This analogy doesn't hold at all. Casinos are engineered for the house to win and the customer to lose because the casino is paying the customer directly. They are disincentivized to let the customer win. If the customer always won, the casino goes out of business very quickly.
Ad auctions are completely different, since they are just a marketplace. A customer's (advertiser's) win comes from a third party, and the point of the ad platform is to connect that third party to the customer. The ad platform would love to be able to make every single customer win all the time. Although, if they did that, it would create a lot more demand and even out the ROI with increased ad prices. But that's not the "evil ad platform" engineering the customers to always lose in the way casinos do, that's simply basic supply/demand economics.
COI disclaimer: My current job is a small part of an ads platform on a website you have heard of. I dislike ads but realize they are a necessity for running some parts of the internet.
Yea, I also work in ads and the article doesn't make sense at all. A way bigger factor than anything mentioned in this article is simply what you as an advertiser can afford as a CAC (cost to acquire a customer). There are some advertisers who are selling low cost/low repeat products that advertising is nearly impossible to make profitable. Meanwhile, other advertisers are selling high cost/high repeat products to that same person. The latter can afford to price the former out of ad inventory.
Many companies are spending significant amounts of money (many millions) and not even telling the advertising platforms the actual revenue they generated from the ads, just the number of "leads" or "customers" because they are paranoid about the very thing in the article (which I don't actually believe to be true anyway).
I'm in the paranoid camp. I've always wondered who in their right mind sends profit data upstream when it's exactly what the networks need to push margins down. Maybe Google and the others don't actually use that for ill intent, but they don't exactly have trustworthy reputations. If it came out tomorrow that they were using the data to inflate their margin, no one would be surprised.
I think you misunderstood the article. They are not claiming the ad platform is actively calibrating prices to take everything you can spare, but that that is an inherent result of the bidding war, at least for product categories where products are at least a little bit interchangable.
The article very heavily implies that ad platforms are actively bad. It draws comparisons to casinos, including explicitly saying that ad platforms are engineered to act the same. It says "we never blame the auction house" in a way that is obviously implying we should blame the auction house.
The point is that natural game theory dynamics lead to the actual seller (any of them) in a competitive market unable to make a profit as they invest incremental dollars in capturing attention, while the marketplace makes a killing.
Container ship capacity has inelastic supply and elastic demand. Natural game theory dynamics in a competitive market should lead to customers being unable to make a profit on having their product shipped, because anyone who can profit will invest incremental dollars in shipping more product. Do we say that overseas shipping is an exploitative industry that should be avoided?
They (certainly Google) have always been pretty clear that they are out to capture most of the surplus value of each ad impression. The auction setup was chosen and carefully tuned to reach this outcome. Within each niche a small number of sellers get a modest positive value from the ads, and a large mass of others loose money and churn.
I don't see why this would be surprising or controversial. It's pretty much how any finite commodity is sold.
> Meanwhile, other advertisers are selling high cost/high repeat products to that same person. The latter can afford to price the former out of ad inventory.
This is correct. The best advertisers are setting the floor prices for ads within their niche. This is why StackOverflow, before their fall, had just a few advertisers like Google, IBM/RedHat, and Azure. Those advertisers were willing to buy pretty much the whole inventory at a price a startup focusing on say a single niche SO tag couldn't compete with.
Disclosure: Work on an ad startup I helped build. I talk to over a dozen advertisers every day. We have similar problems with advertisers trying to buy most/all of our inventory and we explicitly cap the max percentage we'll sell to any one advertiser.
In the 2000s my political activism and business involvements had me in contact with people involved in local newspaper and radio ads in my community.
I always had the impression that there is a tranche of advertiser who is "made of money" who doesn't care what ads cost and... that person drives the market.
For instance in my town there is one guy who owns all the new card dealerships in town and he saturates the airwaves, I think because he loves to hear his name on the radio. For the rest of us it just reminds us that the next time we buy a car we're going to drive 50 miles to a town where one guy doesn't own all the new car dealerships.
If a radio ad costs him $20 I am sure he gets at least $20 of pleasure from hearing his name and he makes monopoly profits so he can afford it!
A similar story was that a tiny ad in the local newspaper cost $100 a day and it was stupendously difficult to make a case that this was worth it for most small businesses, especially when you could get a storefront for that much or hire a part-timer. I'd talk to people who sold ads for the paper and they told me, "we tell them that their customers will think they went out of business if they stop advertising!"
Some of the rapid collapse of "old media" was that the foundations of it were pretty rotten and people didn't see the value they were getting out of it and were ready to pull out of it as soon as they could.
A lot of companies are just running ads and literally losing their shirts on them, because they don't know how to run paid advertising properly.
Now you come along and want to be #1 and you have to try and outbid people who are losing money on every click.
You have to be top of your game to make paid ads profitable in 2026. High margin product and every step of your process super-optimized and split-tested to the nth degree.
It's an entirely AI-generated article. There's almost certainly no human making a good-faith argument here. You're wasting your time arguing with an LLM, which is a pretty good illustration of why I think we should be shunning such content. Only so many hours in a day.
