15 ms·

Given the attitude on HN about iOS14, I don’t think most here grasp the implication of gutting ad attribution. All publishers (news sites, indie games, etc) are

Given the attitude on HN about iOS14, I don’t think most here grasp the implication of gutting ad attribution. All publishers (news sites, indie games, etc) are going to see huge drops in CPM. It could easily be a business ending shift. Sure removing IDFA will help protect user privacy, but let’s not kid ourselves that there will not be collateral damage. Being a third party publisher will be that much harder and walled gardens (FB, Google, Instagram) will be further cemented as the only viable business model.

Spivak 6y ago
If that was really the case then why are FB and Google fighting it so hard? Shouldn’t they want this then?
- gregorythecat 6y ago
  I believe it will still comprise a reduction in their revenue. The internet as a whole will be less profitable, and only the big players will be able to survive. Maybe with time there can be a shift in the economics, but I would not bet on it. I can’t imagine people will start paying for email, search, music streaming...
  - fuzzer37 6y ago
    Or we'll shift back to an internet where people make things because they enjoy making them, not because they're trying to get a million clicks on blog spam. I'd be happy if the internet as a whole was a less profitable place.
    - gregorythecat 6y ago
      Don’t assume I’m against that from the perspective of a netizen. I too grew up reading slashdot and idolizing the FSF. I’m just pointing out this well could be the death knell of SWE as we know it. I will definitely miss the ridiculous TC fueled by the ad-laden web, but at least maybe I won’t work alongside physics and math phds on optimizing clicks....
  - akmarinov 6y ago
    Don’t people already pay for music streaming?
  - Nextgrid 6y ago
    > I can’t imagine people will start paying for email, search, music streaming... There's absolutely now reason why these services can't all fit in a 5-10$/month allowance. E-mail is already ~2$/month with Office 365, and if everyone starts paying economies of scale would make even cheaper plans profitable. Some of these services can just as well be provided by the ISP, like e-mail was back in the days. A lot of ISPs are indeed scum but there isn't technically anything forcing the users to use them (you can pay for a third-party) and I think the move away from e-mail being centralized around a few giants like Google (GMail), Microsoft (Outlook), etc would be a good thing.
    - gregorythecat 6y ago
      FB’s ARPU in US in 2019 was $112. Google’s was $256. By comparison, paid services would cost $30/month. You might be willing to pay that, but many will not. People already share Netflix accounts or bootleg their friend’s HBO accounts. I think a shift to paid services is _possible_ but I won’t bet on it. The economics and market forces that brought us metered data and text usage haven’t really changed. Destroying the advertising-supported offerings will not spawn competition overnight. But in any case I think the big players are positioned to weather the changes. And by the way, $30/month/user is an average. If you are have high disposable income you are likely netting FB et al more than that.
      - Nextgrid 6y ago
        How much of that ARPU is covering overheads of the advertising model (all the infrastructure and middlemen) as opposed to being actual money spent by brands on customer acquisition? If ads didn't exist and people wanted to pay for stuff, wouldn't brands be able to offer lower prices because of easier customer acquisition (not having to pay for ads) and thus people wouldn't need to share accounts if the account costs a couple of bucks a month?
- ocdtrekkie 6y ago
  The claim regulating the big players will actually hurt the little guys is the common cry of the big player hoping they don't actually end up getting regulated. As your comment points out, it really doesn't hold up to a cursory examination of what's being claimed and how they stand to gain. ;)
K0nserv 6y ago
Apple added SKAdnetwork specifically for the attribution use case, presumably because they foresaw this.
- gregorythecat 6y ago
  SKAdNetwork is severely restrictive. In my group we’re still forecasting a major hit to our ability to perform attribution or lift analysis.
  - K0nserv 6y ago
    It's worth it in the balance of user privacy and how accurate attributions are imo. Apple are tilting the balance back in favour of users.
  - Nextgrid 6y ago
    Why can't attribution be solved by including a campaign ID in the links to the product the ad is selling, and then counting the number of purchases per campaign ID? Is it because it's a manufactured problem by this scummy industry where they try to consider anything a "conversion" like "we showed this guy an ad 2 years ago and now he finally happened to look the right way at our client's shop while walking past it, this is definitely because of our ad and we should include this in our conversion stats"?
    - dialtone 6y ago
      There's a limit to how many campaigns you can be a part of as well since one could display an arbitrary set of campaigns to you and understand who they targeted once again. In any case "show this guy an ad" like you quote, won't leave signs or campaign IDs in the apple framework unless there's a way to track that user explicitly. And tagging people that viewed ads is not anywhere in the scheme for apple. As an aside, attribution of online ads is vastly more sophisticated than you paint it and it involve statistical and causal models. https://dl.acm.org/doi/abs/10.1145/2020408.2020453 https://dl.acm.org/doi/abs/10.1145/2020408.2020453 https://pdfs.semanticscholar.org/eaa4/875ae2376d681e156dcc3bc15acce89e08bb.pdf https://pdfs.semanticscholar.org/eaa4/875ae2376d681e156dcc3b... https://research.netflix.com/business-area/marketing-and-growth https://research.netflix.com/business-area/marketing-and-gro...
skratlo 6y ago
> Being a third party publisher And what is it that you publish that helps humanity?
danShumway 6y ago
> I don’t think most here grasp the implication of gutting ad attribution. > All publishers (news sites, indie games, etc) are going to see huge drops in CPM. It could easily be a business ending shift. Your second statement might not imply the first. It could be that publishers are going to see business disrupting shifts in CPM, and we are aware of that fact, and still support the changes. "Why is everyone praising this? Don't they realize that it will make it harder to track them?" At the same time, there's something really telling that after all of the years of the advertising industry talking about how users actually want to be tracked and targeted, they are nonetheless terrified of something as simple as a prominent opt-out button. If advertisers sincerely believed that users wanted what they were offering, then they wouldn't view it as an existential threat for users to have the ability to turn the offer down. If the take is, "users having control over their own data will be devastating for the advertisers", then honestly that's kind of a big signal that the advertising industry is out of control and needs some business disrupting shifts.
- eitland 6y ago
  > At the same time, there's something really telling that after all of the years of the advertising industry talking about how users actually want to be tracked and targeted, they are nonetheless terrified of something as simple as a prominent opt-out button. I actually gave it a chance. The problem was during that time the ads (from Google) were more irrelevant. Back in the old days you could often see how ads were relevant to the topic of the web site you were visiting or the topic you searched for even if they weren't always relevant to me (e.g. selling expensive software while I was a broke student, living outside the US). The next years - sometimes after the Double Click acquisition it seems - I started seeing almost only irrelevant ads everywhere. I don't have anything against companies making money - in fact I liked it when Google were earning good money before Double Click. But when they add invasive tracking and the quality goes way down at the same time something has gone wrong.
  - bjelkeman-again 6y ago
    Then you don’t have to deny tracking. But for the rest of us who want to deny tracking I like this, collateral damage or not.
beervirus 6y ago
I get the implications. It's not that I'm indifferent to them, it's that I'm actively happy about them. I hope every company that makes a living by trying to spy on me goes out of business.
rodgerd 6y ago
If this was twitter, this would be where I reply with the gif of Donald Glover saying "good". Business models come and go. This one isn't as bad as slavery or child labour, but but I'd happily tap-dance on its grave.

