9 ms·

"A new report says that iOS 14 tracking changes could cost the advertising revenue billions if an early trend in spending decline continues." Does it not follo

"A new report says that iOS 14 tracking changes could cost the advertising revenue billions if an early trend in spending decline continues." Does it not follow then that advertisers are saving billions of dollars? Now, admittedly, the advertisers aren't going to be getting now what they used to get, but this Freakonomics Podcast episode, "Does Advertising Actually Work? (Part 2: Digital) (Ep. 441)" https://freakonomics.com/podcast/advertising-part-2/ https://freakonomics.com/podcast/advertising-part-2/ makes me wonder how much it'll really impact the people who pay for advertising.

bpodgursky 5y ago
> Does it not follow then that advertisers are saving billions of dollars? No — if the targeting is less effective, the advertisers are not getting the same incremental value from the ads. The ads are cheap, but draw in less total revenue for the company, saving nothing. The bid prices are just lower on untargeted ad slots, adjusting to the lower CPM / lower lift. Edit: Truly, the best part of HN are the Dunning-Kruger programmers who once read a substack post about advertising, and now believe they have deep hidden insight that a hundred thousand marketing professionals, running marketing campaigns as their job, every day, have not figured out — aha, internet advertising doesn't work!
- thehours 5y ago
  Given how rampant ad fraud is in programmatic, I wonder how much revenue actually drops for these companies, relative to the decreased ad spend. Uber’s [1] experience is one example: “We turned off two thirds of our ad spend – $100m out of annual spend of $150m – and basically saw no change in our number of rider app installs. What we saw is a lot of installs we thought had come through paid channels suddenly came through organic.” [1] https://mackgrenfell.com/blog/how-did-uber-waste-so-much-ad-money https://mackgrenfell.com/blog/how-did-uber-waste-so-much-ad-...
  - toast0 5y ago
    Even without ad fraud, you have things like user wants to install Uber, so searches for Uber in a web search. First link on the page is an ad to install Uber, maybe a couple more ads, then first organic search result is a link to the Uber home page, and maybe the second result is to install it. If the ad weren't there, the user would probably install organically, but since it's there and at the top of the page, it's likely to get clicked and cost money.
    - xuki 5y ago
      The ad in that case is to prevent Lyft from getting that ad spot. Fucked up, I know.
      - JohnJamesRambo 5y ago
        The google shakedown- "It'd be terrible if ya know Lyft, got that top spot..."
      - xxpor 5y ago
        How is that fucked up? There's a certain number of slots, and someone is going to fill them. It's like in a grocery store. If Kellogg's doesn't offer Kroger a price they like, they'll just stock General Mills instead... that's just business.
- chmod775 5y ago
  > No — if the targeting is less effective, the advertisers are not getting the same incremental value from the ads. I mean... No again. Kinda. Advertising isn't magically gonna make money appear in the pockets of your customers. Advertising influences how people distribute their income, influencing them this way or that (at least that's the goal). If advertising becomes worse for everyone, which it did, those weights are just scaled by a fixed amount, arriving at the same end result. The only winners are those who didn't advertise at all and only competed only on value. Those may see an uptick.
  - dividedbyzero 5y ago
    Not sure; targeted ads can work pretty well for small niche businesses, getting the same amount of conversions through untargeted ads may not be feasible. I could imagine those getting hit harder by this change than, say, Uber, so probably not quite the same end result.
    - chmod775 5y ago
      I could see that. But you can still effectively target niches with contextual ads.
      - dividedbyzero 5y ago
        Definitely. Is that something you can buy right now? I was under the impression there was little to nothing substantial in between untargeted and programmatic these days
        morelisp 5y ago
        We've been slowly refactoring our products to offer more contextual possibilities (building more aggregations by domains, keywords, etc. instead of audiences). No one is interested in buying.