I'm now shunning any domain I don't recognize and de facto upranking any domain I recognize from being around for more than five years. Which is unfortunate because not every old domain deserves a positive reputation, they've only gained that by default relative to most of what's "published" today.
Oh I'm very aware of that. I'm not trying to arguing against the article. I'm trying to inform the dozens of people who are upvoting this post and implicitly think it makes any semblance of a good point.
What makes you say it is LLM generated? Having to sift though LLM generated content is part of my day job these days but I don't recognise this as LLM generated. It has a voice and a thread running through it which I've never seen an LLM manage.
It might be a bad analogy and you may not agree with the message but that was a thing well before LLMs entered the everyday vernacular.
They're selling something. Their article is about how "ad funnels are broken" and how unsavory they are (they're preying on your psychology like casinos do), meanwhile their product (from cursory glance) looks like a funnel for sales outreach - instead of using paid ads you'd look for customer companies and reach out to them.
Marketing material like this looks like AI generated content, and are equivalent, even if a human wrote it.
Well, for one thing, Pangram confidently assesses it as 100% AI...
It felt AI-voice to me throughout as well. There's this very particular, like... faux-edgy, faux-dramatic register that the same sorts of people who hustle LinkedIn try to evoke from LLMs. I'm convinced some "buy my course" bro is selling around to all of them a "killer prompt to revenuemaxx your agent in your sleep and hyper-mega-charge your content" or something.
Some of the tells: high drama treatment of dilemmas with low or no stakes; not-really-very-human imagined mental state/reactions projected onto the little-guy "us" in the situation; poorly-modeled human sentiments; some imagined dastardly adversary that's either an impersonal force or motivated by strange alleged incentive structures; a chain of logic made up of individually-credible links that only connect to one other in a Markov-chain-babble sort of way...
> A normal person walking away from a slot machine down ten dollars would recognize they lost. But the ad dashboard doesn't present it as a loss. It presents it as a near-miss. It serves up a carousel of green percentage metrics: impressions up 14%, click-through rate up 0.2%, cost-per-click down 4 cents.
And then, of course, the pitch for the firm, which was clearly not worth a human's eyes:
> "The Global Commerce Graph is a live graph of companies under 100 employees: the quiet majority that databases and ad auctions can't see. Every morning at 6:00 AM ET, it surfaces 3 to 5 target buyers, partners, and audiences for each registered domain."
You have picked one dimension, incentives, and say the analogy doesn't work because on that dimension it's different. Analogies aren't supposed to be exactly the same.
The point of the analogy stands - most players using paid advertising can't win. They don't realize the entire structure of the market makes it close to impossible for them to win. Google (and a small number of others) have an extremely strong position in what is effectively the distribution channel for a lot of businesses, and a pricing mechanism which enables them to take a giant chunk of economics of the transaction.
Sure, we can make some analogy between ads and gambling, and transitively apply that to ad platforms and casinos. But this article frames ad platforms as exploitative entities engineering advertisers to lose money, when the reality is that the ad platforms provide a marketplace, and via the consequences of supply and demand this marketplace can end up being negative ROI for some advertisers.
Ads have a positive EV while gambling has a negative EV, which completely changes how the entities hosting these products behave. Casinos try to get you to stay on by tricking you into thinking you have a positive EV, while ad platforms get you to stay on by genuinely trying to give you a positive EV. If you aren't getting positive results on your ads, you genuinely have the agency to adjust your ads to try to get positive EV. Casinos are explicitly engineered to give you negative EV regardless of how you behave, and kick you out if you try to change that (e.g. card counting).
If we say that you shouldn't buy the products of an ads platform because the platform tries to get you to buy more, we may as well say you shouldn't buy the products of a grocery store because the store tries to get you to buy more.
People who believe ads are necessary have an underlying assumption that commerce is also a requirement. It's a shaky basis, and leads to a bunch of other messy things (ads to demographic to profiles to tracking)
No, they believe running a website has costs and if you aren't selling something your only other options seems to be ads or charity. The assumption is actually the opposite of commerce being a necessity. If you aren't selling something, that's when you need ads to stay afloat.
Not strictly necessary, but functionally necessary. Many websites provide services that are expensive to run and can't be funded by other products from the company, so the website itself has to be profitable. I'm not convinced that these websites (e.g. most social media) are societally necessary, but I don't see any scenario where they continue to exist without advertising (barring extreme measures such as government funding).
It's like casino hosting poker games, like a tournament. Between the players (advertisers) it's zero sum, because there are just so much money (end user attention and money), and house (ad platform) takes a cut in the proceeds.
I use to think we lived in an enlightened age, where we have now learned that we need to use scientific reasoning and analysis to make decisions. It is what lifted us up from superstition and folk tales.
But society, still, collectively, falls for snake oil, over and over again.
Anyone that has an interest in the topic should listen (or read) about what
Steve Tadelis did in the 2012 timeframe. He was an economics professor at the University of California, Berkeley, and then worked at eBay, where he led the famous experiment on Google ad effectiveness.
People are born in ignorance. Reason takes energy. Nature optimizes for energy efficiency.
