3 ms·

I'm sorry but your essay didn't get much attention because it is not particularly written well. I'm not really sure what you are arguing here, except that Goog

I'm sorry but your essay didn't get much attention because it is not particularly written well. I'm not really sure what you are arguing here, except that Google is bad maybe? You aren't clearly arguing an specific thesis, and there are many non sequitur paragraphs that don't flow. Why are you talking about OpenJDK in an essay about Google Adsense? Or is the essay about Doubleclick BMP? If the point of this essay is to provide an overview of the area, I am not sure you are well educated in this area. For example, most of doubleclick publishers dont do retargeted ads. They tend to do BMP only. You didn't mention programmatic at all. You don't talk about ad networks. Also you don't mention that Google has done a lot in the area of malware both on the web in general, and specifically has spent a lot of resources in getting malware off the ad network. This makes you seem disingenuous, because while Google isn't a fully "neutral" actor, they are overall a good citizen in the ad space. Ad targeting isn't inherently evil, a lot of ad targeting is geographically based -- in fact many people would probably argue that showing geographically inappropriate ads is even worse -- others are consumer segment (eg: women/men/age). I would like to have a non-advertisement internet, but it appears that people are just unwilling to do so. Also we may be facing very real cognitive limits, studies on microtransaction fatigue have demonstrated that is not likely to be a successful mechanism for funding webpages -- people just can't make that many monetary decisions every day.

yuhong 8y ago
There are other known problems with the essay as well. For example, I mentioned storage costs but it turned out to be more complex than that. The point is not only to provide an overview though but more importantly to trace back the problems to Larry/Sergey, which is why it is so important. And yes I focus on ad tracking using cookies and nothing else. When I was writing the essay BTW, even tracing back when they began sharing the retargeting data took some time.
- ryanobjc 8y ago
  You're making an argument, but I think you take it for granted that everyone knows what "the problems" are. I'm still not sure what that is! Also, why is it important to trace the problems back to Larry/Sergey? What about the early engineers who worked on adsense? The early PMs? Surely they are just as culpable for "the problems" (which ones again?) as anyone else? I guess they aren't household names? Also, adsense/doubleclick revenue accounts for 14.9% of total company revenue as per the last 10-Q filed. So it's a misrepresentation to discuss adsense/doubleclick as anything other than an important, but not critical, part of Google. Google could, in theory, lose all that revenue and still remain in the black.
  - yuhong 8y ago
    There are also Google Analytics/Urchin though. Part of the point of the essay (and one of the hardest parts for me to write) is to figure out when they merged the tracking data after the acquisition. I would consider the DoubleClick one worse than AdSense BTW, which is part of the point of the essay.
    - ryanobjc 8y ago
      Except, you’re wrong. Ga data is never joined with other data sets.
      - yuhong 8y ago
        What I was referring to is the sharing of remarketing lists. One example: https://analytics.googleblog.com/2015/06/remarketing-lists-for-search-ads.html https://analytics.googleblog.com/2015/06/remarketing-lists-f... https://analytics.googleblog.com/2015/11/share-google-analytics-data-and.html https://analytics.googleblog.com/2015/11/share-google-analyt...
        ryanobjc 8y ago
        Well later in the essay you write that google analytics is merged with ad data in an unqualified statement. That’s what I was reacting to. Remarketing lists is a feature that would be google neutral. I could imagine such a feature working for self hosted analytics.
        yuhong 8y ago
        And I mentioned in the essay that "In many cases, advertisers managed “remarketing” lists of “anonymous” visitors that was being tracked by cookies from a central console without thinking of the privacy problems, treating visitors almost as numbers." I wonder if such a use would comply with the GDPR BTW.
mirimir 8y ago
> I would like to have a non-advertisement internet, but it appears that people are just unwilling to do so. Also we may be facing very real cognitive limits, studies on microtransaction fatigue have demonstrated that is not likely to be a successful mechanism for funding webpages -- people just can't make that many monetary decisions every day. There's no reason why users would need to take all those decisions individually. Just imagine adding user agents to the existing ad-placement system. By default, offer enough to win most bids. If some ad won, prompt the user to increase their bid. And by the way, page prices here would be more like $0.10 than $1.00 or more.

yuhong 8y ago

There are other known problems with the essay as well. For example, I mentioned storage costs but it turned out to be more complex than that. The point is not only to provide an overview though but more importantly to trace back the problems to Larry/Sergey, which is why it is so important. And yes I focus on ad tracking using cookies and nothing else. When I was writing the essay BTW, even tracing back when they began sharing the retargeting data took some time.

ryanobjc 8y ago
You're making an argument, but I think you take it for granted that everyone knows what "the problems" are. I'm still not sure what that is! Also, why is it important to trace the problems back to Larry/Sergey? What about the early engineers who worked on adsense? The early PMs? Surely they are just as culpable for "the problems" (which ones again?) as anyone else? I guess they aren't household names? Also, adsense/doubleclick revenue accounts for 14.9% of total company revenue as per the last 10-Q filed. So it's a misrepresentation to discuss adsense/doubleclick as anything other than an important, but not critical, part of Google. Google could, in theory, lose all that revenue and still remain in the black.
- yuhong 8y ago
  There are also Google Analytics/Urchin though. Part of the point of the essay (and one of the hardest parts for me to write) is to figure out when they merged the tracking data after the acquisition. I would consider the DoubleClick one worse than AdSense BTW, which is part of the point of the essay.
  - ryanobjc 8y ago
    Except, you’re wrong. Ga data is never joined with other data sets.
    - yuhong 8y ago
      What I was referring to is the sharing of remarketing lists. One example: https://analytics.googleblog.com/2015/06/remarketing-lists-for-search-ads.html https://analytics.googleblog.com/2015/06/remarketing-lists-f... https://analytics.googleblog.com/2015/11/share-google-analytics-data-and.html https://analytics.googleblog.com/2015/11/share-google-analyt...
      - ryanobjc 8y ago
        Well later in the essay you write that google analytics is merged with ad data in an unqualified statement. That’s what I was reacting to. Remarketing lists is a feature that would be google neutral. I could imagine such a feature working for self hosted analytics.
        yuhong 8y ago
        And I mentioned in the essay that "In many cases, advertisers managed “remarketing” lists of “anonymous” visitors that was being tracked by cookies from a central console without thinking of the privacy problems, treating visitors almost as numbers." I wonder if such a use would comply with the GDPR BTW.

mirimir 8y ago

> I would like to have a non-advertisement internet, but it appears that people are just unwilling to do so. Also we may be facing very real cognitive limits, studies on microtransaction fatigue have demonstrated that is not likely to be a successful mechanism for funding webpages -- people just can't make that many monetary decisions every day. There's no reason why users would need to take all those decisions individually. Just imagine adding user agents to the existing ad-placement system. By default, offer enough to win most bids. If some ad won, prompt the user to increase their bid. And by the way, page prices here would be more like $0.10 than $1.00 or more.