’re learning some new things about how Bing makes use of clickstream data. Potentially, we’re going to learn more about what Google does with it.

    Stealthrobot, if Bing learns that a page is “relevant” to “users” simply because Google was incredibly smart in (1) finding the page on the web and (2) ranking it above all others, especially if the user entered an ambiguous term, did Bing really learn from the user behavior, or did the user behavior simple serve to funnel back to Bing what Google thinks is right for a query? That’s the crux of the argument.

    Webmirhi, yes. In fact, I commented about something somewhat related at Matt’s blog, saying that while Matt complains that Bing’s data collection tools aren’t giving clear disclosure, neither are some of Google’s.

    When you install the Google Toolbar, you get told:

    “PageRank and future page-related services are part of the enhanced Toolbar. For enhanced Toolbar features to work, Toolbar has to tell us what site you’re visiting by sending Google the URL.”

    Nothing in there discloses that the toolbar will be used to gather site speed information which, in turn, influences how sites rank in Google. Nor does it say that your searches on Bing will get sent back to Google — which they will, even if Google says it doesn’t use them.

    In addition, the toolbar doesn’t tell you to say yes or no. It says “Ask Me Later” or “Enable Enhanced Features.”

    I think it’s a waste of time to be focusing on the disclosure aspect. Google has raised it as evidence of supposed Microsoft wrong-doing. I looked at it in my first article and didn’t find a problem.

  • http://www.michael-martinez.com/ Michael Martinez

    @Danny — I understand why YOU continue to write about the subject. I’m good with that. But everyone else has to fuss over this?

  • webmihir

    @david your rage for IE is justified. That was the Microsoft of the past though – IE9 seems like a step in the right direction. Them enabling H.264 support on all Windows browser is another step in the right direction, instead of Google’s position to use a self-developed standard. I honestly think people need to give Microsoft a second chance rather than constantly looking at the past – especially since Google needs some competition in areas related to privacy.

  • Michael Flaster

    Danny – I think you’re missing a subtle point with the spelling correction. For [torsorophy], on Google’s blog, Bing is only returning 2 results. That’s it. Both Google and Bing have many more results for the correctly spelled word. On Google’s page, they present a link to “did you mean”, to give the user the option to access all the results.

    Of Bing’s 2 results, one is the same result as Google. The other is nonsense. There is no link for “did you mean”. Correct me if I’m wrong, but every spell correction on Bing (and Google) will either show you results for the spell corrected word, and/or give you a link for the spell corrected query. I mean, it makes sense. If Bing knew it was doing a spelling correction, why wouldn’t it give access to all the other pages associated with the correctly spelled word?

    This seems to me to be a clear indication of Bing saying “here’s 1 good result for this query, but I have no idea *why* it’s a good result, so I can’t show you any others”.

  • http://bridgerbay.com Tym Barker

    If Google applies a penalty to Bing, and Bing supposedly copies Google, will this lead to an infinite loop of zero rankings? LOL

    Will they respond to each other’s re-inclusion requests?

  • http://www.stareclips.com/?twitter Bob Bigellow

    Maybe his son will come home and say, “Dad, I didn’t copy my friend’s quiz answers. I merely observed the patterns that his pencil moved in, then I applied those movements as a ‘signal’ to the movements of my own pencil. Don’t worry, though. I was able to answer many of the quiz questions myself. I only observed the movements of my friend’s pencil when I didn’t know the answer to the question on my own.”

    Semantics, the reason Google had a problem with this copying – let’s call it what it is – is that they aren’t really innovating. Google has been pushing their technology with machine learning algorithms to attempt to understand the meaning of textual content in the same way humans do. In this manner, the progress of search engine algorithms moves artificial intelligence forward.

    What Microsoft’s “signal” is doing in this case is they are admitting that:

    1) Their crawlers aren’t smart enough to find the same pages on the Internet that other search engines might find, so they’ll monitor their users visiting these deeper pages to help them build their own directory.