This is likely a hole us meat puppets can never crawl out of. Any moment we stop spending energy to climb out, we're sliding down the ant lion nest to ignorance.
i can't square away all of this with the simple math that if you make a product that makes sense for online advertising, you make boatloads of money by using online ads, and make less as more people compete in it.
this is true of GLP1 ads and for many, many subscription services like video streaming.
i'm not sure it's about falling for something so much as it is that people who like, sell products through amazon, well, they are fucked no matter what.
Big advertisers (especially big retailers) are constantly employing fairly sophisticated methods to validate that they are getting a return on their investment with ads. It really is impossible to square this reality with comments reposting that one single paper from 2012 (?) (which doesn't even really say what they think it says---they ought to actually read it!) because those commenters have absolutely no clue what's going on.
and that somebody is looking seven generations (or even a seventh of a generation) ahead seems so strange in 2026 when so many people think they can enjoy being poor forever (like billions of years) if they don't manage to pound out 15 ai slop blog posts to prove to everyone else that they know where ai is going.
I don't subscribe to the idea that there ever was a reliable source of truth. It's just that having open access to information makes it so people see competing versions of framings some of which are incompatible with each other. For example, Christianity, Judaism and Islam can't all be true. They could all be false, however. Same applies to anything else. Just because you clearly know that this other view is wrong doesn't make your own perspective right.
Unfortunately or maybe fortunately it is entirely possible to live a long and happy life being entirely wrong about a great many important things without realizing.
I can't be sure that what I think is true, is really true. Therefore I don't really want to make the decision on how other people are supposed to live their life and find anyone that thinks himself worthy of doing just that suspect. I prefer someone that exploits people for their own gain, since then we are at least clear about the incentive structure and can measure and verify that gain and also negotiate for a cut. Something that isn't really possible when we are told to do things for our own good because God said so.
Back in the 2000’s I was involved with numerous “social” systems before there was Facebook, the thing they all had in common was that we knew we lived or died based on being viral so the most important knob we could turn in the process was the rate at which people completed the onboarding and recommended new people. So we were obsessive about plugging leaks in our funnel from both a craftsmanship and business standpoint.
Today whether it is mobile games or anything else I’m often left wondering if there is anybody left who remembers what it takes to tune up a funnel. Now I am told that for western mobile games they aren’t even interested in players who would care what the game mechanics are but instead just want to filter for a handful of ‘whales’ who will start spending money when they encounter the slightest speed bump… but still!
Right now I trade it back to the time when Facebook discovered its business model, made virality impossible, and convinced everybody to pay for traffic. You’d think the same motivations exist there, like you can save your money or your investors’ money if you have a good onboarding process so you should but what you find is…. People don’t.
I’ve had people blame the OKR/KPI culture for this which has everybody distracted by sub optimization…. But there’s also the problem that the thing you really need to compare customer acquisition cost to is lifetime customer value. So yeah, you spend $50 to get a customer to spend $20 where your COGS is $18…. That’s the part you know about and it looks awful and you tell yourself that that LCV will be more than $2 which is true because some of those customers will come back but it’s speculative. So you think you are practicing ‘quantitative marketing’ but you are like Coca-Cola in the mid 20th century spending on every kind of advertising because you think the brand equity is worth it.
I don't really think this is true, as if advertising didn't work people wouldn't use it.
That being said, there is something to be said about how extractive it has become.
It used to be you ran a TV ad, and if you made a great ad, you got amazing ROI, you were rewarded for being clever.
Now, it's all captured algorithmically, there are no "great ads", you can only have good ads or bad ads. They know full well if your ad is great then they can charge you whatever they want, inevitably making your ROI only "good enough". Sure go to another platform where the algorithm will price you out eventually.
Look at Him & Hers / Ro / GoodRx / etc. They basically take every dollar they make and throw it into advertising, it's the only way to stay relevant.
> if advertising didn't work people wouldn't use it.
Advertising works, and the unfortunate end result for fungible products is that the product price increases to pay for the advertising.
Producer A and B each cover half the market and compete on price. When A starts advertising to get more market share, the loss is for producer B who's margin decreases due to smaller volume. B also starts advertising. To make ends meet A and B have to increase product price to cover the cost of advertising.
This is very simplified reasoning. The gist is: because advertising works, consumers have to pay more for their products. The more advertising there is, the more prices will increase without products becoming better.
It is worth checking out. Here is a relevant couple of quotes...
HWANG: I think a second thing is, a little bit like in the subprime-mortgage crisis, you do have people who have very perverse incentives, I think, to push the effectiveness of online ads. That’s the ad agencies, the ad platforms themselves, the people who run ad technology. All these people, I think, have a very strong incentive to say, “No, this stuff is way better than earlier generations of advertising, and this is why you should use it.”
FE: If you’ve been listening closely, you’ll notice this is the exact same problem Steve Levitt talked about regarding the TV-ad ecosystem: human beings generally make decisions based on self-interest.
LEVITT: No chief marketing officer is ever going to say, “Hey, I don’t know, maybe ads don’t work. Let’s just not do them and see what happens.”
FE: Or, as the author Upton Sinclair once wrote: “It is difficult to get a man to understand something when his salary depends on his not understanding it.”
