9 ms·

>Facebook tracks virtually every website you visit Is that true though? Many sites have Facebook share buttons that don’t actually load anything from Facebook

>Facebook tracks virtually every website you visit Is that true though? Many sites have Facebook share buttons that don’t actually load anything from Facebook unless you use them - they use a locally hosted button (I do this on my own sites, and got the idea from seeing it on so many other sites). I don’t see like buttons (which are directly loaded from Facebook) with anywhere near the frequency I once did either. In fact, the like button is only used on 0.5% of the world’s websites [1]. Even the Facebook pixel, a common target of the scorn of privacy advocates, is only in use on ~12% of the top 1 million sites [2]. 12% is hardly “virtually every website you visit”. It’s pretty alarmist to say that. [1] https://trends.builtwith.com/widgets/Facebook-Like-Button https://trends.builtwith.com/widgets/Facebook-Like-Button [2] https://trends.builtwith.com/analytics/Facebook-Pixel https://trends.builtwith.com/analytics/Facebook-Pixel

notyourwork 9y ago
I'd like to see traffic distribution against the sites which use facebook. I suspect your p90 for all traffic is against a large proportion of sites that use these and thus "virtually" every site in fact is closer to true than your raw numbers.
- downandout 9y ago
  Well that’s a bit of a self-fulfilling prophecy. Most sites that pay to advertise on Facebook are going to use the pixel for retargeting and conversion tracking. But only 25% of the Quantcast Top 10k sites use the pixel, and the number for the whole Internet is 0.4%. I wouldn’t say that anywhere near most of my browsing originates from Facebook ads...in fact I’m not entirely sure I have ever clicked on one except to view competitors’ landing pages. I’m guessing that most other people don’t limit the sites they visit to those they are shown in paid Facebook ads either. So the overall numbers are extremely relevant.
  - notyourwork 9y ago
    You missed my point. I’m suggesting that the 75th percentile of traffic goes to sites that probably have Facebook integrated via like button, ads, auth. (Way that fb can track you.) Therefore most traffic on internet is traceable. Key being traffic not count of sites in my analysis.
    - downandout 9y ago
      Well, that's what the Quantcast numbers are for...the top 10K sites are the top 10K trafficked sites in the world. And only about 25% of them have the Facebook Pixel.
      - notyourwork 9y ago
        Pixel isn't the only way to embed Facebook on the page mate. Facebook plugins and facebook auth also.
        downandout 9y ago
        I went through the stats for those above as well. I used the Pixel because it's by far more prevalent than any other Facebook-hosted stuff (plugins etc).
JumpCrisscross 9y ago
> Many sites have Facebook share buttons that don’t actually load anything from Facebook unless you use them - they use a locally hosted button This is incorrect. "When you load up your site with a host of sharing buttons you're – unwittingly perhaps – enabling those companies to track your visitors, whether they use the buttons and their accompanying social networks or not" [1]. > the Facebook pixel, a common target of the scorn of privacy advocates, is only in use on ~12% of the top 1 million sites So only tens of millions of people whose browsing history is being siphoned off without their consent? [1] https://www.wired.com/2013/03/social-sharing-buttons-that-respect-your-visitors-privacy/ https://www.wired.com/2013/03/social-sharing-buttons-that-re...
- downandout 9y ago
  >So only tens of millions of people whose browsing history is being siphoned off without their consent? If you have an issue with this, disable javascript and third party cookies on your browser - take some personal responsibility. Also, go setup a protest on Google’s front lawn, because Google Analytics (which also tracks you) is in use on nearly 75% of the Quantcast Top 100k sites [1]. That represents billions of people whose “browser history is being siphoned off with their consent” as you put it and over 6x the number of sites that the Facebook pixel is on. Since you seem to have a significant issue with Facebook, I’d suggest that you block all connections to them, which should solve your problem. Most firewalls, including the free Windows Firewall, enable you to easily block all connections to any root domain. You can configure your firewall to block *.Facebook.com for example, and then you will no longer have to worry about Zuck monitoring you. [1] https://trends.builtwith.com/analytics/Google-Analytics https://trends.builtwith.com/analytics/Google-Analytics
  - JumpCrisscross 9y ago
    > If you have an issue with this, disable javascript and cookies on your browser - take some personal responsibility This is a problematic attitude. It's the car dealer telling the lemon [1] buyer to "take personal responsibility" for being sold junk. The point of government is we don't stand alone. Facebook is a menace, and I block them. But I shouldn't have to. And neither should my mother. > it’s hypocritical to only have a problem with Facebook when Google is doing it on a scale that Facebook can only imagine Whatabout whatabout [2]? "You can't arrest me because there is another arsonist in town" isn't a valid excuse. (In any case, third parties calling out some, but not all, bad actors in a category isn't hypocrisy. It's prioritization. Hypocrisy would be Google calling out Facebook's advertising model, or an Enron executive criticising Facebook employees' complicity.) [1] https://en.wikipedia.org/wiki/Lemon_%28automobile%29 https://en.wikipedia.org/wiki/Lemon_%28automobile%29 [2] https://en.wikipedia.org/wiki/Whataboutism https://en.wikipedia.org/wiki/Whataboutism
extrapolate 9y ago
Here are some numbers [1]: > Over the last week ... the Like button appeared on 8.4M websites covering 2.6B webpages, the Share button on 931K websites covering 275M webpages, and there were 2.2M Facebook pixels installed on websites [1] https://twitter.com/facebook/status/985999292862152708 https://twitter.com/facebook/status/985999292862152708
- downandout 9y ago
  Yes, but that 8.4M websites is out of a universe of ~375 million websites - so it's a very small percentage (I am using the number of sites that BuiltWith tracks, which probably isn't all websites, but it's a minimum number). The pages number is meaningless, because sites that use it will use it on every page they have - if I have a site with 2 million auto-generated pages, then I account for 2 million of those web pages.
alanh 9y ago
If the denominator of that 12% number includes adult websites, which presumably know you don’t want to share to facebook, then this number may mean something a little different than many will assume it does
lozenge 9y ago
You won't see the locally hosted button suggested on Facebook's website, so it's far less common than the tracking version. 12% of websites can still reveal a lot about you - a few news sites and they'll know your political leaning and topics of interest. A recipe site and they could deduce your race. A lyrics site can indicate your gender, race and age. Etc.

