6 ms·

Google kills off FLoC, replaces it with Topics

62 points by adwmayer 5y ago

mrkramer 5y ago
Low key this can be opportunity for micropayments; if website knows your interests it can offer you to unlock some of its content for something like $0.20. I had similar idea like this but not for web advertising but for micropayments and content unlocking.
- acdha 5y ago
  My fear is that very few people are actually willing to make micropayments, and publishers will continue to build their decision around ad sales. It'd be nice if there was a very easy micropayment option but if it didn't see substantial adoption I'm not sure many sites would consider it.
  - mrkramer 5y ago
    I was thinking something like for example if my interests are Economics and I stumble upon paywalled article which is economic in nature website can say "This article matches your interest of Economics would you like to unlock it for $0.20?". I was aiming at paywalls because from my personal experience I bump into so many articles from variety of different websites that are behind paywalls that it is not feasible from consumer point of view to subscribe to dozens of publications when I read their articles every now and then and not all of them. So throwing a few cents(let's say from $0.05 to $0.20) for accessing an article wouldn't bother me. It would be something like PPV(pey per view) model. But now when I stumble upon an article which is behind paywall I can either choose to leave it or to view de facto pirated version on archive.is or on some other mirroring service.
  - GekkePrutser 5y ago
    Isn't this what Web3 is trying to introduce with integrated cryptowallets?
    - acdha 5y ago
      The problem is that it's really the other way around: what's being sold as web3 is a rebranding attempt to shed the bad reputation which cryptocurrencies developed and find new buyers. There are a number of hard problems in this space which are even harder to solve when your primary question is “how can I find someone to buy these otherwise worthless hashes?” rather than “what is a good solution for the user?”. In addition to the business model question about whether enough people would actually pay directly to make it worth building and participating in the system, there are also serious challenges around how to prevent blockchain privacy issues from turning it into a great tracking system, scaling the number of transactions the system is capable of up by many orders of magnitude, and dealing with various avenues for abuse.
    - mrkramer 5y ago
      Yea that was one of the intentions of Bitcoin.
- canadianfella 5y ago
- SahAssar 5y ago
  > if website knows your interests it can offer you to unlock some of its content for something like $0.20 Why would a website need to know my interests for me to be able to pay for content?
  - mrkramer 5y ago
    It is a UX trade-off; it's easier for the consumers to receive offers based on their preferences than it would be for them to seek out actively content that is interesting to them. My idea was when you are casually and randomly surfing the web you can receive offers to unlock some paywalled or premium content that would be appealing to you and otherwise locked and never accessed. The same concept is similar to advertising and particularly targeted advertising; why would advertisers spend millions and billions when consumers already know what they are interested in? It's not that easy because consumers experience Opportunity cost and Information asymmetry meaning they are simply not aware of everything that is available to them and when they are they do not quite know which product or service is better than the other and they have to choose atm which one to buy and which one they are not going to buy.
    - SahAssar 5y ago
      > My idea was when you are casually and randomly surfing the web you can receive offers to unlock some paywalled or premium content that would be appealing to you and otherwise locked and never accessed. I don't get the scenario, if I got linked a particular article then you already know what article I want to unlock. If I landed on the frontpage of the site then showing snippets/headlines/previews will guide me to what article I want to unlock. What is the scenario where there is no signal on what a visitor wants to read but that visitor would still be willing to pay for it?
      - mrkramer 5y ago
        The thing is now the only way to access that linked paywalled article is to subscribe to the publication, there in no way to pay to read only that particular article.
- rchaud 5y ago
  How exactly? The barrier to micropayments is friction. People don't want to have to put in a CC number or sign into Paypal to read a website. At the same time, publishers don't want to nickel and dime users, because the transaction costs of a $0.20 article aren't worth it. Google's changes don't affect either of these issues. The changes are purely about how audiences are targeted for Google ad products.
2Gkashmiri 5y ago
why should i, as a user be okay with tech like this? "For sites that it hasn’t categorized before, a lightweight machine learning algorithm in the browser will take over and provide an estimated topic based on the name of the domain." yeah right
- oaiey 5y ago
  If it fully works, it is at least under my control what data I handover to the advertisers. So it is better. I just think that it will not achieve the endgoal because every user will just deactivate it and the ad business will find another way to track me.
  - lopis 5y ago
    You're assuming advertisers will just give up fingerprinting technology? This tech sounds acceptable in a vacuum, but in the end it's just one more neat data point.
    - Macha 5y ago
      Or that advertisers will trust this data is another point. A lot of the bloat is from companies not trusting each other so every step along the chain including multiple libraries to check what's going on in all the other steps and fingerprint likely bots or fraud
    - oaiey 5y ago
      I am not but I was daydreaming.
  - dbbk 5y ago
    > I just think that it will not achieve the endgoal because every user will just deactivate it Most mainstream people don't care.
    - hammyhavoc 5y ago
      Source?
ohgodplsno 5y ago
Oh look, Google is looking for another way to fuck you over when using their spyware.
- iamstupidsimple 5y ago
  People generally aren't antsy about targeted ads but data collection and tracking. This avoids that.
cute_boi 5y ago
Haha these big techs shenanigans always intrigues me. First facebook tried to whitewash their tarnished reputation using buzz word "Meta". And now Google is trying to rename FLoC with topics. It would be so weird to say "Topic" is violating the privacy. Same snake with different color.
- mananaysiempre 5y ago
  > Same snake with different color. Yes and no. Yes, it still amounts to building user-tracking functionality into user agents. No, it’s not the same approach, because (IIUC) FLoC was doing unsupervised clustering of users, whereas Topics allocates users into predetermined clusters. This makes it potentially more transparent and also seems designed to work around the “protected categories” objection to FLoC, although in a world of proxy metrics and indicators I’m doubtful.
gkbrk 5y ago
Google didn't really kill off FLoC, but rather no other browser implemented such a user-hostile feature. Hard to sell ad targeting to your customers if iPhone users or Firefox users or Safari users can't be targeted.
- jjulius 5y ago
  Forgive my ignorance, but how does Topics differ from FLoC in that regard? It seems like Topics will also be in-browser, but perhaps I'm misunderstanding.
  - gkbrk 5y ago
    It doesn't differ too much in a meaningful way. Most likely it will end up the same way, pretty much all other browsers will either 1. Not implement it at all and depend on web developers having heard of error handling or feature detection 2. Implement it as a stub and return no data 3. Implement it to return random / meaningless data.
    - jjulius 5y ago
      So I guess where my brain is going with this is... If Google is aware that no other browsers were going to implement FLoC and that's why they're putting the kibosh on it, why would their alternative then be just another system that those same browsers wouldn't implement? Maybe this is more of a rhetorical question, since others in this thread have suggested all this amounts to is a name change? I dunno, this all just seems so... dumb.
      - GekkePrutser 5y ago
        I think the answer to this is legislation. Tracking cookies are really frowned upon in the EU and more legislation is incoming. I think Google is protecting their business model of targeted ads and perhaps also as a reason to lobby for tracking ("look, we tried to propose alternatives but nobody adopted it so we need tracking"). I personally don't like targeted ads even if I am anonymous so I'm not happy with the changes still. If there's an option for generic ads I'll always choose it.
mchusma 5y ago
I like this proposal overall. I like generally personalized ads, so this seems like a good way to give users more control over what they see.
popfastful 5y ago