    2) Rather than merely crawl this page to determine which keywords might apply, they will look at the specific keyword which was used on the other search engine which resulted in that search engine’s algorithm to determine relevancy.

    3) Their index can’t always determine which pages are more relevant, so they’ll just watch where people go and consider that relevant.

    I’m not suggesting that monitoring human behavior doesn’t result in relevancy, it’s just not innovative. It’s the technology from the 1990′s. C’mon, Microsoft, it’s time to out-innovate Google, not out-maneuver.

  • Bo. Xi. Mon.

    Microsoft’s “surfstream” tracking in IE8 works very simply by default via InPrivate Filtering, not (only) via Bing Toolbar /Suggested Sites! That means, IE8 tracks all the surfstreams as long as the user does not tick that box:

    “By default, InPrivate Filtering analyzes the websites you visit and the content providers they use,..”
    http://windows.microsoft.com/en-US/Windows7/InPrivate-frequently-asked-questions

  • http://sitesell.com Ken Evoy

    Danny, great post.

    I posted to Matt Cutts yesterday. It didn’t get published. Here’s what I said (paraphrasing, since I didn’t keep the original)…

    Bing has every right to track a competitor, especially a competitor it believes is superior. The Japanese did it when they got into the luxury car-making business (let’s leave illegal industrial espionage out of this discussion, we’re talking about information that is legally available).

    Tracking Google’s results, then comparing them to their own, enables Bing to measure deltas and presumably improve their own results over time IF an algorithm determines that a set of results from Google is superior.

    Naturally, this type of tracking will show up more obviously when returning results for outliers. I doubt if Bing could care less about the outliers.

    This has nothing to do with copying.

    On Matt’s issue about Bing not being forthright about what folks agree to when they ask for permission to track your clickstream, DOH? Like Google is? NO company is. It’s immature and disingenuous — the “I agree” rate would drop precipitously if every company gave a full disclosure (in a prominent location) of what they do with the information. THAT type of stuff is up to Congress, not Matt.

    But, now that Matt let that cat out of the bag, he should be asking for the same type of disclosure from Google for theirs, no?

    Google should have just thanked Bing for the compliment and moved on, trying to figure out ways to block or send misleading signals (if that’s at all possible). THAT would have been a nice PR win. Instead, they look afraid.

    The whole thing is pretty sad.

    All the best,
    Ken

  • http://www.johnbeagle.com John Beagle

    Nice work Danny Sullivan. I appreciate your keeping us informed on this topic. To ignore certain channel data, means you will never be the best.

  • incessant

    Hi Danny,

    I would just like to take a moment to applaud how I felt you’ve stayed fairly neutral while writing this article, even though Google was the only one to have originally approached you about the story. I am in no way saying you have any responsibility for this, but it is unfortunate that many news sites and writers have written about this in such a biased way, either intentionally or unintentionally, based on the original article and blog post by Google. I remember reading some articles by reputable news sites that took everything Google said for granted and reported it as fact, which will invariably lead to the vast majority of people who read those stories to take those reported words as fact.

    I don’t know who is “wrong” or “right” in this situation, but at the very least, I feel that Google’s side of the story is not entirely accurate and is misleading. I feel it is quite unfair for Bing’s team that most people will not take a moment to consider this possibility, though, and simply read the first news articles and move on. Thank you for continuing to cover the issue.

    Tony

  • PasserBy

    Let’s consider this in terms of intellectual property. When Bing uses clicks on the Google search it clearly uses Google’s intellectual property. Users just refine what Google offers them. So in a click there is a Google’s contribution and a user’s contribution. I would say that user’s contribution is fairly small compared to Google’s.

    So in my opinion it is stealing.

    Another thing is that Bing is probably excluding its own searches from the clickstream. Otherwise they would introduce a positive feedback since users tend to click on top links. So not all clicks are treated equally.

    And one more note: watch spammers exploit this feature. Bing will soon (if not already) be fighting with the influx of robot-generated clickstreams.

