After GDPR, The New York Times cut off ad exchanges and kept growing ad revenue • Hacker News Board

Re: After GDPR, The New York Times cut off ad exchanges and kept growing ad revenue

#51

Post by jarfil » Wed, Jan 16, 2019, 2:05 PM UTC

pytyper2 wrote: post #31

bitofhope wrote: post #6

I honestly would not be surprised if the difference in effectiveness of targeted vs untargeted advertisement turns out to be none, negligible or even unfavorable to targeted. Sure, NYT is not enough data to be significant, especially when compensating for my confirmation bias, but I would really like to see more sites stop tracking, at least from an ethical standpoint.

This makes sense to me. If you are targeting me, I have told you I like it, and therefore don't need you to tell me about it.

The worst kind of targeting is when I go to a website, browse some similar items comparing their features and price, buy one... and now you decide to show me ads about those same products for the next week even if it should be clear that I've already made up my mind (I'm looking at you, Amazon).

Re: After GDPR, The New York Times cut off ad exchanges and kept growing ad revenue

#52

Post by MarkMc » Wed, Jan 16, 2019, 2:05 PM UTC

bitofhope wrote: post #6

I honestly would not be surprised if the difference in effectiveness of targeted vs untargeted advertisement turns out to be none, negligible or even unfavorable to targeted. Sure, NYT is not enough data to be significant, especially when compensating for my confirmation bias, but I would really like to see more sites stop tracking, at least from an ethical standpoint.

I would be absolutely astounded if targeted ads did not provide significant long-term advantage to a big player like Google. How can it be that having intimate knowledge of someone would not allow you to sell them more stuff? John has a 10th wedding anniversary coming up. Mary is single and goes to the gym every day. Steve just got a promotion and likes the BMW 3 series. Michael is overweight but has just gone on a diet. Are such details of no significance determining what ads to show these people?

Having said that, I'm constantly surprised by how bad Google is at targeted advertising. For example, today when I visit nytimes.com I see an ad from Google with the ad text in French. Hey Google, despite my recent visit to Paris I don't speak French - maybe your AI experts could analyse my 13 years of Gmail and search history to figure that out!

PS: I'm a Google shareholder, so my confirmation bias is in the other direction :)

Re: After GDPR, The New York Times cut off ad exchanges and kept growing ad revenue

#53

Post by cyborgx7 » Wed, Jan 16, 2019, 2:07 PM UTC

throwawaymath wrote: post #47

darkpuma wrote: post #34

Earlier quoted context omitted.

Say what you want, but sweaters don't spy on people.

Then in a decade our new sardonic message board reference will be, "Yes, the software engineering industry cratered. But for a beautiful moment in time, we stopped serving ads." I don't think the parent commenter disagrees with you about spying. I think they're trying to highlight the fundamental tension of the is-ought problem at play here. Your normative claim is that it's good to limit activity which reduces consu…

How about we built a society in which people don't have to choose between loosing their livelyhood and being a moral hazard. In the meantime, I'm not going to cry about the deaths of immoral industries.

Re: After GDPR, The New York Times cut off ad exchanges and kept growing ad revenue

#54

Post by SilasX » Wed, Jan 16, 2019, 2:08 PM UTC

reilly3000 wrote: post #8

pjc50 wrote: post #3

This is the thing the ad industry really didn't want people to discover: what if the behavioural and personalised targeting wasn't actually worth the cost?

The ad industry mostly already knows that. I would pay $4-6 CPM to run a behavioral campaign that barely filled and rarely performed. It was almost always more efficient to be at $0.75 cpm with simple contextual targeting, ie. publisher reported content category against a tight site list. The data is pretty much shit, and the way its sourced is fully opaque to the advertiser. It was pulling teeth to even find a one-s…

Good points, but that doesn't mean targeting is ineffective, only that they currently charge too much for it (and some ad sellers hide behind undisclosed flaky algorithms for doing so).

I'd be interested in knowing if there are other effects: if you remove targeting, does that also mean you're turning off the extreme bloat on sites? If so, I can totally believe that costing you enough users to kill the value added by the ad precision (contra the conventional wisdom of "lol I'm only marketing to people with the latest hardware anyway").

Re: After GDPR, The New York Times cut off ad exchanges and kept growing ad revenue

#55

Post by wlesieutre » Wed, Jan 16, 2019, 2:08 PM UTC

jarfil wrote: post #51

pytyper2 wrote: post #31

Earlier quoted context omitted.

This makes sense to me. If you are targeting me, I have told you I like it, and therefore don't need you to tell me about it.

The worst kind of targeting is when I go to a website, browse some similar items comparing their features and price, buy one... and now you decide to show me ads about those same products for the next week even if it should be clear that I've already made up my mind (I'm looking at you, Amazon).