        tormeh 5y ago
- danShumway 5y ago
  Maybe you're right. But if less advertising means that less money is going into businesses, then ads don't make products free, because they increase consumer spending and decrease the efficiency of consumer spending. If what you're saying is true, then advertising has a tangible cost for consumers. There are three possible theories about how advertising works: A) Advertising increases consumer spending and makes consumers spend dollars less efficiently based on attributes like branding that are unrelated to quality. In which case, all of this "advertising makes apps free" talk has been really deceptive, because consumers who are viewing ads are spending more money for worse/inflated products to compensate. In this world, it's worth asking what the costs of advertising are and whether we want it to be the primary way of funding the Internet. OR B) Consumer spending/efficiency doesn't change, consumers are smart enough to buy the best product anyway, and the only thing that changes is which companies get that money and have brand recognition. In which case, it starts to become difficult to argue that consumers opting out of ads they don't want to see is particularly harmful. In the world where advertising is just making people aware of new information and helping them make informed purchasing choices, then it's probably a good thing that across the board everyone is being forced to cut down on the pointless noise that amounts to companies trying to scream over each other and increasingly hyper-target people in creepy ways. If advertising does not increase consumer spending, and if everyone is reducing spending and effectiveness by the same amount on iOS, then this should be good for advertisers -- they'll spend less money on ads, and consumers will still spend the same amount of money themselves. The only reason that wouldn't be the case is if highly targeted ads actually changes consumer behavior and spending habits in meaningful ways -- but if we admit that, we need to admit that the advertising industry is a lot more problematic and a lot less "free" for consumers than it likes to claim. OR C) Advertising is a public good, and consumers actually want to know about these products, and ultimately people want targeted ads. But this is obviously not the case for the majority of advertising. If it was the case, than iOS's changes wouldn't be a big deal. If consumers want targeted ads, they have the option to keep allowing them in iOS. The whole reason we're in this situation is because consumers have overwhelmingly demonstrated that they do not want targeted ads. So at best, the advertising industry can position itself as a kind of metaphorical "vegetable", claiming that everyone hates them but ultimately the world would be better if more things were commercialized. But it's just not a realistic theory, the data is pretty clear that consumers don't like ads and it's very hard to argue that unrestrained commercialism and forced targeting is good for society. There are more efficient ways we could be doing this if our goal was really just to make it easier for people to compare products and learn about new solutions to their problems.
- stingraycharles 5y ago
  On the other hand, for the publishers this may be a net win. I’ve always argued that very narrow targeting mostly benefits the advertisers, and the publishers are left with a very small audience that yields a lot, and a very large audience that is difficult to sell. Of course, this may not be the case for Facebook et al, but it’s definitely the case for a lot of apps and websites that live off advertising. Now, without super narrow targeting, they are suddenly able to sell a lot more inventory.
  - gleb 5y ago
    Unlikely. You make more money by breaking a beef into multiple cuts for different customers than grinding the whole thing and selling burgers only.
    - thebean11 5y ago
      Assuming you can find a buyer for enough of those individual cuts yes. If most of them have no buyers you're better off selling burgers which I think was the point.
dillondoyle 5y ago