  • ML

    Google is really , really nervous…. I’ve used Bing and it is quite good and I switched to Bing about 8 months ago and may be once or twice went to Google search. Suddenly, MS share jumped from 12 to 29 with Yahoo which is not insignificant when you take the userbase into account – millions of millions users. There is some truth behind it and you cannot ignore it no matter which camp you are. Google is becoming, or better yet, already become, ME TOO, company and wants to be everywhere.
    Youtube-worlds number one pirated website. No one can get away with it except Google.
    Google Books – another “copied” – again “copied” books from libraries around the world without permission. Oh yes, if the book is yours, go write to Google to hide pages. Wonderful.

    Not sure if Google has “Do No Evil” motto anymore or if it cares.

  • Miguel Ferreyra

    Bing does not copy, is Microsoft Windows. There is a difference of what he says Google, why? Well, buy a netbook and puts Google as the default browser after opening Internet Explorer (any version of Windows) and obvious that the Google logo does not appear but is a custom OS for IE. Microsoft Windows.
    Another fact, because Microsoft does not give the data out of Yahoo! which in turn, these data are already in the DBA. of Google’s because it uses servers that are over Yahoo and Microsoft.

  • DavidStaver

    Danny,

    It seems like nobody has pointed out yet that Microsoft’s defense, while proving that they don’t outright copy Google’s results in particular, actually continues to implicate them. They admit that their search engine is taking Google’s results in to account, but somehow that is okay, because it’s also taking a peek at what users are clicking on at all of the other sites too.

    And personally I can say that a very large chunk of my browsing begins with a search query of some kind, be it Google, Wikipedia, shopping, or anything else I’m doing. I’m sure that a very sizable chunk of the data that comes in from the “clickstream” is based on searches.

    So really, not only is Bing mimicking Google’s results a little bit, but actually EVERY search engine’s results (A little bit). Albeit in a roundabout way, and officially accidentally.

    As an example, I will draw a similar analogy as a user above me. This is a situation where a professor gives an exam, and notices that one of the students got eerily similar examples to the star student sitting next to him.

    He confronts the student, and asks him if he copied his neighbor’s answers. The student explains that, well, that isn’t really true, because what really happened is this — he was glancing around the room and thinking about the questions, and accidentally saw everyone’s answers, and then averaged them against his own, to get his final result.

    While true that he did see his neighbor’s answers, it would be untrue to say that he outright copied them, because he really saw everybody’s papers, and really they were only a small part of what he took in to account when answering his questions.

    This really seems pretty clear to me. Bing is still copying search results. It’s an unintended result from what could be seen as a clever idea, but it’s still an issue.

    Am I way off here? Thanks.

  • HuskyFromAlaska

    @DavidStaver
    He confronts the student, and asks him if he copied his neighbor’s answers. The student explains that, well, that isn’t really true, because what really happened is this — he was glancing around the room and thinking about the questions, and accidentally saw everyone’s answers, and then averaged them against his own, to get his final result.

    This is not a good analogy, because in an examination a student is not supposed to glance around the room. Period. Full stop.

    A closer (though still not the perfect) analogy would be: this is a term project for a music-major where every student is supposed to compose a song. Every one can spend one month to compose the piece. A student, Page Gugle, starts to suspect the other student, Bing Bin, was copying his idea, so Gugle goes outside Bin’s dorm window, and for 20 evenings Gugle plays a special rhythm only Bin can hear, though Bin has no idea that what he hears is a specially designed rhythm that nobody else will compose. Of course in the meantime Bin also listens to all the other kinds of music.

    Later the professor collects each student’s work, and Gugle stands up, “Professor, Bin is copying my work! You see, I played 100 pieces outside his window, and you can find 7 rhythms in Bin’s work”. Bin would not (and cannot) deny that the 7 rhythms by Gugle show in his work, but Bin insists, this is not copying, that he did not know these rhythms were Gugle’s. “I just quiet down in the evening, thinking of the 1000 rhythms I experience every day, and compose my work. I never deliberately copy anyone’s work, especially Gugle’s.”

