6 ms·

I don't want to go into much detail, but I work for a major company in this space and nost companies in the industry can track you with reasonable success even

I don't want to go into much detail, but I work for a major company in this space and nost companies in the industry can track you with reasonable success even if you are logged out over multiple devices. Your (approximate) location, browsing habits and patterns are good enough data to predict what kind of stuff you buy. If you want to not be tracked, turn off JavaScript for a start.

Left4Yee 7y ago
Turning off js in turn locks you out of normieweb completely. Google captchas dont work, fat sites break...
- mylons 7y ago
  it is how almost all of the data collection happens though
- cameronbrown 7y ago
  I've solved Google CAPTCHAs without JS, so I don't think it's a requirement.
  - pmoriarty 7y ago
    Really? Could you go in to detail on how you did this? I thought JS was an absolute requirement for them.
    - tempguy9999 7y ago
      HN had a google captcha it subjected me to for a while. It did not need JS[0] At the same time I have come across captchas that do need JS. I guess google offers both. [0] if it had, I would have stopped posting here as I never accept browser JS outside of a VM.
    - cameronbrown 7y ago
      Whenever I've solved them without JS, it reverts to standard HTML forms. Since it's in an iframe, when you click submit it doesn't reload the whole page, only the frame. When you've been validated it gives you a code which you can copy and paste into the host site. Note, I've only seen this on Cloudflare pages so far. Disclaimer: I work for Google, but not on reCAPTCHA.
      - misterdoubt 7y ago
        Every time I've tried one of these non-JS versions, it gives me no evidence that it's working or progressing and just keeps setting up new comparisons for me to copy and paste the code. So I get tired of doing that and stop using whatever website is demanding it.
- tempguy9999 7y ago
  Totally not true. It does limit you somewhat, but a large part of the web, in HN, slashdot, bbc, most news sites, TheRegister, wikipedia, most of stackexchange, and a shedload more all work fine.
- thosakwe 7y ago
  This is definitely true to an extent. However, NoScript allows you to whitelist domains, or temporarily "allow all" for a site. Sometimes it takes a little bit of trial and error, but for most sites, you can get full functionality with only first-party scripts and common CDN's (cdnjs, jsdeliver, etc). Fat sites do break though, and I don't think there's really any way around that other than temporary whitelisting. The Web is a lot faster for me now, though.
pavel_lishin 7y ago
> good enough data to predict what kind of stuff you buy. Then why are the ads that _do_ sneak through my adblockers or onto Instagram, etc., such hot, moist garbage? Is it just a lack of people wanting to advertise at my demographic?
- soumyadeb 7y ago
  Ad re-targeting (showing products you browsed) is generally cookie based I would think and adblockers block that. Anything other than re-targeting is anyway too broad and generally crap for me - even on Facebook which knows a lot I would think. Other data vendors are totally crap. You can check your BlueKai profile here https://datacloudoptout.oracle.com/registry/ https://datacloudoptout.oracle.com/registry/
- Joe8Bit 7y ago
  Several reasons: Many large, higher quality ad buyers (agencies and direct sold) can get enough return from non-greyarea inventory like this. There's no point in them taking (even a small) perceived risk of buying this grey market inventory. However, they likely do use this tracking data second or third hand via their third party data providers brokering and including it in larger data sets they use for cross site targeting via DMPs. Finally, there's simple economics: people who go through enough effort to block trackers and other things are (perceived) to be less likely to engage with ads (e.g. they see it as you do, when one 'slips through'). So there's no point in paying $$$ for those eyeballs. All of this in combination means you get low quality, barrel bottoms ads.
  - alexbanks 7y ago
    People forget (or just never knew) that tracking and ad-serving isn't necessarily about giving every screenviewer the best possible ads you can all the time. Inventory and return on ad spend are both big factors in determining where to spend. The tracking enables ad placement companies to know that even with the best possible ad (which cost inventory to place), specific users just will not engage. And if that's true, why spend that ad on that person?
soared 7y ago
> Your (approximate) location, browsing habits and patterns are good enough data to predict what kind of stuff you buy. Missing component here is "when compared against massive existing data sets". That info alone isn't very powerful.
jlmx 7y ago
Easier for some people but not for the masses. Not everyone will know how to do this.