Claude doesn't understand that the net revenue per conversion (including variable revenue minus all other expenses) is different for everyone bidding on the same ad slot.
The gambling analogy doesn’t quite hold here. Gambling is a zero sum game. Advertising is not a zero sum game. There is room for both sides to walk away with gains from trade.
The article explains why that isn't the case: it's because advertisers (Google's clients who run ads) are competing with each other to push the cost of advertising up to the point where the cost of an ad and the value it provided reach a rough equilibrium.
And in economic theory, perfect competition drives the margin to zero. The gambling part is not explaining that theory, but rather the effect on human behavior when the short term results hover around that zero margin line.
The margin for the marginal participant! The more profitable participants will still make a profit. You need some argument like "the profitable participants are going to seek to expand their business until it isn't profitable anymore" to truly eradicate profits in theory.
If you look at the oil business, for instance, the Saudis have a much lower cost than anybody else in the business so they could be profitably selling oil in a world where everybody else is losing money.
> advertisers (Google's clients who run ads) are competing with each other to push the cost of advertising up to the point where the cost of an ad and the value it provided reach a rough equilibrium.
Yup. That's econ 101, except that the price you pay is not the value you would have gotten, it is the price one bid increment higher than the value your competition would have gotten/was willing to pay. You can still have consumer surplus on your side of the trade as long as the sale is a proxy or vickery auction - like most ad sales are.
> You can still have consumer surplus on your side of the trade as long as the sale is a proxy or vickery auction - like most ad sales are.
You can still make a profit on your side of the trade if it's a regular "winner pays what they bid" auction too. No one can stop you.
People can bid more than the ad is worth to you, and stop you from running the ad at all, but they can't make you bid more than the ad is worth and take a loss.
That wasn't the analogy. It's gambling in the sense "the house always wins".
It's not a simple "both sides" issue - the scenario has multiple sides. You find a good keyword, someone finds you, you both benefit. Great!
Then a third party finds and uses your strategy, and bids higher in the ad auction. This reduces profits. As more parties join in, the profits decrease.
The thesis is that those profits don't simply go to zero, but can actually result in a net loss. In the meanwhile, those higher bids end up as profits for the auction house.
Uh, yeah the house always wins because it is zero sum. In a competitive market there is gains from trades. Rational actors can drive consumer and producer surplus to zero but once rational actors won't continue to buy at a loss.
It's often constant sum, because in ads can move sales between organisations instead of creating value.
With the automated bidder maximising ad prices for shareholder returns it can become zero sum equivalent. Every gain is matched by a loss elsewhere.
This changes if ads are creating new demand, or if consumers are getting value they otherwise wouldn't have found. But in many markets that won't be the case.
Gambling, advertising, trading, and AI have one thing in common: if you don’t do well it’s your fault. You didn’t prompt right, or target the right audience, or you don’t have the right harness, and the only way to remedy is always to spend more money.
The only way to keep out of those traps is to have real metrics of your own and benchmarks of your own. Performance changes over time so you can never trust the current state will continue.
Psychological elements aside, this sort of misses that goods and services require multiple inputs from different counterparties (or different markets) that have different pricing power (some of which might be close to monopolies). It’s implausible that whenever there is one monopoly seller anywhere that all added economic value (or more than all, as in the article??) has to go to that one.
If I have to necessarily buy ads from google and water from the local water company, where does the economic value go in equilibrium? Clearly not all to the water company and not all just to google.
That’s just victimhood mentality. There are tons of examples of companies printing with ads. I’m not saying that there aren’t players that play net-negative games. And those playing hurt everyone but the advertising sellers, but the game is not fundamentally flawed.
The whole ecosystem is extractive with the goal of improving quarterly earnings by removing control from the people spending the money. Online ads started out being trackable and much higher ROI than traditional media. Then they ran traditional media out of business while continously raising their own rates (opaque quality score, nebulous targeting options) and then go even further, essentially stealing publishers' content for their own walled garden models.
So glad not having to look at this damn dashboard and all its myriad of tools, interconnects and whatnots, which even claim to be able to attribute a sale to some bizarre ins and outs of the search. Of course if all cookies, tags, and pixels that you have to install beforehand work. And also validations, and confirmations, legal checks, etc, etc. It reaaally is a giant slot machine. Really, first you have to be very lucky to get an actually good connectivity of all the needed info, and then you need to see Google in its lucky day actually driving this traffic that your struggling business so much needs.
As my godfather put it once - dude, you better give $10 each customer to come and buy, rather than spend $420 to get three. It will work better.
I very much hope this all dies for good with AI, something tells me we're in for v2 where all the bullshit is injected into LLM streams, and then other local LLMs would need to adblock it. Sad story. Not the internet I got fascinated about as a kid, tempting to attribute it to age, but there was surely more awe and fun in it than this never-ending slop of irrelevancy, and I'm not even talking AI.
I tried social media ads years ago for my project, and it was all bots. Wasted hundreds of dollars, so I'm afraid to try again today now that there is AI.