    Professor then asked, “in that case, why don’t you exclude anything related to Gugle?”

    Bin replied, “professor, I spent 64% of time every day in the same class as Gugle. You want me to ignore everything I hear during the 64% time?”

  • HuskyFromAlaska

    @PasserBy
    Let’s consider this in terms of intellectual property. When Bing uses clicks on the Google search it clearly uses Google’s intellectual property. Users just refine what Google offers them. So in a click there is a Google’s contribution and a user’s contribution. I would say that user’s contribution is fairly small compared to Google’s.
    So in my opinion it is stealing.

    Not at all. You can go to Google and search for the video/image/web page as you wish. How you click each result — you spend 15 seconds or 45 seconds before clicking a link, or you click “Translate this page”, etc, belongs to *YOU*, not Google. Therefore you can install Bing Toolbar and *OPT-IN* to provide your clickstream to Bing voluntarily. You do not need to ask permission from Google before you install Bing Toolbar because your clickstream does not belong to Google.

    @PasserBy
    Another thing is that Bing is probably excluding its own searches from the clickstream. Otherwise they would introduce a positive feedback since users tend to click on top links. So not all clicks are treated equally.

    That’s a speculation. I believe Danny has already explained. Say for one keyword, Bing gives a bad result with an irrelevant URL on top. 12% of the people using Bing skip over the top URL and click on the second or the third, so gradually Bing “learns” the top URL is ignored by most users, so the top URL will gradually fall down while the second and the third URL will float up.

    Note that Bing can do this even without IE Suggested Sites or Bing Toolbar, and Google is doing the same.

    @PasserBy
    And one more note: watch spammers exploit this feature. Bing will soon (if not already) be fighting with the influx of robot-generated clickstreams.

    Probably no. A smart spammer will invest his time to tweak search ranks on a search engine with 64% market share rather than the little guy with 12% market share, just like hackers target Windows rather than Mac OS X.

  • DavidStaver

    The point is that people are getting too caught up in the Google vs. Microsoft eDrama, and missing the bigger picture — Bing is using data from other search engines. Google were just the only ones to notice.

    They are passing it off as merely being data from user behavior, which is not untrue, but the result looks different. Bing isn’t aware of what it is doing, but when it looks at a search query, and then looks at where a user goes, it is also getting information about the results of the search query. Somebody could design a similar program to intentionally mimic the results of another search engine too. It would appear to be superior, but all it would really be doing is piggybacking on the original search engine’s code, and then prioritizing the results in a more useful manner.

    And in fact, Bing gets information from all of the search programs on the internet. Not just Google. I’m sure there is a programmer at Amazon.com that spends eight hours a day figuring out how to get people the products they want, and Bing is taking a peek at his search results too.

    I really doubt that this outcome was unforeseen by Microsoft, considering how much of browsing today involves search queries. And somebody had to sit down and write code that parsed a URL for keywords, and then made connections based on where the user went from there. The programmers had to realize that they might tap in to information from other search engines this way. It’s a pretty obvious conclusion to draw.

    That may be why the clickstream data is factored so low in to their ranking — they realized that the code was a little dangerous. It would be really easy to take it and just make a search engine that was designed to one up the rest of them. In fact, every time the other ones got better, yours would get better too.

    The question here isn’t whether or not they are using data from other search engines, but really whether it is ethical or not to do so. Basically, how much is too much?

    Personally I am inclined to refrain from taking sides on the issue, because of what’s already been reported. Google was just pretty alarmed and didn’t know what was going on, and Microsoft didn’t mean it to happen this way. But the fact remains that whether intentionally or not, Bing is still using search engine results from the rest of the web to better their own. I don’t find it to be infringement, but it’s still a little shady to me.

    And about the music analogy, when we are talking about copyright law, accidental copying is still copying. There is actually a very famous case involving a song written by George Harrison that was like this.