Spivak 6y ago

If that was really the case then why are FB and Google fighting it so hard? Shouldn’t they want this then?

gregorythecat 6y ago
I believe it will still comprise a reduction in their revenue. The internet as a whole will be less profitable, and only the big players will be able to survive. Maybe with time there can be a shift in the economics, but I would not bet on it. I can’t imagine people will start paying for email, search, music streaming...
- fuzzer37 6y ago
  Or we'll shift back to an internet where people make things because they enjoy making them, not because they're trying to get a million clicks on blog spam. I'd be happy if the internet as a whole was a less profitable place.
  - gregorythecat 6y ago
    Don’t assume I’m against that from the perspective of a netizen. I too grew up reading slashdot and idolizing the FSF. I’m just pointing out this well could be the death knell of SWE as we know it. I will definitely miss the ridiculous TC fueled by the ad-laden web, but at least maybe I won’t work alongside physics and math phds on optimizing clicks....
- akmarinov 6y ago
  Don’t people already pay for music streaming?
- Nextgrid 6y ago
  > I can’t imagine people will start paying for email, search, music streaming... There's absolutely now reason why these services can't all fit in a 5-10$/month allowance. E-mail is already ~2$/month with Office 365, and if everyone starts paying economies of scale would make even cheaper plans profitable. Some of these services can just as well be provided by the ISP, like e-mail was back in the days. A lot of ISPs are indeed scum but there isn't technically anything forcing the users to use them (you can pay for a third-party) and I think the move away from e-mail being centralized around a few giants like Google (GMail), Microsoft (Outlook), etc would be a good thing.
  - gregorythecat 6y ago
    FB’s ARPU in US in 2019 was $112. Google’s was $256. By comparison, paid services would cost $30/month. You might be willing to pay that, but many will not. People already share Netflix accounts or bootleg their friend’s HBO accounts. I think a shift to paid services is _possible_ but I won’t bet on it. The economics and market forces that brought us metered data and text usage haven’t really changed. Destroying the advertising-supported offerings will not spawn competition overnight. But in any case I think the big players are positioned to weather the changes. And by the way, $30/month/user is an average. If you are have high disposable income you are likely netting FB et al more than that.
    - Nextgrid 6y ago
      How much of that ARPU is covering overheads of the advertising model (all the infrastructure and middlemen) as opposed to being actual money spent by brands on customer acquisition? If ads didn't exist and people wanted to pay for stuff, wouldn't brands be able to offer lower prices because of easier customer acquisition (not having to pay for ads) and thus people wouldn't need to share accounts if the account costs a couple of bucks a month?
ocdtrekkie 6y ago
The claim regulating the big players will actually hurt the little guys is the common cry of the big player hoping they don't actually end up getting regulated. As your comment points out, it really doesn't hold up to a cursory examination of what's being claimed and how they stand to gain. ;)

