Microsoft Vehemently Denies Google's "Bing Sting"

eldavojohn writes "In a blog post titled 'Setting the Record Straight,' Microsoft's senior vice president of online services, Yusuf Mehdi, addressed Google's 'Bing Sting' operation saying, 'We do not copy results from any of our competitors. Period. Full stop. We have some of the best minds in the world at work on search quality and relevance, and for a competitor to accuse any one of these people of such activity is just insulting.' Mehdi went on to claim that Google engaged in 'click fraud' in order to rig up their alleged 'experiment.' Mehdi added, 'That's right, the same type of attack employed by spammers on the web to trick consumers and produce bogus search results. What does all this cloak and dagger click fraud prove? Nothing anyone in the industry doesn't already know.' The struggle for Bing to usurp Google as number one in search continues."

Comments

Response from Another VP

[eldavojohn]

When I submitted the story, I had not noticed an additional response from Harry Shum, VP of Bing, who said:

To be clear, we learn from all of our customers. What we saw in today’s story was a spy-novelesque stunt to generate extreme outliers in tail query ranking. It was a creative tactic by a competitor, and we’ll take it as a back-handed compliment. But it doesn’t accurately portray how we use opt-in customer data as one of many inputs to help improve our user experience.

Apparently Google's accusations are viewed by some as a backhanded compliment.

Re:Response from Another VP

[Anonymous Coward]

Apparently Google's accusations are viewed by some as a backhanded compliment.

I don't thing they're wrong. I remember years and years ago how excited the Slashdot crowd got when Microsoft started directly addressing their superiority over Linux in their marketing propaganda. It meant Linux was enough of a threat that Microsoft was taking it seriously.

When was the last time you heard Google talk about other search engines? When it comes to searching, Google's been the undisputed market leader for a long time. For them to seriously acknowledge Bing, even if it's solely in the form of criticism, is still a big step.

Re:Response from Another VP

[LordLimecat]

Step 3: Microsoft gets the data and compares it against relevent information for that search term.

So what youre saying is, they ARE copying google, just with the user's opt-in consent? Ah, thats different then.

Re:Response from Another VP

[RobNich]

While Google search results look like links directly to their targets (because they are, right up until they are clicked), Google uses javascript to dynamically rewrite the link target to google URL which includes the target page URL and search terms, which is how Google tracks the click throughs (this Google page then redirects to the real target with a 302 response.)

Interesting theory, but demonstrably untrue. Install Live HTTP Headers and do a Google search, then click a result. There's no such redirect.

They track clicks of search results using Javascript, using the mousedown event on each search result link. There doesn't seem to be a server-side call, so they're probably setting a cookie with the click information and then reading the cookie later, when you return to Google.

I agree

[Haedrian]

The fact that microsoft technology has advanced to the point of linking

"delhipublicschool40 chdjob"
to a Credit Union website

is simply showing how well they understand their potential customers, and has nothing to do with the fact that Google set them up at all.

Re:I agree

[Xest]

I have to say that response made me chuckle

Feigned insult followed by a sleight of hand in trying to associate Google's research with spammers, fraudsters, and criminals.

What a terrible attempt at denial, it's not like they actually gave any evidence in their defence. They just pretended to be offended, and then tried to change the subject.

I'm usually quite supportive of Microsoft because I honestly believe some of their products (e.g. Visual Studio) are best of breed, but this is just a joke. They seem to have been caught red handed and have no idea how to deal with it, they'd have been better off just staying quiet and letting the story fade into obscurity than crying out like this without being able to offer the slightest bit of real actual defence such as an explanation of why they ended up with an obscure search term in their search results that Google had manufactured on their search engine.

Re:I agree

[KarmaMB84]

They did offer a defense: it's the customer data. What happens is even if that customer data is only weighted as 0.001% as important as their other metrics, if that customer data is the ONLY data they have for these bogus search terms, this would happen. Google used obviously bogus search terms which have exaggerated the weighting of that data. In reality, that data might only move a page up or down a ranking on page 10 of a real search on Bing for all we know.

Unless they come up with some actual evidence of real copying, this is a non-story. The #1 complaint around here all the time seems to be that Bing ISN'T giving the same results as Google so obviously that customer data isn't be weighted as important enough!

Re:I agree

[Haedrian]

Here's the thing.

Search engines work like this

Query ---> [S.E.] --> Results

A clever search engine (one with high precision and recall) will give you the results you want. If you click on a result, you are assumed to have found that link useful. Google does that on its own search engine - and that's fair enough.

Now if I capture "Query" and "Correct Result", I am basically using the other search engine's technology (which is used to supply that good result) and the result of all the data collection, research and whatever - in order to improve my search results. That's not a very fair game. If my search results heavily depend on Google's search results - I am piggybacking off them.

I've nothing wrong with Google or Bing reading my searches I input into them and improving their product that way - I don't think its fair if the other company steals this data off other search engines.

Re:I agree

[suomynonAyletamitlU]

Well, you'll have to make a distinction for me.

What Microsoft is tantamount to admitting is that "customer data" includes searches on a rival engine, and the relevant results. In other words, "When our competitor successfully finds a result for you, we want to know what it is." Clearly, Microsoft never asked Google if this was okay, or there would be no shock and no argument. Instead, they're using users of their opt-in program (henceforth known as mules) in a distributed effort to get a mapping of search queries to useful results. However, those useful results were generated thanks to Google's long-standing competence in the field, and not by ANY process Bing has a hand in. Therefore it is still to be argued that Bing is appropriating, without due request or apology, a mapping of google's results weighted by the relevance to users of google's site. So tell me--how is that not copying results?