notyourwork 9y ago

I'd like to see traffic distribution against the sites which use facebook. I suspect your p90 for all traffic is against a large proportion of sites that use these and thus "virtually" every site in fact is closer to true than your raw numbers.

downandout 9y ago
Well that’s a bit of a self-fulfilling prophecy. Most sites that pay to advertise on Facebook are going to use the pixel for retargeting and conversion tracking. But only 25% of the Quantcast Top 10k sites use the pixel, and the number for the whole Internet is 0.4%. I wouldn’t say that anywhere near most of my browsing originates from Facebook ads...in fact I’m not entirely sure I have ever clicked on one except to view competitors’ landing pages. I’m guessing that most other people don’t limit the sites they visit to those they are shown in paid Facebook ads either. So the overall numbers are extremely relevant.
- notyourwork 9y ago
  You missed my point. I’m suggesting that the 75th percentile of traffic goes to sites that probably have Facebook integrated via like button, ads, auth. (Way that fb can track you.) Therefore most traffic on internet is traceable. Key being traffic not count of sites in my analysis.
  - downandout 9y ago
    Well, that's what the Quantcast numbers are for...the top 10K sites are the top 10K trafficked sites in the world. And only about 25% of them have the Facebook Pixel.
    - notyourwork 9y ago
      Pixel isn't the only way to embed Facebook on the page mate. Facebook plugins and facebook auth also.
      - downandout 9y ago
        I went through the stats for those above as well. I used the Pixel because it's by far more prevalent than any other Facebook-hosted stuff (plugins etc).