I drive by an Ikea on my way to work. They just installed a video wall in view of the highway. It is huge, four stories tall and square. It gets noticed but is not intrusive. It is not animated. In daylight it just looks like a big painting. At night it isn't any brighter than the parking lot around it. Yesterday it was advertising an end table for 65$. This morning it was something with a plant in the background. I actually saw and remember that ad. I almost never see ads on websites and certainly don't remember those I do. But that Ikea wall... THAT I remember.
Business are trees. Customers are sunlight. Advertisements are the trunks.
Incredibly wasteful and as tall as they can physically be. If trees could talk to each other and decide to limit the heights of the trunks arbitrarily there would be so much energy saved that could be put into seeds and fruit to spread the forest faster.
But the trees of course can't agree because any defector would gain advantage over the cooperating trees.
In case of the economy the government should step in and replace the whole advertising industry with a single always up to date catalog of goods and services. If comapnies couldn't get ahead by inflating marketing budgets they would be forced to compete on quality and price instead what would be beneficial for everything including the health of those companies.
Amazon got fined a bajillion dollars for deciding to add undisclosed reserve prices for their ad auctions. They just decided that too many auctions ended too low, so they added fake reserve prices and just chose what the price would be...
Google and Meta have also both been caught in advertising shenanigans -- I forget the details but I recall getting some (tiny) class action settlement check.
This analogy doesn't hold at all. Casinos are engineered for the house to win and the customer to lose because the casino is paying the customer directly. They are disincentivized to let the customer win. If the customer always won, the casino goes out of business very quickly.
Ad auctions are completely different, since they are just a marketplace. A customer's (advertiser's) win comes from a third party, and the point of the ad platform is to connect that third party to the customer. The ad platform would love to be able to make every single customer win all the time. Although, if they did that, it would create a lot more demand and even out the ROI with increased ad prices. But that's not the "evil ad platform" engineering the customers to always lose in the way casinos do, that's simply basic supply/demand economics.
COI disclaimer: My current job is a small part of an ads platform on a website you have heard of. I dislike ads but realize they are a necessity for running some parts of the internet.
Yea, I also work in ads and the article doesn't make sense at all. A way bigger factor than anything mentioned in this article is simply what you as an advertiser can afford as a CAC (cost to acquire a customer). There are some advertisers who are selling low cost/low repeat products that advertising is nearly impossible to make profitable. Meanwhile, other advertisers are selling high cost/high repeat products to that same person. The latter can afford to price the former out of ad inventory.
Many companies are spending significant amounts of money (many millions) and not even telling the advertising platforms the actual revenue they generated from the ads, just the number of "leads" or "customers" because they are paranoid about the very thing in the article (which I don't actually believe to be true anyway).
I'm in the paranoid camp. I've always wondered who in their right mind sends profit data upstream when it's exactly what the networks need to push margins down. Maybe Google and the others don't actually use that for ill intent, but they don't exactly have trustworthy reputations. If it came out tomorrow that they were using the data to inflate their margin, no one would be surprised.
I think you misunderstood the article. They are not claiming the ad platform is actively calibrating prices to take everything you can spare, but that that is an inherent result of the bidding war, at least for product categories where products are at least a little bit interchangable.
The article very heavily implies that ad platforms are actively bad. It draws comparisons to casinos, including explicitly saying that ad platforms are engineered to act the same. It says "we never blame the auction house" in a way that is obviously implying we should blame the auction house.
The point is that natural game theory dynamics lead to the actual seller (any of them) in a competitive market unable to make a profit as they invest incremental dollars in capturing attention, while the marketplace makes a killing.
Maybe that is bad?
Container ship capacity has inelastic supply and elastic demand. Natural game theory dynamics in a competitive market should lead to customers being unable to make a profit on having their product shipped, because anyone who can profit will invest incremental dollars in shipping more product. Do we say that overseas shipping is an exploitative industry that should be avoided?
They (certainly Google) have always been pretty clear that they are out to capture most of the surplus value of each ad impression. The auction setup was chosen and carefully tuned to reach this outcome. Within each niche a small number of sellers get a modest positive value from the ads, and a large mass of others loose money and churn.
I don't see why this would be surprising or controversial. It's pretty much how any finite commodity is sold.
> Meanwhile, other advertisers are selling high cost/high repeat products to that same person. The latter can afford to price the former out of ad inventory.
This is correct. The best advertisers are setting the floor prices for ads within their niche. This is why StackOverflow, before their fall, had just a few advertisers like Google, IBM/RedHat, and Azure. Those advertisers were willing to buy pretty much the whole inventory at a price a startup focusing on say a single niche SO tag couldn't compete with.
Disclosure: Work on an ad startup I helped build. I talk to over a dozen advertisers every day. We have similar problems with advertisers trying to buy most/all of our inventory and we explicitly cap the max percentage we'll sell to any one advertiser.
In the 2000s my political activism and business involvements had me in contact with people involved in local newspaper and radio ads in my community.
I always had the impression that there is a tranche of advertiser who is "made of money" who doesn't care what ads cost and... that person drives the market.
For instance in my town there is one guy who owns all the new card dealerships in town and he saturates the airwaves, I think because he loves to hear his name on the radio. For the rest of us it just reminds us that the next time we buy a car we're going to drive 50 miles to a town where one guy doesn't own all the new car dealerships.
