Ads are part of the enshitification of ChatGPT. In the short-term it might offer them some kind of lifeline and I'm sure in their next investor call they'll highlight their revenue from Ads.
But if their competitors (including open-weight models) can offer similar quality and ease of use, Ads is just another reason _not_ to use them.
I saw them today in the app and yeah… first idea - enshitification. Ads were cringe. But google works the same way - cringe ads. I use google in 3-4 different languages and it’s completely lost.
I occasionally enjoy well executed ads, like something that’s creative enough to make me curious, but nowadays, I’m seeing a lot of ads on YouTube that’s obviously low-effort AI slop, they are boring, fake, and depressing.
This is not really true. Ads, especially the classifieds, were often the only way that newspapers and news organizations stayed afloat aside from public funding. What happened was the major ad-tech companies captured the revenue and the subscriber data that came from ads and viewers that would normally go to those organizations. "Real journalism" often didn't make much if not any money. "Real journalism" at least for most of the 20th century and the beginning of the 21st has always existed alongside advertising. Magazines literally survived on advertising, their entire product to readers was long form content, while, to brands, a higher quality product and more targeted subscriber base. It's pure romanticism to say that ads themselves were the problem. Ad-tech, it's monopoly control in the hands of 2-3 behemoths, is the real issue.
Ads, in general, have one main purpose: to make the public, and especially your prospective customers, aware of your product and how it fits and solves problems for them.
By their nature, they should be a positive.
However, media in general (from classical like newspapers, magazines, radio and TV to more modern like news sites and streaming platforms) has been leveraging sound psychological research to influence people to start holding views the editorial policy wants them to. And translated to ads, to them wanting and paying for things they do not need and have not originally even wanted. Back when Windows 95 came out, there was an urban legend how people waited in lines in front of stores to buy it even if they did not have a computer — I never cared enough to check if this was true, but even back then it was acknowledged how you could influence people to simply spend money where it does not bring them any value.
Tech industry had a large problem to enter the market and sustain itself on what was largely volunteer, non-profit environment of the early Internet and web days. The dot-com boom was over (and crashed), people were starting to build real products, but sustaining them was hard. And enter ad-tech, with profiling and targeting enriching the already existing psychological research about how you get people to spend when they don't really have a need or problem to address.
So I believe you are complaining more about the state of media today — I was actually hoping guerilla journalism through personal blogs gets us the real thing back, but oh boy was I wrong :) — than ads or ad-tech directly.
> ChatGPT recommends bidding somewhere around by £2 per click, but I bid between £0.15 and £0.07 and was still able to get traffic. However 0.46% is a shockingly low conversion rate.
I'm not an ad guru by any means, but isn't it expected that you will get junk clicks with a bid like that? Seems like a you-get-what-you-pay-for scenario.
Well I think your experiment more shows that while Open AI is trying to position this as a viable marketing platform, in reality folks with marketing dollars do spend aren’t buying what they’re selling so you’re able to win at pricing far below where they really want/need the market to be.
The cost of clicks tends to rise over time with advertising platforms (Google 'the law of shitty clickthrus'). This effect is very noticeable with Google Adwords, for example. So it makes sense to get in early with a new platform and try and get some traffic, before the big corporates push the prices up.
This will end in another 'adpocalypse' episode. Something like: ChatGPT generates recipes from the Anarchist Cookbook and adds Home Depot ads at the end, or inserts Amazon affiliate links for the Turner Diaries.
I have no idea why anyone would click on any ad anywhere other than by accident, not knowing if the link might take them to some malware site these days.
I would never click on the weird ads that the supposedly "brightest minds of our generation" have delivered to me via some nonsensical middleman auction system.
But I go to some blogs (philosophy, jazz, and men's clothing) that have ads that are literally just image links. The ads are actually relevant to the content, and because I go to the blog of my own volition, that means that more often than not the ads are at least somewhat relevant and usually worth a look, like the curated ones you see in a magazine. Most other ads online are so irrelevant it's almost comical, and I have no idea why people pay to use those services. All these tech giants who make enormous sums of money from advertising are emperors with no clothes - a hard look at the quality of their product reveals it's unanimously considered garbage by the public. Their promise of delivering hyper targeted advertisement was obviously a lie predicated on market capture and becoming the only middlemen in town.