62 points by adwmayer 5y ago

Low key this can be opportunity for micropayments; if website knows your interests it can offer you to unlock some of its content for something like $0.20. I had similar idea like this but not for web advertising but for micropayments and content unlocking.

acdha 5y ago
My fear is that very few people are actually willing to make micropayments, and publishers will continue to build their decision around ad sales. It'd be nice if there was a very easy micropayment option but if it didn't see substantial adoption I'm not sure many sites would consider it.
- mrkramer 5y ago
  I was thinking something like for example if my interests are Economics and I stumble upon paywalled article which is economic in nature website can say "This article matches your interest of Economics would you like to unlock it for $0.20?". I was aiming at paywalls because from my personal experience I bump into so many articles from variety of different websites that are behind paywalls that it is not feasible from consumer point of view to subscribe to dozens of publications when I read their articles every now and then and not all of them. So throwing a few cents(let's say from $0.05 to $0.20) for accessing an article wouldn't bother me. It would be something like PPV(pey per view) model. But now when I stumble upon an article which is behind paywall I can either choose to leave it or to view de facto pirated version on archive.is or on some other mirroring service.
- GekkePrutser 5y ago
  Isn't this what Web3 is trying to introduce with integrated cryptowallets?
  - acdha 5y ago
    The problem is that it's really the other way around: what's being sold as web3 is a rebranding attempt to shed the bad reputation which cryptocurrencies developed and find new buyers. There are a number of hard problems in this space which are even harder to solve when your primary question is “how can I find someone to buy these otherwise worthless hashes?” rather than “what is a good solution for the user?”. In addition to the business model question about whether enough people would actually pay directly to make it worth building and participating in the system, there are also serious challenges around how to prevent blockchain privacy issues from turning it into a great tracking system, scaling the number of transactions the system is capable of up by many orders of magnitude, and dealing with various avenues for abuse.
  - mrkramer 5y ago
    Yea that was one of the intentions of Bitcoin.
canadianfella 5y ago
SahAssar 5y ago
> if website knows your interests it can offer you to unlock some of its content for something like $0.20 Why would a website need to know my interests for me to be able to pay for content?
- mrkramer 5y ago
  It is a UX trade-off; it's easier for the consumers to receive offers based on their preferences than it would be for them to seek out actively content that is interesting to them. My idea was when you are casually and randomly surfing the web you can receive offers to unlock some paywalled or premium content that would be appealing to you and otherwise locked and never accessed. The same concept is similar to advertising and particularly targeted advertising; why would advertisers spend millions and billions when consumers already know what they are interested in? It's not that easy because consumers experience Opportunity cost and Information asymmetry meaning they are simply not aware of everything that is available to them and when they are they do not quite know which product or service is better than the other and they have to choose atm which one to buy and which one they are not going to buy.
  - SahAssar 5y ago
    > My idea was when you are casually and randomly surfing the web you can receive offers to unlock some paywalled or premium content that would be appealing to you and otherwise locked and never accessed. I don't get the scenario, if I got linked a particular article then you already know what article I want to unlock. If I landed on the frontpage of the site then showing snippets/headlines/previews will guide me to what article I want to unlock. What is the scenario where there is no signal on what a visitor wants to read but that visitor would still be willing to pay for it?
    - mrkramer 5y ago
      The thing is now the only way to access that linked paywalled article is to subscribe to the publication, there in no way to pay to read only that particular article.
rchaud 5y ago
How exactly? The barrier to micropayments is friction. People don't want to have to put in a CC number or sign into Paypal to read a website. At the same time, publishers don't want to nickel and dime users, because the transaction costs of a $0.20 article aren't worth it. Google's changes don't affect either of these issues. The changes are purely about how audiences are targeted for Google ad products.