If a radio ad costs him $20 I am sure he gets at least $20 of pleasure from hearing his name and he makes monopoly profits so he can afford it!
A similar story was that a tiny ad in the local newspaper cost $100 a day and it was stupendously difficult to make a case that this was worth it for most small businesses, especially when you could get a storefront for that much or hire a part-timer. I'd talk to people who sold ads for the paper and they told me, "we tell them that their customers will think they went out of business if they stop advertising!"
Some of the rapid collapse of "old media" was that the foundations of it were pretty rotten and people didn't see the value they were getting out of it and were ready to pull out of it as soon as they could.
Is this why during a show I'll see the SAME ad multiple times? Like if i ignored it the first time, why would I care the second time.
Some ads are designed for awareness, rather than for direct action. Repetition creates a stronger long term memory.
The kars4kids jingle just began repeating itself in my head
If you block a lot of trackers, you get the same ads over and over on YouTube.
A lot of companies are just running ads and literally losing their shirts on them, because they don't know how to run paid advertising properly.
Now you come along and want to be #1 and you have to try and outbid people who are losing money on every click.
You have to be top of your game to make paid ads profitable in 2026. High margin product and every step of your process super-optimized and split-tested to the nth degree.
aren't business losses deductible? this is probably yet another way we are subsidizing tech and surveillance.
It's an entirely AI-generated article. There's almost certainly no human making a good-faith argument here. You're wasting your time arguing with an LLM, which is a pretty good illustration of why I think we should be shunning such content. Only so many hours in a day.
I'm now shunning any domain I don't recognize and de facto upranking any domain I recognize from being around for more than five years. Which is unfortunate because not every old domain deserves a positive reputation, they've only gained that by default relative to most of what's "published" today.
Oh I'm very aware of that. I'm not trying to arguing against the article. I'm trying to inform the dozens of people who are upvoting this post and implicitly think it makes any semblance of a good point.
What makes you say it is LLM generated? Having to sift though LLM generated content is part of my day job these days but I don't recognise this as LLM generated. It has a voice and a thread running through it which I've never seen an LLM manage.
It might be a bad analogy and you may not agree with the message but that was a thing well before LLMs entered the everyday vernacular.
They're selling something. Their article is about how "ad funnels are broken" and how unsavory they are (they're preying on your psychology like casinos do), meanwhile their product (from cursory glance) looks like a funnel for sales outreach - instead of using paid ads you'd look for customer companies and reach out to them.
Marketing material like this looks like AI generated content, and are equivalent, even if a human wrote it.
Well, for one thing, Pangram confidently assesses it as 100% AI...
It felt AI-voice to me throughout as well. There's this very particular, like... faux-edgy, faux-dramatic register that the same sorts of people who hustle LinkedIn try to evoke from LLMs. I'm convinced some "buy my course" bro is selling around to all of them a "killer prompt to revenuemaxx your agent in your sleep and hyper-mega-charge your content" or something.
Some of the tells: high drama treatment of dilemmas with low or no stakes; not-really-very-human imagined mental state/reactions projected onto the little-guy "us" in the situation; poorly-modeled human sentiments; some imagined dastardly adversary that's either an impersonal force or motivated by strange alleged incentive structures; a chain of logic made up of individually-credible links that only connect to one other in a Markov-chain-babble sort of way...
> A normal person walking away from a slot machine down ten dollars would recognize they lost. But the ad dashboard doesn't present it as a loss. It presents it as a near-miss. It serves up a carousel of green percentage metrics: impressions up 14%, click-through rate up 0.2%, cost-per-click down 4 cents.
And then, of course, the pitch for the firm, which was clearly not worth a human's eyes:
> "The Global Commerce Graph is a live graph of companies under 100 employees: the quiet majority that databases and ad auctions can't see. Every morning at 6:00 AM ET, it surfaces 3 to 5 target buyers, partners, and audiences for each registered domain."
You have picked one dimension, incentives, and say the analogy doesn't work because on that dimension it's different. Analogies aren't supposed to be exactly the same.
The point of the analogy stands - most players using paid advertising can't win. They don't realize the entire structure of the market makes it close to impossible for them to win. Google (and a small number of others) have an extremely strong position in what is effectively the distribution channel for a lot of businesses, and a pricing mechanism which enables them to take a giant chunk of economics of the transaction.
Sure, we can make some analogy between ads and gambling, and transitively apply that to ad platforms and casinos. But this article frames ad platforms as exploitative entities engineering advertisers to lose money, when the reality is that the ad platforms provide a marketplace, and via the consequences of supply and demand this marketplace can end up being negative ROI for some advertisers.
Ads have a positive EV while gambling has a negative EV, which completely changes how the entities hosting these products behave. Casinos try to get you to stay on by tricking you into thinking you have a positive EV, while ad platforms get you to stay on by genuinely trying to give you a positive EV. If you aren't getting positive results on your ads, you genuinely have the agency to adjust your ads to try to get positive EV. Casinos are explicitly engineered to give you negative EV regardless of how you behave, and kick you out if you try to change that (e.g. card counting).
If we say that you shouldn't buy the products of an ads platform because the platform tries to get you to buy more, we may as well say you shouldn't buy the products of a grocery store because the store tries to get you to buy more.
