Last week, we talked about allegations made by Google that Bing is stealing its search results. This doesn’t mean that Bing’s results are identical to Google’s – in fact, the overlap between first-place results in a test group was only about 9 %. However, something does seem to be going on, and Microsoft isn’t quite denying the charges. Here’s the statement that Microsoft issued after Google made its accusations:
We use over 1,000 different signals and features in our ranking algorithm. A small piece of that is clickstream data we get from some of our customers, who opt-in to sharing anonymous data as they navigate the web in order to help us improve the experience for all users.
Translated into plain English, here’s what’s probably happening.
As we outlined last week, Google created “honeypot” search terms to try to catch Bing in the act of copying search results. In this context, “honeypot” terms are nonsense words that didn’t actually appear in any websites (i.e. so overlap in results couldn’t be explained by good optimization practices) and that had never been searched for before. Next, Google manually assigned search results to those made-up search words. Then, Google’s programmers went to their home computers and started searching for those made-up terms and clicking on the results that appeared.
That’s the key to this little “sting” operation: when Google’s team searched for the honeypot terms at home, they made sure to use Internet Explorer, Microsoft’s Internet browser, and to make sure that the Bing toolbar was enabled.
The reason they did so is because they suspected that Microsoft was doing what it later confirmed in the statement quoted above: it was monitoring and recording “clickstream data” (i.e. a record of what users click on as they navigate around the web). That data almost certainly came from people who agreed to allow Microsoft to monitor what they do online. (If you said yes to “Turn on Suggested Sites” when you set up Internet Explorer on your computer, you too agreed to this.) If Microsoft users used Google search and clicked on Google results, that information was sent to Microsoft, which may have incorporated it into the Bing algorithm.
In some ways, this practice isn’t so far off from “retweeting”, in which content from one source is reproduced in another place by a user who thinks it’s good stuff. In the world of the web, this is generally considered to be a compliment – as long as the source is attributed. While not acknowledging Google’s contribution is definitely poor sportsmanship, it’s probably not illegal, since (based on the small percentage of overlap) Google’s input was probably only a small part of Bing’s algorithm, and therefore Bing’s use of Google results probably doesn’t quite violate intellectual property laws.
That being said, it’s pretty hypocritical that Bing’s “Decision Engine” advertising campaign tries to suggest that Bing is superior to the search engine from which it’s “borrowing” data.
However, if we worried about the ethics of every giant corporation does, none of us would have time for anything else. We at iMarket think we’ll best serve our clients by focusing on the practical: how will Bing’s likely use of Google in its search algorithm impact businesses that use the web to generate leads?
The answer is, it will probably benefit those businesses. The first benefit is that Bing seems to be becoming a better search engine. We’ve noticed some improvements recently, particularly in how fast Bing adds new websites to its search results. That’s been good news for our clients.
The second way businesses might benefit is that their good performance in Google may be rewarded with good performance in Bing as well. If you look at it as a numbers game, here’s the way it could work: Internet Explorer is used by approximately one-quarter of searchers, and Google is used by approximately two-thirds of searchers. For the sake of this exercise, let’s assume that half of the people who set up Internet Explorer on their computers opt in to Microsoft’s “Suggested Sites” click-tracking program and the Bing toolbar. Now, let’s do some down-and-dirty math: If you take half of two-thirds of one-quarter, you get about eight percent of all web searchers sending information about their Google results to Microsoft via Internet Explorer’s toolbar. (Interestingly, that’s about the percentage of overlap between Bing’s and Google’s results.)
So, in our view, the most important thing we can do in light of all this is to make sure that Internet Explorer users send some tasty “Google juice” over to Bing for our clients. We’ll do that by continuing to do what we’ve already been doing successfully: working hard to get strong placement for our clients in Google.
Of course, we’ve never ignored Bing in our optimization work. Next week we’ll talk about how Bing and Google differ and how we respond to those differences.