It annoys me beyond belief that every single digital ads HN thread this is the most upvoted hot take and it rarely comes from someone working in buy side let alone ads in general. Good or bad or neutral, FB's targeting specifically works amazingly well and better than anything else for many buyers. Especially for direct response ads. If you can't make money selling your product from digital ads, you're not saving dollars you're losing revenue. This is true especially for 'internet only' upstart products, for example everyone who gets stuck in for instance a $60 athletic short ad loop as an example!! Sometimes you hear examples here about a new product or kickstarter who pre-this type of digital would probably not be able to launch or it would be much harder. Brand advertising is different and targeting can definitely argue doesn't matter as much. But these changes has already changed how we spend money, and we're tiny in the grand scheme. The other point that always gets thrown out: It's also incredibly simple to measure effectiveness or lift. Staying with FB as an example, if you are selling a product it can be as simple as using a unique link for your FB ads. ROI = Purchases with that link / FB spend. FB also has a large set of conversion tracking JS and offline like loading your retail conversions to the platform to match. But given the new rules FB is now 'modeling' ROAS and it has decreased accuracy for us from something like previously capturing 80-95% of conversions to sometimes less than half. This makes a big difference in auto-optimizing on their end and thus our ROAS. And again sure, it gets complicated and definitely more snake oil to buy some 3rd party ad tech which claims to model across all your channels online and off. A lot of times that's gamed or just a bunch of BS. But another way to measure is say you're a larger brand buying TV, digital, OOH etc the works. You can simply control by making only one single change segregated to a control and treatment geo/group before and after, measure delta.
- rubyfan 5y ago
  I’m not deep in this space but was thinking something along the lines of your response. If I understand it right, advertisers will lose efficiency here but I’m skeptical that it’ll cost ad networks much because advertisers will still need to reach an audience but will now suddenly be less effective at targeting.
- nojito 5y ago
  >Good or bad or neutral, FB's targeting specifically works amazingly well and better than anything else for many buyers. Especially for direct response ads. A recent study says otherwise. https://pubsonline.informs.org/doi/pdf/10.1287/mksc.2019.1188 https://pubsonline.informs.org/doi/pdf/10.1287/mksc.2019.118... Do you have any studies supporting your claim that it "works amazingly well"?
yohannparis 5y ago
I agree with you, I hate when news report as "cost", a loss of revenue they never going to get is not a "cost". They just need to adjust to the market.
tormeh 5y ago
The blanket statement "Advertisement does not work", which I really hope is a strawman, is trivially wrong. If you start a new B2C business the only way to do inbound sales is advertisement, as by definition no one knows about your company/brand/products/services. Outbound sales is (usually) impossible in B2C for cost reasons. Almost every successful B2C company is evidence that advertisement works. There are some exceptions like banking and utilities, where the LTV of a retail customer is high enough to do outbound sales, but most B2C companies can't afford this.
- iwitaly 5y ago
  +1
- amelius 5y ago
  Only under the assumption that no alternatives to advertising exist. Ban advertising, and suddenly this statement is not true anymore: "If you start a new B2C business the only way to do inbound sales is advertisement, as by definition no one knows about your company/brand/products/services." ... for the simple reason that people will look for other means to find products online. For example non-biased search engines, product catalogues, review sites, etc. etc. And no, listing your product with objective info is not the same as advertising.
  - tgv 5y ago
    And then those other means will become commercially very interesting and you'll have to buy your way into them. But let's assume advertising doesn't work. Two companies start with the same product and same market at the same time. One advertises, the other doesn't. According to you, the former one should lose, because it wastes money on ads. I'm betting that (on average), the latter will lose, though. I know, it's a very black-and-white hypothetical situation.
    - nlh 5y ago
      Heh we wrote basically the same reply at exactly the same moment. Cheers :)