> I dislike ads but realize they are a necessity for running some parts of the internet.
Can you expand on this? Why are they necessary for some parts?
People who believe ads are necessary have an underlying assumption that commerce is also a requirement. It's a shaky basis, and leads to a bunch of other messy things (ads to demographic to profiles to tracking)
No, they believe running a website has costs and if you aren't selling something your only other options seems to be ads or charity. The assumption is actually the opposite of commerce being a necessity. If you aren't selling something, that's when you need ads to stay afloat.
Not strictly necessary, but functionally necessary. Many websites provide services that are expensive to run and can't be funded by other products from the company, so the website itself has to be profitable. I'm not convinced that these websites (e.g. most social media) are societally necessary, but I don't see any scenario where they continue to exist without advertising (barring extreme measures such as government funding).
It's like casino hosting poker games, like a tournament. Between the players (advertisers) it's zero sum, because there are just so much money (end user attention and money), and house (ad platform) takes a cut in the proceeds.
Right, and casinos generally don't like poker because they can't really exploit it. They only host poker tables to get people in the door.
Yeah its almost like we shouldn't look to a page of AI slop for advice about how to run a business.
I use to think we lived in an enlightened age, where we have now learned that we need to use scientific reasoning and analysis to make decisions. It is what lifted us up from superstition and folk tales.
But society, still, collectively, falls for snake oil, over and over again.
This topic was talked about in 2019. https://news.ycombinator.com/item?id=21465873
Freakanomics did a great podcast about if Internet Advertising actually works. https://news.ycombinator.com/item?id=47906485
Anyone that has an interest in the topic should listen (or read) about what Steve Tadelis did in the 2012 timeframe. He was an economics professor at the University of California, Berkeley, and then worked at eBay, where he led the famous experiment on Google ad effectiveness.
People are born in ignorance. Reason takes energy. Nature optimizes for energy efficiency.
This is likely a hole us meat puppets can never crawl out of. Any moment we stop spending energy to climb out, we're sliding down the ant lion nest to ignorance.
Great analogy.
i can't square away all of this with the simple math that if you make a product that makes sense for online advertising, you make boatloads of money by using online ads, and make less as more people compete in it.
this is true of GLP1 ads and for many, many subscription services like video streaming.
i'm not sure it's about falling for something so much as it is that people who like, sell products through amazon, well, they are fucked no matter what.
Big advertisers (especially big retailers) are constantly employing fairly sophisticated methods to validate that they are getting a return on their investment with ads. It really is impossible to square this reality with comments reposting that one single paper from 2012 (?) (which doesn't even really say what they think it says---they ought to actually read it!) because those commenters have absolutely no clue what's going on.
The average IQ is 100. These expectations are totally unreasonable.
Anyone with an IQ in that range or lower absolutely needs a reliable source to just tell them what to do and how to live.
Social media and weaponized right wing "news" have set all those reliable sources on fire.
The trouble is people lacking wisdom more than people lacking intelligence.
Like lately I have been going to talks by Native Americans whose spirituality is rooted in place, relationships, commitment and all that
https://americanindian.si.edu/environment/pdf/01_02_Thanksgi...
and that somebody is looking seven generations (or even a seventh of a generation) ahead seems so strange in 2026 when so many people think they can enjoy being poor forever (like billions of years) if they don't manage to pound out 15 ai slop blog posts to prove to everyone else that they know where ai is going.
I don't subscribe to the idea that there ever was a reliable source of truth. It's just that having open access to information makes it so people see competing versions of framings some of which are incompatible with each other. For example, Christianity, Judaism and Islam can't all be true. They could all be false, however. Same applies to anything else. Just because you clearly know that this other view is wrong doesn't make your own perspective right.
Unfortunately or maybe fortunately it is entirely possible to live a long and happy life being entirely wrong about a great many important things without realizing.
I can't be sure that what I think is true, is really true. Therefore I don't really want to make the decision on how other people are supposed to live their life and find anyone that thinks himself worthy of doing just that suspect. I prefer someone that exploits people for their own gain, since then we are at least clear about the incentive structure and can measure and verify that gain and also negotiate for a cut. Something that isn't really possible when we are told to do things for our own good because God said so.
For a long time I’ve been chewing on this…
Back in the 2000’s I was involved with numerous “social” systems before there was Facebook, the thing they all had in common was that we knew we lived or died based on being viral so the most important knob we could turn in the process was the rate at which people completed the onboarding and recommended new people. So we were obsessive about plugging leaks in our funnel from both a craftsmanship and business standpoint.
Today whether it is mobile games or anything else I’m often left wondering if there is anybody left who remembers what it takes to tune up a funnel. Now I am told that for western mobile games they aren’t even interested in players who would care what the game mechanics are but instead just want to filter for a handful of ‘whales’ who will start spending money when they encounter the slightest speed bump… but still!
Right now I trade it back to the time when Facebook discovered its business model, made virality impossible, and convinced everybody to pay for traffic. You’d think the same motivations exist there, like you can save your money or your investors’ money if you have a good onboarding process so you should but what you find is…. People don’t.