JumpCrisscross 9y ago

> Many sites have Facebook share buttons that don’t actually load anything from Facebook unless you use them - they use a locally hosted button This is incorrect. "When you load up your site with a host of sharing buttons you're – unwittingly perhaps – enabling those companies to track your visitors, whether they use the buttons and their accompanying social networks or not" [1]. > the Facebook pixel, a common target of the scorn of privacy advocates, is only in use on ~12% of the top 1 million sites So only tens of millions of people whose browsing history is being siphoned off without their consent? [1] https://www.wired.com/2013/03/social-sharing-buttons-that-respect-your-visitors-privacy/ https://www.wired.com/2013/03/social-sharing-buttons-that-re...

downandout 9y ago
>So only tens of millions of people whose browsing history is being siphoned off without their consent? If you have an issue with this, disable javascript and third party cookies on your browser - take some personal responsibility. Also, go setup a protest on Google’s front lawn, because Google Analytics (which also tracks you) is in use on nearly 75% of the Quantcast Top 100k sites [1]. That represents billions of people whose “browser history is being siphoned off with their consent” as you put it and over 6x the number of sites that the Facebook pixel is on. Since you seem to have a significant issue with Facebook, I’d suggest that you block all connections to them, which should solve your problem. Most firewalls, including the free Windows Firewall, enable you to easily block all connections to any root domain. You can configure your firewall to block *.Facebook.com for example, and then you will no longer have to worry about Zuck monitoring you. [1] https://trends.builtwith.com/analytics/Google-Analytics https://trends.builtwith.com/analytics/Google-Analytics
- JumpCrisscross 9y ago
  > If you have an issue with this, disable javascript and cookies on your browser - take some personal responsibility This is a problematic attitude. It's the car dealer telling the lemon [1] buyer to "take personal responsibility" for being sold junk. The point of government is we don't stand alone. Facebook is a menace, and I block them. But I shouldn't have to. And neither should my mother. > it’s hypocritical to only have a problem with Facebook when Google is doing it on a scale that Facebook can only imagine Whatabout whatabout [2]? "You can't arrest me because there is another arsonist in town" isn't a valid excuse. (In any case, third parties calling out some, but not all, bad actors in a category isn't hypocrisy. It's prioritization. Hypocrisy would be Google calling out Facebook's advertising model, or an Enron executive criticising Facebook employees' complicity.) [1] https://en.wikipedia.org/wiki/Lemon_%28automobile%29 https://en.wikipedia.org/wiki/Lemon_%28automobile%29 [2] https://en.wikipedia.org/wiki/Whataboutism https://en.wikipedia.org/wiki/Whataboutism

extrapolate 9y ago

Here are some numbers [1]: > Over the last week ... the Like button appeared on 8.4M websites covering 2.6B webpages, the Share button on 931K websites covering 275M webpages, and there were 2.2M Facebook pixels installed on websites [1] https://twitter.com/facebook/status/985999292862152708 https://twitter.com/facebook/status/985999292862152708

downandout 9y ago
Yes, but that 8.4M websites is out of a universe of ~375 million websites - so it's a very small percentage (I am using the number of sites that BuiltWith tracks, which probably isn't all websites, but it's a minimum number). The pages number is meaningless, because sites that use it will use it on every page they have - if I have a site with 2 million auto-generated pages, then I account for 2 million of those web pages.

alanh 9y ago

If the denominator of that 12% number includes adult websites, which presumably know you don’t want to share to facebook, then this number may mean something a little different than many will assume it does

lozenge 9y ago

You won't see the locally hosted button suggested on Facebook's website, so it's far less common than the tracking version. 12% of websites can still reveal a lot about you - a few news sites and they'll know your political leaning and topics of interest. A recipe site and they could deduce your race. A lyrics site can indicate your gender, race and age. Etc.