bpodgursky 5y ago

> Does it not follow then that advertisers are saving billions of dollars? No — if the targeting is less effective, the advertisers are not getting the same incremental value from the ads. The ads are cheap, but draw in less total revenue for the company, saving nothing. The bid prices are just lower on untargeted ad slots, adjusting to the lower CPM / lower lift. Edit: Truly, the best part of HN are the Dunning-Kruger programmers who once read a substack post about advertising, and now believe they have deep hidden insight that a hundred thousand marketing professionals, running marketing campaigns as their job, every day, have not figured out — aha, internet advertising doesn't work!

thehours 5y ago
Given how rampant ad fraud is in programmatic, I wonder how much revenue actually drops for these companies, relative to the decreased ad spend. Uber’s [1] experience is one example: “We turned off two thirds of our ad spend – $100m out of annual spend of $150m – and basically saw no change in our number of rider app installs. What we saw is a lot of installs we thought had come through paid channels suddenly came through organic.” [1] https://mackgrenfell.com/blog/how-did-uber-waste-so-much-ad-money https://mackgrenfell.com/blog/how-did-uber-waste-so-much-ad-...
- toast0 5y ago
  Even without ad fraud, you have things like user wants to install Uber, so searches for Uber in a web search. First link on the page is an ad to install Uber, maybe a couple more ads, then first organic search result is a link to the Uber home page, and maybe the second result is to install it. If the ad weren't there, the user would probably install organically, but since it's there and at the top of the page, it's likely to get clicked and cost money.
  - xuki 5y ago
    The ad in that case is to prevent Lyft from getting that ad spot. Fucked up, I know.
    - JohnJamesRambo 5y ago
      The google shakedown- "It'd be terrible if ya know Lyft, got that top spot..."
    - xxpor 5y ago
      How is that fucked up? There's a certain number of slots, and someone is going to fill them. It's like in a grocery store. If Kellogg's doesn't offer Kroger a price they like, they'll just stock General Mills instead... that's just business.
chmod775 5y ago
> No — if the targeting is less effective, the advertisers are not getting the same incremental value from the ads. I mean... No again. Kinda. Advertising isn't magically gonna make money appear in the pockets of your customers. Advertising influences how people distribute their income, influencing them this way or that (at least that's the goal). If advertising becomes worse for everyone, which it did, those weights are just scaled by a fixed amount, arriving at the same end result. The only winners are those who didn't advertise at all and only competed only on value. Those may see an uptick.
- dividedbyzero 5y ago
  Not sure; targeted ads can work pretty well for small niche businesses, getting the same amount of conversions through untargeted ads may not be feasible. I could imagine those getting hit harder by this change than, say, Uber, so probably not quite the same end result.
  - chmod775 5y ago
    I could see that. But you can still effectively target niches with contextual ads.
    - dividedbyzero 5y ago
      Definitely. Is that something you can buy right now? I was under the impression there was little to nothing substantial in between untargeted and programmatic these days
      - morelisp 5y ago
        We've been slowly refactoring our products to offer more contextual possibilities (building more aggregations by domains, keywords, etc. instead of audiences). No one is interested in buying.
        tormeh 5y ago
danShumway 5y ago
Maybe you're right. But if less advertising means that less money is going into businesses, then ads don't make products free, because they increase consumer spending and decrease the efficiency of consumer spending. If what you're saying is true, then advertising has a tangible cost for consumers. There are three possible theories about how advertising works: A) Advertising increases consumer spending and makes consumers spend dollars less efficiently based on attributes like branding that are unrelated to quality. In which case, all of this "advertising makes apps free" talk has been really deceptive, because consumers who are viewing ads are spending more money for worse/inflated products to compensate. In this world, it's worth asking what the costs of advertising are and whether we want it to be the primary way of funding the Internet. OR B) Consumer spending/efficiency doesn't change, consumers are smart enough to buy the best product anyway, and the only thing that changes is which companies get that money and have brand recognition. In which case, it starts to become difficult to argue that consumers opting out of ads they don't want to see is particularly harmful. In the world where advertising is just making people aware of new information and helping them make informed purchasing choices, then it's probably a good thing that across the board everyone is being forced to cut down on the pointless noise that amounts to companies trying to scream over each other and increasingly hyper-target people in creepy ways. If advertising does not increase consumer spending, and if everyone is reducing spending and effectiveness by the same amount on iOS, then this should be good for advertisers -- they'll spend less money on ads, and consumers will still spend the same amount of money themselves. The only reason that wouldn't be the case is if highly targeted ads actually changes consumer behavior and spending habits in meaningful ways -- but if we admit that, we need to admit that the advertising industry is a lot more problematic and a lot less "free" for consumers than it likes to claim. OR C) Advertising is a public good, and consumers actually want to know about these products, and ultimately people want targeted ads. But this is obviously not the case for the majority of advertising. If it was the case, than iOS's changes wouldn't be a big deal. If consumers want targeted ads, they have the option to keep allowing them in iOS. The whole reason we're in this situation is because consumers have overwhelmingly demonstrated that they do not want targeted ads. So at best, the advertising industry can position itself as a kind of metaphorical "vegetable", claiming that everyone hates them but ultimately the world would be better if more things were commercialized. But it's just not a realistic theory, the data is pretty clear that consumers don't like ads and it's very hard to argue that unrestrained commercialism and forced targeting is good for society. There are more efficient ways we could be doing this if our goal was really just to make it easier for people to compare products and learn about new solutions to their problems.
stingraycharles 5y ago
On the other hand, for the publishers this may be a net win. I’ve always argued that very narrow targeting mostly benefits the advertisers, and the publishers are left with a very small audience that yields a lot, and a very large audience that is difficult to sell. Of course, this may not be the case for Facebook et al, but it’s definitely the case for a lot of apps and websites that live off advertising. Now, without super narrow targeting, they are suddenly able to sell a lot more inventory.
- gleb 5y ago
  Unlikely. You make more money by breaking a beef into multiple cuts for different customers than grinding the whole thing and selling burgers only.
  - thebean11 5y ago
    Assuming you can find a buyer for enough of those individual cuts yes. If most of them have no buyers you're better off selling burgers which I think was the point.