I agree, before the age of enshittification, when I went to actual blogs, sometimes I did see ads worth my time. But these days most blogs I would visit are hosted on sites that just use ad farms, and the targeting is awful (which, hey, I am on my apple devices most of the time and I disable all tracking) but mostly I would never feel confident that clicking the ad wouldn't bring me to some spyware infested hole. I am far more leery of where I would go when I click it than I am concerned that I'm not the target demographic.
Sometimes (I especially notice this on the App Store, but sometimes Google), companies bid on their own name. So you'll search something like "Discord" and the first result will be an ad for Discord, and then below is the organic results which has them first anyways.
I'd really like for some country to ban all ads, even if just for a while, so people could observe the quasi-experiment.
My personal bet would be the country would suffer for it quite a lot, similar to places that have banned e.g. loans, but it'd be very cool if I were wrong!
If you saw a product from a brand on a billboard on your commute, and a product from a brand you've never seen before, you'd probably be able to pick which one you've never seen before.
I might recognize a color theme or similar. I had a friend make a joke about a lawyer who was apparently on multiple billboards for years. The name didn't click at all, so he had to explain. I did notice later due to the joke priming me.
Do they though? It is often very hard to know with ads. My experience is that most people spending money on ads have no idea if they are getting any sort of return. Are there any studies that show billboard ads work?
BTW That why money off vouchers are so common with ads. Because this is an actual way you can check if the ad is working. Although there are confounding factors - maybe they were going to buy from you before they saw the voucher?
Yes, ChatGPT ads are a heaping pile of garbage. I feel sorry for anyone that’s wasted perfectly good marketing spend on such nonsense.
Outside a captive platform like Amazon, most ads are garbage and a total waste of money.
If I’m on Amazon trying to buy X and my search for X “promotes” an option to the top of the list yeah that probably influences me buying that.
But the broader online ad ecosystem is a total waste of money. Yes, I once searched a question about golf balls… no I don’t want to see “targeted ads” about golf balls following me all over the internet for the next three weeks.
Speaking as a consumer, I’ve always found ad targeting to be way off for Google, YouTube or even facebook. The few times I got ads I ChatGPT it was also off. It’s kinda ironic that with all that AI, they choose not to use any of that intelligence to target ads better. It makes you wonder what all those engineers are working on if not the main revenue driver.
I understand why it's hard, but it's very funny to get ads for a product I've bought for weeks/months after completing the purchase.
I looked for rain boots, and I bought a pair of rain boots. Getting ads for rain boots now isn't going to drive any purchasing, right? Meta or Alphabet get paid still, so the advertisers are getting hosed.
My experience with performance adtech was that 95% of the energy went into targeting whales, with a focus on products that drive massive revenue (mobile gaming, gambling, crypto). It's truly insane what these whales are spending.
Sure, there's still a lot of money in the 5%, especially for a surface like ChatGPT where people are making retail decisions. But I'm not particularly shocked if they're a little behind.
I've been down the same thought path, but my conclusion was that it's a fallacy. Google and Facebook make literal billions with data-driven ad targeting. I think the more apt conclusion is that what would you and I would consider a good or at least tolerable ad experience is not what maximizes their profits.
I vaguely remember a blog post by a former FAANG engineer on HN lamenting the tyranny of designing for the marginal user (ie the one that gets swayed by an epsilon of optimization). That marginal user has incredibly short attention span, makes incredibly dumb decisions and overall has the cognitive capacity of some plankton.
We can't use ChatGPT Ads because ChatGPT banned our company for ads for political content even though we are a CRM/Compliance tool for campaigns and the only ad we had was the autogenerated ChatGPT one they made. No chance at an appeal either. We had this issue with Bing Ads but microsoft fixed it and even personally called to apologize about the mixup.
The main place I see ads is YouTube on iOS. I used to have ad targeting turned off, but I’ve since opted into all of googles tracking and ad-targeting, in attempt to curate the ads I see and improve them.
