Beating a dead horse
This post is a (no doubt futile) response to Google engineer, PR drone wannabe and whiner extraordinaire Matt Cutt’s latest attempt to stoke the fire. The five fundamental flaws in his argument are as follows:
-
Data Ownership & Origin: Google’s results as published on the open web are to be cordoned off by Bing from its web click activity data collection system. Let alone that Bing isn’t even using Google’s results, only those misdirected links in the results which Google engineers purposely clicked on to inject fake info into Microsoft’s data collection system. Bing is using user activity data as their source, not the actual contents of Google results, hence the users have ownership of that data when they give it to Microsoft, and Google does not.
-
Bing’s algorithm: Assuming Bing is targeting and separating click data on Google results pages in the first place, asking Bing to disclose how that particular data is weighted among the 1000 signals Bing uses is like asking Bing to disclose their result-generation algorithm. Google guards its own algorithm as one of its most precious trade-secrets, and would never disclose it. This whole issue is supposed to be about Bing benefiting off of Google’s algorithm’s uniqueness improperly, according to Google. And they expect Bing to disclose their algorithm?
-
Reverse-engineering: The fact that there exists a research paper that shows how competing search-engines’ spell-check mechanisms can be reverse-engineered using URL patterns does not mean that this technique is actually being used by Bing today. Bing could clarify this point quickly.
-
Screenshots as evidence: Matt uses a few screenshots as evidence, without disclosing that only a fraction of these so-called “honeypots” actually affected Bing results in a discernible way and that by their very nature (being nonsense terms & being rare), the “honeypot” terms were more likely to affect them.
-
Respect & PR: Finally, Matt attempts to threaten Microsoft with embarrassment. He says that he respects the Bing engineers, yet his, Amit Singhal’s and Google’s behavior around dispersing their allegations as factual findings without having Microsoft & independent third-parties verify them speaks of nothing but childish behavior, contempt & perhaps jealousy & a desire for vindication.
Google’s timing of releasing this allegations was suspect at best (sitting on them for a few months, then releasing on the morning of the event), and Matt’s subsequent goading of Harry Shum into playing his dirty little game and attempt at misdirecting the focus of the event does not show any respect for the Bing people whatsoever. Thankfully, most of the independent bright minds and innovators in the room still discussed the future of search in a post-Google era.
The only valid point in the whole post is that both Microsoft & Google could make the possible uses of their data-collection systems more obvious to customers. How they do this without overloading the customer with info & driving them away from opting in to these systems is a rational conversation both companies can have.