dillondoyle 5y ago

It annoys me beyond belief that every single digital ads HN thread this is the most upvoted hot take and it rarely comes from someone working in buy side let alone ads in general. Good or bad or neutral, FB's targeting specifically works amazingly well and better than anything else for many buyers. Especially for direct response ads. If you can't make money selling your product from digital ads, you're not saving dollars you're losing revenue. This is true especially for 'internet only' upstart products, for example everyone who gets stuck in for instance a $60 athletic short ad loop as an example!! Sometimes you hear examples here about a new product or kickstarter who pre-this type of digital would probably not be able to launch or it would be much harder. Brand advertising is different and targeting can definitely argue doesn't matter as much. But these changes has already changed how we spend money, and we're tiny in the grand scheme. The other point that always gets thrown out: It's also incredibly simple to measure effectiveness or lift. Staying with FB as an example, if you are selling a product it can be as simple as using a unique link for your FB ads. ROI = Purchases with that link / FB spend. FB also has a large set of conversion tracking JS and offline like loading your retail conversions to the platform to match. But given the new rules FB is now 'modeling' ROAS and it has decreased accuracy for us from something like previously capturing 80-95% of conversions to sometimes less than half. This makes a big difference in auto-optimizing on their end and thus our ROAS. And again sure, it gets complicated and definitely more snake oil to buy some 3rd party ad tech which claims to model across all your channels online and off. A lot of times that's gamed or just a bunch of BS. But another way to measure is say you're a larger brand buying TV, digital, OOH etc the works. You can simply control by making only one single change segregated to a control and treatment geo/group before and after, measure delta.

rubyfan 5y ago
I’m not deep in this space but was thinking something along the lines of your response. If I understand it right, advertisers will lose efficiency here but I’m skeptical that it’ll cost ad networks much because advertisers will still need to reach an audience but will now suddenly be less effective at targeting.
nojito 5y ago
>Good or bad or neutral, FB's targeting specifically works amazingly well and better than anything else for many buyers. Especially for direct response ads. A recent study says otherwise. https://pubsonline.informs.org/doi/pdf/10.1287/mksc.2019.1188 https://pubsonline.informs.org/doi/pdf/10.1287/mksc.2019.118... Do you have any studies supporting your claim that it "works amazingly well"?

yohannparis 5y ago

I agree with you, I hate when news report as "cost", a loss of revenue they never going to get is not a "cost". They just need to adjust to the market.

tormeh 5y ago

The blanket statement "Advertisement does not work", which I really hope is a strawman, is trivially wrong. If you start a new B2C business the only way to do inbound sales is advertisement, as by definition no one knows about your company/brand/products/services. Outbound sales is (usually) impossible in B2C for cost reasons. Almost every successful B2C company is evidence that advertisement works. There are some exceptions like banking and utilities, where the LTV of a retail customer is high enough to do outbound sales, but most B2C companies can't afford this.

iwitaly 5y ago
+1
amelius 5y ago
Only under the assumption that no alternatives to advertising exist. Ban advertising, and suddenly this statement is not true anymore: "If you start a new B2C business the only way to do inbound sales is advertisement, as by definition no one knows about your company/brand/products/services." ... for the simple reason that people will look for other means to find products online. For example non-biased search engines, product catalogues, review sites, etc. etc. And no, listing your product with objective info is not the same as advertising.
- tgv 5y ago
  And then those other means will become commercially very interesting and you'll have to buy your way into them. But let's assume advertising doesn't work. Two companies start with the same product and same market at the same time. One advertises, the other doesn't. According to you, the former one should lose, because it wastes money on ads. I'm betting that (on average), the latter will lose, though. I know, it's a very black-and-white hypothetical situation.
  - nlh 5y ago
    Heh we wrote basically the same reply at exactly the same moment. Cheers :)