Left4Yee 7y ago

Turning off js in turn locks you out of normieweb completely. Google captchas dont work, fat sites break...

mylons 7y ago
it is how almost all of the data collection happens though
cameronbrown 7y ago
I've solved Google CAPTCHAs without JS, so I don't think it's a requirement.
- pmoriarty 7y ago
  Really? Could you go in to detail on how you did this? I thought JS was an absolute requirement for them.
  - tempguy9999 7y ago
    HN had a google captcha it subjected me to for a while. It did not need JS[0] At the same time I have come across captchas that do need JS. I guess google offers both. [0] if it had, I would have stopped posting here as I never accept browser JS outside of a VM.
  - cameronbrown 7y ago
    Whenever I've solved them without JS, it reverts to standard HTML forms. Since it's in an iframe, when you click submit it doesn't reload the whole page, only the frame. When you've been validated it gives you a code which you can copy and paste into the host site. Note, I've only seen this on Cloudflare pages so far. Disclaimer: I work for Google, but not on reCAPTCHA.
    - misterdoubt 7y ago
      Every time I've tried one of these non-JS versions, it gives me no evidence that it's working or progressing and just keeps setting up new comparisons for me to copy and paste the code. So I get tired of doing that and stop using whatever website is demanding it.
tempguy9999 7y ago
Totally not true. It does limit you somewhat, but a large part of the web, in HN, slashdot, bbc, most news sites, TheRegister, wikipedia, most of stackexchange, and a shedload more all work fine.
thosakwe 7y ago
This is definitely true to an extent. However, NoScript allows you to whitelist domains, or temporarily "allow all" for a site. Sometimes it takes a little bit of trial and error, but for most sites, you can get full functionality with only first-party scripts and common CDN's (cdnjs, jsdeliver, etc). Fat sites do break though, and I don't think there's really any way around that other than temporary whitelisting. The Web is a lot faster for me now, though.

pavel_lishin 7y ago

> good enough data to predict what kind of stuff you buy. Then why are the ads that _do_ sneak through my adblockers or onto Instagram, etc., such hot, moist garbage? Is it just a lack of people wanting to advertise at my demographic?

soumyadeb 7y ago
Ad re-targeting (showing products you browsed) is generally cookie based I would think and adblockers block that. Anything other than re-targeting is anyway too broad and generally crap for me - even on Facebook which knows a lot I would think. Other data vendors are totally crap. You can check your BlueKai profile here https://datacloudoptout.oracle.com/registry/ https://datacloudoptout.oracle.com/registry/
Joe8Bit 7y ago
Several reasons: Many large, higher quality ad buyers (agencies and direct sold) can get enough return from non-greyarea inventory like this. There's no point in them taking (even a small) perceived risk of buying this grey market inventory. However, they likely do use this tracking data second or third hand via their third party data providers brokering and including it in larger data sets they use for cross site targeting via DMPs. Finally, there's simple economics: people who go through enough effort to block trackers and other things are (perceived) to be less likely to engage with ads (e.g. they see it as you do, when one 'slips through'). So there's no point in paying $$$ for those eyeballs. All of this in combination means you get low quality, barrel bottoms ads.
- alexbanks 7y ago
  People forget (or just never knew) that tracking and ad-serving isn't necessarily about giving every screenviewer the best possible ads you can all the time. Inventory and return on ad spend are both big factors in determining where to spend. The tracking enables ad placement companies to know that even with the best possible ad (which cost inventory to place), specific users just will not engage. And if that's true, why spend that ad on that person?

soared 7y ago

> Your (approximate) location, browsing habits and patterns are good enough data to predict what kind of stuff you buy. Missing component here is "when compared against massive existing data sets". That info alone isn't very powerful.

jlmx 7y ago

Easier for some people but not for the masses. Not everyone will know how to do this.