  • HuskyFromAlaska

    @DavidStaver
    The point is that people are getting too caught up in the Google vs. Microsoft eDrama, and missing the bigger picture — Bing is using data from other search engines. Google were just the only ones to notice.

    As far as I know Google and Bing are the only two major players who published toolbars to collect user clickstreams (not sure about Yahoo), and thus Google will cry foul.

    @DavidStaver
    They are passing it off as merely being data from user behavior, which is not untrue, but the result looks different. Bing isn’t aware of what it is doing, but when it looks at a search query, and then looks at where a user goes, it is also getting information about the results of the search query. Somebody could design a similar program to intentionally mimic the results of another search engine too. It would appear to be superior, but all it would really be doing is piggybacking on the original search engine’s code, and then prioritizing the results in a more useful manner.

    I do not know if you are a programmer, but maybe you can explain how to write a piggybacking search program to generate different results on the keyword “pontneddfechan” while showing a difference as Danny explained in his blog. Bing deliberately randomize the piggyback results a little so it does not look like copying?

    @DavidStaver
    And in fact, Bing gets information from all of the search programs on the internet. Not just Google. I’m sure there is a programmer at Amazon.com that spends eight hours a day figuring out how to get people the products they want, and Bing is taking a peek at his search results too.

    The difference is Amazon search engine can index pages on Amazon servers. Google cannot. Bing cannot. Google scans books and index the results. You think Google is so generous to allow other search engines to index the books she paid to get scanned?

    @DavidStaver
    I really doubt that this outcome was unforeseen by Microsoft, considering how much of browsing today involves search queries. And somebody had to sit down and write code that parsed a URL for keywords, and then made connections based on where the user went from there. The programmers had to realize that they might tap in to information from other search engines this way. It’s a pretty obvious conclusion to draw.

    Danny wrote, “Even if Bing excluded data from Google’s own sites, Google’s results from ISP partners, portal partners and others might still flow through.” Did you read Danny’s blog before presenting questions?

    @DavidStaver
    That may be why the clickstream data is factored so low in to their ranking — they realized that the code was a little dangerous. It would be really easy to take it and just make a search engine that was designed to one up the rest of them. In fact, every time the other ones got better, yours would get better too.

    You are once again speculating. How do you know the clickstream is factored low?
    ……

    @DavidStaver
    Personally I am inclined to refrain from taking sides on the issue, because of what’s already been reported. Google was just pretty alarmed and didn’t know what was going on, and Microsoft didn’t mean it to happen this way. But the fact remains that whether intentionally or not, Bing is still using search engine results from the rest of the web to better their own. I don’t find it to be infringement, but it’s still a little shady to me.

    It is fine if you feel it shady. Some people believe Google News show articles from news media all over the world without asking is shady too, and some news site even protested and asked Google to stop.

    @DavidStaver
    And about the music analogy, when we are talking about copyright law, accidental copying is still copying. There is actually a very famous case involving a song written by George Harrison that was like this.

    Google does not own anything shown on its search results, so please do not drag copyright laws here.

  • Registration000452

    Wakey, wakey, people!

    Microsoft gets to farm search information from every single search instance, in the entire world, that uses a PC IOR uses IE IOR uses “Bing”. (And it uses only about 8% of the instances! What fantastic economy! That’s only about 270,000,000,000/day!)

    That information includes not only which answer a user selected, but also what the answer itself was.

    Helloooo??!

    If I went and wrote a search engine, I would have to actually write a search engine that had some nous, and it would be only as good as I made it. As for result relevance: I could use all and only the information about what my clients clicked on — about which it is only a good guess, that it is actually what they really wanted.

    You can call it clever to leverage a monopoly position if you like… but this is not only generally considered to be immoral (by people who actually notice that it is happening), but also actually explicitly illegal. (Not that anything being illegal ever had any material impact on Microsoft.)

    Do I need to explicitly mention how much easier it is to write a complex algorithm when you can check your results against the right answer? (I’ll leave you to work out what the term “right answer” means, here.)