Further, the mules in this attack are legitimate Google users who are acting on good faith. And indeed, perhaps the weighting on the algorithm is such that until or unless the weighting changes, this mapping does so little to Bing's results as to be utterly innocent. However, if Bing gained dominance (at the expense of Google) because of this mapping, or if for any other reason this caused Google's service to falter or become unprofitable, those users of Google's service will have unwittingly caused its downfall, and they caused that downfall by being satisfied with Google as a product.

I'm probably overstating it, but it still leaves a nasty taste in the mouth as far as I'm concerned. There's a difference in Bing's policy versus whether or not they're successful at it. If they consider it good policy to sit on the threshold of stealing someone else's results, but then simply not weighting those results highly enough to cause trouble, then I take issue with them. It remains their prerogative to explain themselves if they want to reverse that opinion.

Re:Ouch

[bluefoxlucid]

Well Microsoft's response is sort of self-incriminating really. I mean the summary here basically paints their response as simple posturing and trying to get out of getting caught doing something they're not supposed to. "How dare you! We're better than them! We're smarter than they are! Those people are just trying to make us look bad! That's it, THEY'RE cheating! They're rigging tests and accusing us of things! They're trying to make US look bad because THEY know we're BETTER and it gets their pants all in a knot! Why would WE ever do something like that?!"

We have some of the best minds in the world... after Google, who invented some truly creative and innovative search methods, and then patented them. We have to find a completely different direction that works the same way, kind of, then improve on it.

So the answer is yes

[Trelane]

They do spy on (sorry, gather 'click stream' data from) IE users (through IE itself, or one of its add-ons). Read those EULAs veeery carefully, folks!

Somehow this extremely relevant part of the story keeps getting skipped over whenever it's being told.

The 'click fraud' accusation is hilarious and quite arguably libelous as fraud (and click fraud) is a real criminal act.

http://en.wikipedia.org/wiki/Click_fraud
"Click fraud is a type of Internet crime that occurs in pay per click online advertising when a person, automated script or computer program imitates a legitimate user of a web browser clicking on an ad, for the purpose of generating a charge per click without having actual interest in the target of the ad's link. Click fraud is the subject of some controversy and increasing litigation due to the advertising networks being a key beneficiary of the fraud.

Use of a computer to commit this type of Internet fraud is a felony in many jurisdictions, for example, as covered by Penal code 502 in California, USA."
(also claimed to be a felony at http://www.legalmatch.com/law-library/article/click-fraud.html with claims of arrests.)

"Period. Full stop". stop. over. out.

[unity100]

If, he hasnt used 'period' and 'full stop' and created enough dramatic pause, i wouldnt have believed him.

but now, i believe him, despite bing has been caught red handed, denied it without showing ANY proof, and then went on to accuse google of something totally irrelevant.

Evidence and Explanation

[tzhuge]

Ok, I'm quite irked by this story, and I got modded troll a bunch of times by trying to point out that Google's experiment doesn't really support their accusation. I know some people will immediately label me a shill or apologist just for having a different opinion. What's stupid is I use Google search, and never Bing.

Anyways, the following is my understanding and some opinion. The secret knowledge of the search engine is the association of a search term and a result (usually a url). So to say that Bing is copying (I think 'cheating' might have the what was used, but copying is a lot of people's interpretation), implies they are acquiring Google's association data; conversely if the Bing search comes to the same result coincidentally, then they can't be 'cheating'. It wouldn't be that surprising if two search engines return same results for certain words. However, Google did their sting with fake terms... so obviously Bing is copying right?

So let's talk about their sting. They created (100?) honeypot search terms where a fake word would return a real link 'sss4yxyxy -> returns www.myresult.com'. Then they had 20 employees using IE and Bing toolbar w/ Google search and kept using these fake terms, then clicking the resulting link. Some time later, some of these fake terms return the same results on Bing.

A few things: Google employees opted into tracking w/ the Bing toolbar. (This is somewhat beside the point anyways, since Google isn't exactly in a position to point the finger about tracking.) Note that my understanding is that few of the (100?) honeypot terms actually worked on Bing.

The explanation from MS is that the Google employees gamed their user tracking mechanism to produce a result which makes it appear as if Bing is 'copying' Google. Basically they tracked the user search term, then the link they clicked through, and used this as part of the data for Bing. Google successfully gamed this because those terms are fake, and therefore the only data about them came from the sting.

So my opinion is that this isn't copying. If 100 of 100 honeypots showed up on Bing then that would support their accusation better. If their 20 employees only used Google normally from IE, without going through the toolbar, then that would strengthen the case. Without these, I have a hard time understanding how even the people at Google have rationalized their own accusation. Now maybe MS is lying and I'm a chump, but at least I'm taking the time to consider the evidence as presented.

Re:Evidence and Explanation

[tzhuge]

Sigh... they are acquiring association data from the tracked users. These fake users entered 'delhipublicschool40 chdjob' into the Bing search bar, then clicked on a link to 'a Credit Union website'. If they were copying directly from Google, then 100% of honeypot search terms should have worked...

It's not like that explanation even makes MS look good per se, but I'm almost guaranteed to get modded down again.

Diversion

[QuietLagoon]

We do not copy results from any of our competitors. Period. Full stop.

Parsing that carefully, that is not what Microsoft was accused of. So, in effect, Microsoft is saying that they did not do something that they were not accused of.

It's Microsoft's typical tactic, try to move the discussion over to a slightly different topic when Microsoft is caught with its hand in the cookie jar.

More disturbing for me during this whole mess is the fact that Microsoft is capturing my mouse clicks and visited links when I am using the browser, and sending that captured data back to Microsoft.