I’ve had all the Google tracking and ad-targeting turned on for close to a year now. Google still can’t even target me ads that are in English. I can’t read Chinese or Turkish or Spanish but I still get ads in those languages. Most of the ads are still scams or questionably legal products. And nearly all advertisements are blatantly targeted toward the wrong demographics. I get both “you can buy this multi million dollar hotel” and “you can use food stamps at our store”. I sometimes see advertisements for local politicians, which might be targeted (same state) but I just as frequently see ads for politicians I cannot possibly vote for, (opposite side of the country)
HN isn't really a good place to have an informed discussion on ads efficacy unfortunately. In this thread we have many many people talking about how online ads in general are completely worthless. All of the top comments I see are harping on things like Google Ads having bad targeting for reasons like you are getting recommendations on products you just bought (having experience in the ads industry, yes this is actually a good signal that you might buy another one, even if you don't want to believe it).
Meanwhile, the article is actually discussing ChatGPT ads specifically compared to Google Ads, which they do have success with.
Anyway, my guess is that it's a combination of 1. OpenAI being new and not having the massive work needed yet to set up a fully successful ads network, but they're recruiting a lot of talent from the area + recsys so it's probably not too far off. And 2. we don't really know in general yet how best to do in context ad recommendations. I think this will also improve pretty rapidly with how much talent is working in the space. The best ads will probably look a lot more like Google Search per-request targeted ads rather than Instagram style per-user targeted ads, but I believe we'll see lots of gains fast.
Regarding showing ads for something you just bought, I’ve always assumed the principle here is that because conversion is extremely unlikely for a given ad, the super unlikely case you’ll buy something twice is actually more likely than a random other thing.
54 comments
[ 2.1 ms ] story [ 23.6 ms ] threadBut if their competitors (including open-weight models) can offer similar quality and ease of use, Ads is just another reason _not_ to use them.
- Why is online news just partisan rage-baiting? Ads
- Why is real journalism dying? Ads
- Why are your web search results less helpful? Ads
- Why are recipes 2,000 words long? Ads
- Why is social media engineered for scroll-maxxing?
- Why is your car/TV/refrigerator spamming you?
- Why are pharmaceuticals so expensive?
- Why is your population voting for assholes?
Future generations are going to look back at how we live with advertising the same way we look back on leaded gasoline and asbestos.
This is not really true. Ads, especially the classifieds, were often the only way that newspapers and news organizations stayed afloat aside from public funding. What happened was the major ad-tech companies captured the revenue and the subscriber data that came from ads and viewers that would normally go to those organizations. "Real journalism" often didn't make much if not any money. "Real journalism" at least for most of the 20th century and the beginning of the 21st has always existed alongside advertising. Magazines literally survived on advertising, their entire product to readers was long form content, while, to brands, a higher quality product and more targeted subscriber base. It's pure romanticism to say that ads themselves were the problem. Ad-tech, it's monopoly control in the hands of 2-3 behemoths, is the real issue.
By their nature, they should be a positive.
However, media in general (from classical like newspapers, magazines, radio and TV to more modern like news sites and streaming platforms) has been leveraging sound psychological research to influence people to start holding views the editorial policy wants them to. And translated to ads, to them wanting and paying for things they do not need and have not originally even wanted. Back when Windows 95 came out, there was an urban legend how people waited in lines in front of stores to buy it even if they did not have a computer — I never cared enough to check if this was true, but even back then it was acknowledged how you could influence people to simply spend money where it does not bring them any value.
Tech industry had a large problem to enter the market and sustain itself on what was largely volunteer, non-profit environment of the early Internet and web days. The dot-com boom was over (and crashed), people were starting to build real products, but sustaining them was hard. And enter ad-tech, with profiling and targeting enriching the already existing psychological research about how you get people to spend when they don't really have a need or problem to address.
So I believe you are complaining more about the state of media today — I was actually hoping guerilla journalism through personal blogs gets us the real thing back, but oh boy was I wrong :) — than ads or ad-tech directly.
I'm not an ad guru by any means, but isn't it expected that you will get junk clicks with a bid like that? Seems like a you-get-what-you-pay-for scenario.
No. Bidding low means less or no clicks, but should not affect traffic quality.
Did they try bidding more? It's possible they are just only getting the dregs. (In which case the ads would be expensive instead of garbage)
OP here. No. The cheapest version of my PerfectTablePlan software is £19.95 + VAT, one time fee. So I would be insane to bid £2 per click.
