Whoa. Shit just got real.

Google have a post on their blog entitled “Microsoft’s Bing uses Google search results - and denies it”. It’s a pretty interesting post that explains how Google set up search results for nonsense terms in Google, then repeatedly searched for them using IE8. After a week or two, Bing began returning the same results for these nonsense terms.

Microsoft responded with “We do not copy Google’s results” and then later with this post which implies that they sorta kinda do.

In Microsoft’s defence, and the point that everyone seems to be rallying around, is that the Bing toolbar simply, anonymously, collects habits about people’s browsing and uses that to build it’s search index. It just so happens that Google is in that mix of data.

On the face of it, it doesn’t seem so bad that Microsoft uses this data to build up a tapestry of information about the web, but there is one thing that sticks out for me in all of this.

Bing’s advertising campaign is centered around bringing the user meaningful results, illustrated by people spouting irrelevant fact-chains in the TV adverts. How, then, does that USP tally with a search engine that cannot spot a result that has zero relevance to the search term? The only link between Google’s bogus searches, and the resultant sites, was Google itself. And yet, that’s enough for Bing. It doesn’t do it’s own cross-checking with other results.

I wonder, how much of Bing is actually a search engine, and how much of it is reliant on users surfing the web. According to Microsoft, Bing uses over “1000 signals” to determine its results. This sounds like 1000 sources that Bing takes into account when deciding on what to show for a given search term. These 1000 signals could be anything, but it does make me wonder if each one is based on observation of users.

We just have to take Microsoft’s word that the data is used anonymously, and they are hardly alone in that kind of practice, but Microsoft’s response to this makes me wonder if Bing is anything other than a curator of the general public’s search habits. Does it do any crawling of it’s own? Does it analyse the data it gets, or is it purely weighted by clicks from users?

The fact that Google can just make something up and have Bing index it indicates that the latter may well be the case.

Either way it certainly makes Bing look bad. The discussion of whether or not this is an edge case won’t be as widespread as the shouty headlines about it.