One of the long-standing dilemmas any search engine has to grapple with is how to rate their own products in search results. Various rumors have been circulating that a particular engine or another is playing favorites by giving a search ranking boost to their own products. Ben Edelman, an assistant professor at the Harvard Business School and search engine “bird-watcher” of sorts aims to shed some light on these rumors via his recently-published study. By comparing search engines’ results for a set list of terms, he looked for patterns that suggested bias. Some of his results can be seen in the table below.

Study Table 1
[Partial table of results from Edelman's study.]

Google and Yahoo! both showed evidence of favoring their respective products in results over those of the competition. Of the three major players, Bing was the only search engine who did not appear to rank their own products more frequently. Is this a case of ego or merely coincidental? Edelman, who it is important to note consulted for Microsoft and initiated a lawsuit against Google, is quick to single out Google as intentionally skewing these results.

The abstract really does not delve sufficiently into the possibility that the users of a particular search are more likely to use other services by that engine, thus making same-company products rise organically in search results. Furthermore, is a company promoting its own services ultimately a bad thing? Surely both Google and Yahoo! each believe that their email program is the best out there and therefore the most relevant to users. So, company ego or legitimate self-promotion? Sound off in the comments.

One Response to “Narcissus’ Search Engine: Are Engines Biased Towards Their Own Products?”

  1. KB says:

    I too feel that the study is inconclusive. I would think that terms that arn’t bread and butter for the company would be a better choice to search for. Example would be that if I were to search web browser on Google and get links back to chrome in the top sections that would be an issue. I would give this professor credence if the terms used had not been ambiguous at best to something that people use every day from the engine.

    As I just did this test myself on Google the top ad is indeed for Chrome. however the top ranked search result is firefox. Again picking on Google I tested “smartphone” and not one link on the page directed me to Android.

    To this professor I call shenanigans.

Leave a Reply



© 2011