Google and Information: Another Aberration or Genuine Insight??
September 1, 2017
I read “Yes, Google Uses Its Power to Quash Ideas It Doesn’t Like—I Know Because It Happened to Me.” How many “damore” of these allegations, misunderstandings, and misinterpretations will flow into my monitoring systems? It appears that a person who once labored for Forbes, the capitalist tool, is combining memory, the methods of Malcolm Gladwell, and a surfboard ride on the anti-Google wave.
The write up recounts this recollection of conversations with marketing and PR people, allegedly real, live Googlers:
I asked the Google people if I understood correctly: If a publisher didn’t put a +1 button on the page, its search results would suffer? The answer was yes. After the meeting, I approached Google’s public relations team as a reporter, told them I’d been in the meeting, and asked if I understood correctly. The press office confirmed it, though they preferred to say the Plus button “influences the ranking.” They didn’t deny what their sales people told me: If you don’t feature the +1 button, your stories will be harder to find with Google. With that, I published a story headlined, “Stick Google Plus Buttons On Your Pages, Or Your Search Traffic Suffers,” that included bits of conversation from the meeting.
If accurate, the method of determining search results runs counter to the information I presented in the Google Legacy,* which I wrote in 2003. In that monograph, I tallied about 100 “signals” that Google used to provide data to its objective algorithm for determining the importance of hits in a results list.
As part of my research for that monograph, I read patent documents stuffed with interesting discussions of what was wrong with certain approaches to search and retrieval issues. (You can find some juicy factoids in the discussion of the background of an invention. The pre-2007 Google patent documents strike me as more informative than Google’s most recent patent documents, but that’s just my opinion.) I recall that Google went to great lengths to explain the objectivity of the methods. I pointed out that judgment was involved in Google’s ranking methods because humans selected which numerical recipes to use and what threshold settings to use for certain procedures. In my lectures about the exploitable “holes” in the most common numerical recipes used by Google and other, the machine-based methods could be fiddled. But overall, the Google was making clear that automation for cost reduction and efficiency was more important than human editorial fiddling.
If the statement extracted from the Gizmodo write up is accurate, Google seems to have machine-based methods, but these can be used by humans to add the lieutenant’s favorite foods to the unit’s backpacks.
The Gizmodo article reveals:
Google never challenged the accuracy of the reporting. Instead, a Google spokesperson told me that I needed to unpublish the story because the meeting had been confidential, and the information discussed there had been subject to a non-disclosure agreement between Google and Forbes. (I had signed no such agreement, hadn’t been told the meeting was confidential, and had identified myself as a journalist.) It escalated quickly from there. I was told by my higher-ups at Forbes that Google representatives called them saying that the article was problematic and had to come down. The implication was that it might have consequences for Forbes, a troubling possibility given how much traffic came through Google searches and Google News.
With this non algorithmic interaction, the Gizmodo story depicts Google as a frisky outfit indeed. The objective system can be punitive. Really?
When I step back from this bit of “real” reporting, enlivened with the immediacy of an anecdote which seems plausible, I am thinking about the disconnect between my analysis is the Google Legacy and the events in the Gizmodo story.
Several questions arise:
- If the story is accurate, how “correct” are other articles about Google? Perhaps Google influenced many stories so that the person doing research is working with a stacked deck?
- If I assume that my research was correct in the 2002 to 2003 period when I was actively compiling data for the Google Legacy, what has caused this “objective method” to morph into a tool suitable for intimidation? If the shift did happen, what management actions at Google allowed objective methods to relax their grip?
- Why, after 20 years, are “real” news organizations now running stories about Google’s power, its machinations, and the collateral damage from Google employees who are far removed from the messy cubicles and Foosball games among Google’s elite engineers? Hey, those smart people were the story. Now it is the behavior of sales and public relations types who are making news? What’s this say about “news”? What’s this say about Google?
My hunch is that a large, 20 year old company is very different from the outfit that hired folks from AltaVista, refugees from Bell Labs, and assorted wizards whose life’s work was of interest to 50 people at an ACM special interest group.
Perhaps the problem is a result of Google’s adding people with degrees in art history and political science? There may even be one or two failed middle school teachers among Google’s non technical staff. Imagine. Liberal arts or education majors in Google satellite offices. I can conjure a staff meeting which involves presentations with low contrast slides, not the wonky drawings that Jeff Dean once favored in his lectures about Big Table.
Google’s staffing has shifted over the years from 99 percent engineers and scientists to a more “balanced” blend of smart people. (I don’t want to say “watered down”, however.) One possibility is that these “stories” about the Google’s alleged punitive actions may be less about the Google technical system and methods and more about what happens when hiring policies change and the firms’ technical past is lost in the haze of success.
Could Google’s sales, marketing, and PR professionals, not the engineers and scientists, are the problem? The fix is easy. More math, more algorithms, more smart software. Does Google need staff who can be easily be categorized as “overhead”? I want to think about this question.
Stephen E Arnold, September 1, 2017
* If you want a pre publication copy of the Google Legacy from 2003, just write benkent2020 at yahoo dot com. Something can be worked out. Yes, this monograph still sells, just slowly.
Comments
2 Responses to “Google and Information: Another Aberration or Genuine Insight??”
-
- Subscribe to Beyond Search
Feature archive
News archive
-
may lam mat nao tot
ERROR: The requested URL could not be retrieved
Haller is chauffeured from courtroom to courtroom across Los Angeles by Earl (Lawrence Mason), a former client now offering his services in lieu of legal fees.