>It's possible they are just only getting the dregs.
Bidding higher generally means more clicks, but the quality could be worse due to winning bids when your content is less relevant.
But I go to some blogs (philosophy, jazz, and men's clothing) that have ads that are literally just image links. The ads are actually relevant to the content, and because I go to the blog of my own volition, that means that more often than not the ads are at least somewhat relevant and usually worth a look, like the curated ones you see in a magazine. Most other ads online are so irrelevant it's almost comical, and I have no idea why people pay to use those services. All these tech giants who make enormous sums of money from advertising are emperors with no clothes - a hard look at the quality of their product reveals it's unanimously considered garbage by the public. Their promise of delivering hyper targeted advertisement was obviously a lie predicated on market capture and becoming the only middlemen in town.
I suppose depends how broadly “ad” is taken to be. Is a webpage an ad? Presumably no if it’s your landing page.
OP is presumably talking about invasive ads that try to take the attention of people who aren’t looking for the product in question
My personal bet would be the country would suffer for it quite a lot, similar to places that have banned e.g. loans, but it'd be very cool if I were wrong!
I wonder why you think the country would suffer?
I don't think any single form of advertising is effective on the whole population.
BTW That why money off vouchers are so common with ads. Because this is an actual way you can check if the ad is working. Although there are confounding factors - maybe they were going to buy from you before they saw the voucher?
Outside a captive platform like Amazon, most ads are garbage and a total waste of money.
If I’m on Amazon trying to buy X and my search for X “promotes” an option to the top of the list yeah that probably influences me buying that.
But the broader online ad ecosystem is a total waste of money. Yes, I once searched a question about golf balls… no I don’t want to see “targeted ads” about golf balls following me all over the internet for the next three weeks.
I looked for rain boots, and I bought a pair of rain boots. Getting ads for rain boots now isn't going to drive any purchasing, right? Meta or Alphabet get paid still, so the advertisers are getting hosed.
My experience with performance adtech was that 95% of the energy went into targeting whales, with a focus on products that drive massive revenue (mobile gaming, gambling, crypto). It's truly insane what these whales are spending.
Sure, there's still a lot of money in the 5%, especially for a surface like ChatGPT where people are making retail decisions. But I'm not particularly shocked if they're a little behind.
I vaguely remember a blog post by a former FAANG engineer on HN lamenting the tyranny of designing for the marginal user (ie the one that gets swayed by an epsilon of optimization). That marginal user has incredibly short attention span, makes incredibly dumb decisions and overall has the cognitive capacity of some plankton.
I couldn't possibly be more par for course for OpenAI
1. Click fraud is rampant, and the ad platforms seem to not care in the least as long as they can drain your budget.
2. The ad bidding process is completely opaque, and no auditing seems to be done, my inclination is that there is large scale fraud here too.
The main place I see ads is YouTube on iOS. I used to have ad targeting turned off, but I’ve since opted into all of googles tracking and ad-targeting, in attempt to curate the ads I see and improve them.
I’ve had all the Google tracking and ad-targeting turned on for close to a year now. Google still can’t even target me ads that are in English. I can’t read Chinese or Turkish or Spanish but I still get ads in those languages. Most of the ads are still scams or questionably legal products. And nearly all advertisements are blatantly targeted toward the wrong demographics. I get both “you can buy this multi million dollar hotel” and “you can use food stamps at our store”. I sometimes see advertisements for local politicians, which might be targeted (same state) but I just as frequently see ads for politicians I cannot possibly vote for, (opposite side of the country)
I get mostly mistargeted ads, but I’ve put a lot of effort into it.
That means you can't be tracked because it's too stupid to do so.
Meanwhile, the article is actually discussing ChatGPT ads specifically compared to Google Ads, which they do have success with.
Anyway, my guess is that it's a combination of 1. OpenAI being new and not having the massive work needed yet to set up a fully successful ads network, but they're recruiting a lot of talent from the area + recsys so it's probably not too far off. And 2. we don't really know in general yet how best to do in context ad recommendations. I think this will also improve pretty rapidly with how much talent is working in the space. The best ads will probably look a lot more like Google Search per-request targeted ads rather than Instagram style per-user targeted ads, but I believe we'll see lots of gains fast.