Customers who bought this television also bought:

    1) different television
    2) another television
    3) alternate listing of the same television and 4x the
       price because it's controlled by some idiot algorithm
    4) more television
    5) yet another television

Suuuuure they did

Re: After GDPR, The New York Times cut off ad exchanges and kept growing ad revenue

#56

Post by Nursie » Wed, Jan 16, 2019, 2:09 PM UTC

MarkMc wrote: post #52

bitofhope wrote: post #6

I honestly would not be surprised if the difference in effectiveness of targeted vs untargeted advertisement turns out to be none, negligible or even unfavorable to targeted. Sure, NYT is not enough data to be significant, especially when compensating for my confirmation bias, but I would really like to see more sites stop tracking, at least from an ethical standpoint.

I would be absolutely astounded if targeted ads did not provide significant long-term advantage to a big player like Google. How can it be that having intimate knowledge of someone would not allow you to sell them more stuff? John has a 10th wedding anniversary coming up. Mary is single and goes to the gym every day. Steve just got a promotion and likes the BMW 3 series. Michael is overweight but has just gone on a d…

>Are such details of no significance determining what ads to show these people?

Perhaps the question is whether that turns to sales and engagement, or if people who don't click ads don't click ads regardless of content, and those that do, do.

Re: After GDPR, The New York Times cut off ad exchanges and kept growing ad revenue

#57

Post by SiempreViernes » Wed, Jan 16, 2019, 2:09 PM UTC

nemothekid wrote: post #49

detaro wrote: post #37

Earlier quoted context omitted.

Why do weird niche websites need more tracking? I'd assume their user base is way more specific, and thus can be targeted better without needing tracking. That I visit the NYT tells you way less about my interests than that I'm reading your blog about aquarium filters.

I can only give you speculation on what I’ve seen in online video. If you are a smaller publisher, unless you get particularly active in courting advertisers (i.e. spend time going into sales), the niche advertisers will never even look at your content. The 20-something digital ad manager has no motivation to risk his job doing speculative ad spend. As a result your ads are less relevant and your inventory’s CPM drop…

If you are a small publisher, why would your ad prices endanger any but the smallest of advertising budgets?

Re: After GDPR, The New York Times cut off ad exchanges and kept growing ad revenue

#58

Post by baddox » Wed, Jan 16, 2019, 2:10 PM UTC

pytyper2 wrote: post #31

bitofhope wrote: post #6

I honestly would not be surprised if the difference in effectiveness of targeted vs untargeted advertisement turns out to be none, negligible or even unfavorable to targeted. Sure, NYT is not enough data to be significant, especially when compensating for my confirmation bias, but I would really like to see more sites stop tracking, at least from an ethical standpoint.

This makes sense to me. If you are targeting me, I have told you I like it, and therefore don't need you to tell me about it.

That sounds vastly oversimplified. So you’ve discovered that I’m into hiking/camping gear. That doesn’t mean I will never buy another piece of outdoor gear or that I have heard of every brand and product in that category.

Re: After GDPR, The New York Times cut off ad exchanges and kept growing ad revenue

#59

Post by maxxxxx » Wed, Jan 16, 2019, 2:10 PM UTC

MarkMc wrote: post #52

bitofhope wrote: post #6

I honestly would not be surprised if the difference in effectiveness of targeted vs untargeted advertisement turns out to be none, negligible or even unfavorable to targeted. Sure, NYT is not enough data to be significant, especially when compensating for my confirmation bias, but I would really like to see more sites stop tracking, at least from an ethical standpoint.

I would be absolutely astounded if targeted ads did not provide significant long-term advantage to a big player like Google. How can it be that having intimate knowledge of someone would not allow you to sell them more stuff? John has a 10th wedding anniversary coming up. Mary is single and goes to the gym every day. Steve just got a promotion and likes the BMW 3 series. Michael is overweight but has just gone on a d…

Maybe the cost for effective targeted advertising is too high. You have to collect a ton of data and analyze it which costs a lot of money. Maybe untargeted are more cost effective.

Re: After GDPR, The New York Times cut off ad exchanges and kept growing ad revenue

#60

Post by Nursie » Wed, Jan 16, 2019, 2:11 PM UTC

jarfil wrote: post #51

pytyper2 wrote: post #31

Earlier quoted context omitted.

This makes sense to me. If you are targeting me, I have told you I like it, and therefore don't need you to tell me about it.

The worst kind of targeting is when I go to a website, browse some similar items comparing their features and price, buy one... and now you decide to show me ads about those same products for the next week even if it should be clear that I've already made up my mind (I'm looking at you, Amazon).

We notice you're interested in car insurance, how about some car insurance?

Yeah, I was, you're right, I spent a couple of hours looking at car insurance stuff and now it's done for the year, so you're spending your advertising budget chasing people who are the absolute least likely to be interested...