2Gkashmiri 5y ago

why should i, as a user be okay with tech like this? "For sites that it hasn’t categorized before, a lightweight machine learning algorithm in the browser will take over and provide an estimated topic based on the name of the domain." yeah right

oaiey 5y ago
If it fully works, it is at least under my control what data I handover to the advertisers. So it is better. I just think that it will not achieve the endgoal because every user will just deactivate it and the ad business will find another way to track me.
- lopis 5y ago
  You're assuming advertisers will just give up fingerprinting technology? This tech sounds acceptable in a vacuum, but in the end it's just one more neat data point.
  - Macha 5y ago
    Or that advertisers will trust this data is another point. A lot of the bloat is from companies not trusting each other so every step along the chain including multiple libraries to check what's going on in all the other steps and fingerprint likely bots or fraud
  - oaiey 5y ago
    I am not but I was daydreaming.
- dbbk 5y ago
  > I just think that it will not achieve the endgoal because every user will just deactivate it Most mainstream people don't care.
  - hammyhavoc 5y ago
    Source?

ohgodplsno 5y ago

Oh look, Google is looking for another way to fuck you over when using their spyware.

iamstupidsimple 5y ago
People generally aren't antsy about targeted ads but data collection and tracking. This avoids that.

cute_boi 5y ago

Haha these big techs shenanigans always intrigues me. First facebook tried to whitewash their tarnished reputation using buzz word "Meta". And now Google is trying to rename FLoC with topics. It would be so weird to say "Topic" is violating the privacy. Same snake with different color.

mananaysiempre 5y ago
> Same snake with different color. Yes and no. Yes, it still amounts to building user-tracking functionality into user agents. No, it’s not the same approach, because (IIUC) FLoC was doing unsupervised clustering of users, whereas Topics allocates users into predetermined clusters. This makes it potentially more transparent and also seems designed to work around the “protected categories” objection to FLoC, although in a world of proxy metrics and indicators I’m doubtful.

gkbrk 5y ago

Google didn't really kill off FLoC, but rather no other browser implemented such a user-hostile feature. Hard to sell ad targeting to your customers if iPhone users or Firefox users or Safari users can't be targeted.

jjulius 5y ago
Forgive my ignorance, but how does Topics differ from FLoC in that regard? It seems like Topics will also be in-browser, but perhaps I'm misunderstanding.
- gkbrk 5y ago
  It doesn't differ too much in a meaningful way. Most likely it will end up the same way, pretty much all other browsers will either 1. Not implement it at all and depend on web developers having heard of error handling or feature detection 2. Implement it as a stub and return no data 3. Implement it to return random / meaningless data.
  - jjulius 5y ago
    So I guess where my brain is going with this is... If Google is aware that no other browsers were going to implement FLoC and that's why they're putting the kibosh on it, why would their alternative then be just another system that those same browsers wouldn't implement? Maybe this is more of a rhetorical question, since others in this thread have suggested all this amounts to is a name change? I dunno, this all just seems so... dumb.
    - GekkePrutser 5y ago
      I think the answer to this is legislation. Tracking cookies are really frowned upon in the EU and more legislation is incoming. I think Google is protecting their business model of targeted ads and perhaps also as a reason to lobby for tracking ("look, we tried to propose alternatives but nobody adopted it so we need tracking"). I personally don't like targeted ads even if I am anonymous so I'm not happy with the changes still. If there's an option for generic ads I'll always choose it.

mchusma 5y ago

I like this proposal overall. I like generally personalized ads, so this seems like a good way to give users more control over what they see.

popfastful 5y ago