K0nserv 6y ago

Apple added SKAdnetwork specifically for the attribution use case, presumably because they foresaw this.

gregorythecat 6y ago
SKAdNetwork is severely restrictive. In my group we’re still forecasting a major hit to our ability to perform attribution or lift analysis.
- K0nserv 6y ago
  It's worth it in the balance of user privacy and how accurate attributions are imo. Apple are tilting the balance back in favour of users.
- Nextgrid 6y ago
  Why can't attribution be solved by including a campaign ID in the links to the product the ad is selling, and then counting the number of purchases per campaign ID? Is it because it's a manufactured problem by this scummy industry where they try to consider anything a "conversion" like "we showed this guy an ad 2 years ago and now he finally happened to look the right way at our client's shop while walking past it, this is definitely because of our ad and we should include this in our conversion stats"?
  - dialtone 6y ago
    There's a limit to how many campaigns you can be a part of as well since one could display an arbitrary set of campaigns to you and understand who they targeted once again. In any case "show this guy an ad" like you quote, won't leave signs or campaign IDs in the apple framework unless there's a way to track that user explicitly. And tagging people that viewed ads is not anywhere in the scheme for apple. As an aside, attribution of online ads is vastly more sophisticated than you paint it and it involve statistical and causal models. https://dl.acm.org/doi/abs/10.1145/2020408.2020453 https://dl.acm.org/doi/abs/10.1145/2020408.2020453 https://pdfs.semanticscholar.org/eaa4/875ae2376d681e156dcc3bc15acce89e08bb.pdf https://pdfs.semanticscholar.org/eaa4/875ae2376d681e156dcc3b... https://research.netflix.com/business-area/marketing-and-growth https://research.netflix.com/business-area/marketing-and-gro...

skratlo 6y ago

> Being a third party publisher And what is it that you publish that helps humanity?

danShumway 6y ago

> I don’t think most here grasp the implication of gutting ad attribution. > All publishers (news sites, indie games, etc) are going to see huge drops in CPM. It could easily be a business ending shift. Your second statement might not imply the first. It could be that publishers are going to see business disrupting shifts in CPM, and we are aware of that fact, and still support the changes. "Why is everyone praising this? Don't they realize that it will make it harder to track them?" At the same time, there's something really telling that after all of the years of the advertising industry talking about how users actually want to be tracked and targeted, they are nonetheless terrified of something as simple as a prominent opt-out button. If advertisers sincerely believed that users wanted what they were offering, then they wouldn't view it as an existential threat for users to have the ability to turn the offer down. If the take is, "users having control over their own data will be devastating for the advertisers", then honestly that's kind of a big signal that the advertising industry is out of control and needs some business disrupting shifts.

eitland 6y ago
> At the same time, there's something really telling that after all of the years of the advertising industry talking about how users actually want to be tracked and targeted, they are nonetheless terrified of something as simple as a prominent opt-out button. I actually gave it a chance. The problem was during that time the ads (from Google) were more irrelevant. Back in the old days you could often see how ads were relevant to the topic of the web site you were visiting or the topic you searched for even if they weren't always relevant to me (e.g. selling expensive software while I was a broke student, living outside the US). The next years - sometimes after the Double Click acquisition it seems - I started seeing almost only irrelevant ads everywhere. I don't have anything against companies making money - in fact I liked it when Google were earning good money before Double Click. But when they add invasive tracking and the quality goes way down at the same time something has gone wrong.
- bjelkeman-again 6y ago
  Then you don’t have to deny tracking. But for the rest of us who want to deny tracking I like this, collateral damage or not.

beervirus 6y ago

I get the implications. It's not that I'm indifferent to them, it's that I'm actively happy about them. I hope every company that makes a living by trying to spy on me goes out of business.

rodgerd 6y ago

If this was twitter, this would be where I reply with the gif of Donald Glover saying "good". Business models come and go. This one isn't as bad as slavery or child labour, but but I'd happily tap-dance on its grave.