I’ve had people blame the OKR/KPI culture for this which has everybody distracted by sub optimization…. But there’s also the problem that the thing you really need to compare customer acquisition cost to is lifetime customer value. So yeah, you spend $50 to get a customer to spend $20 where your COGS is $18…. That’s the part you know about and it looks awful and you tell yourself that that LCV will be more than $2 which is true because some of those customers will come back but it’s speculative. So you think you are practicing ‘quantitative marketing’ but you are like Coca-Cola in the mid 20th century spending on every kind of advertising because you think the brand equity is worth it.
I don't really think this is true, as if advertising didn't work people wouldn't use it.
That being said, there is something to be said about how extractive it has become.
It used to be you ran a TV ad, and if you made a great ad, you got amazing ROI, you were rewarded for being clever.
Now, it's all captured algorithmically, there are no "great ads", you can only have good ads or bad ads. They know full well if your ad is great then they can charge you whatever they want, inevitably making your ROI only "good enough". Sure go to another platform where the algorithm will price you out eventually.
Look at Him & Hers / Ro / GoodRx / etc. They basically take every dollar they make and throw it into advertising, it's the only way to stay relevant.
> if advertising didn't work people wouldn't use it
If animal sacrifice didn't appease the gods then people wouldn't do it, says the neolithic man leading his goat to the altar.
Was the neolithic man able to gather a statistically significant amount of data or is this just a bad analogy?
> if advertising didn't work people wouldn't use it.
Advertising works, and the unfortunate end result for fungible products is that the product price increases to pay for the advertising.
Producer A and B each cover half the market and compete on price. When A starts advertising to get more market share, the loss is for producer B who's margin decreases due to smaller volume. B also starts advertising. To make ends meet A and B have to increase product price to cover the cost of advertising.
This is very simplified reasoning. The gist is: because advertising works, consumers have to pay more for their products. The more advertising there is, the more prices will increase without products becoming better.
You wrote > I don't really think this is true, as if advertising didn't work people wouldn't use it.
In another post I pointed to a great podcast by Freakanomics. https://freakonomics.com/podcast/does-advertising-actually-w...
It is worth checking out. Here is a relevant couple of quotes...
HWANG: I think a second thing is, a little bit like in the subprime-mortgage crisis, you do have people who have very perverse incentives, I think, to push the effectiveness of online ads. That’s the ad agencies, the ad platforms themselves, the people who run ad technology. All these people, I think, have a very strong incentive to say, “No, this stuff is way better than earlier generations of advertising, and this is why you should use it.”
FE: If you’ve been listening closely, you’ll notice this is the exact same problem Steve Levitt talked about regarding the TV-ad ecosystem: human beings generally make decisions based on self-interest.
LEVITT: No chief marketing officer is ever going to say, “Hey, I don’t know, maybe ads don’t work. Let’s just not do them and see what happens.”
FE: Or, as the author Upton Sinclair once wrote: “It is difficult to get a man to understand something when his salary depends on his not understanding it.”
An AI written article with a bad analogy
Claude doesn't understand that the net revenue per conversion (including variable revenue minus all other expenses) is different for everyone bidding on the same ad slot.
The gambling analogy doesn’t quite hold here. Gambling is a zero sum game. Advertising is not a zero sum game. There is room for both sides to walk away with gains from trade.
The article explains why that isn't the case: it's because advertisers (Google's clients who run ads) are competing with each other to push the cost of advertising up to the point where the cost of an ad and the value it provided reach a rough equilibrium.
this is competition, not gambling.
And in economic theory, perfect competition drives the margin to zero. The gambling part is not explaining that theory, but rather the effect on human behavior when the short term results hover around that zero margin line.
The margin for the marginal participant! The more profitable participants will still make a profit. You need some argument like "the profitable participants are going to seek to expand their business until it isn't profitable anymore" to truly eradicate profits in theory.
If you look at the oil business, for instance, the Saudis have a much lower cost than anybody else in the business so they could be profitably selling oil in a world where everybody else is losing money.
I don't deny that the article is an exaggeration of the situation. Just wanted to point out that the criticism is the combination of
1) market making mechanisms implemented by the platforms drive down ROI
2) when ROI is around 1, irrational behavior can look similar to gambling
Also, where's the proof that ad platforms do not increase prices based on estimated profitability, previous willingness to pay, etc.?
The gambling part is how the competition between advertisers is implemented.
> advertisers (Google's clients who run ads) are competing with each other to push the cost of advertising up to the point where the cost of an ad and the value it provided reach a rough equilibrium.
Yup. That's econ 101, except that the price you pay is not the value you would have gotten, it is the price one bid increment higher than the value your competition would have gotten/was willing to pay. You can still have consumer surplus on your side of the trade as long as the sale is a proxy or vickery auction - like most ad sales are.
> You can still have consumer surplus on your side of the trade as long as the sale is a proxy or vickery auction - like most ad sales are.
You can still make a profit on your side of the trade if it's a regular "winner pays what they bid" auction too. No one can stop you.
People can bid more than the ad is worth to you, and stop you from running the ad at all, but they can't make you bid more than the ad is worth and take a loss.
That wasn't the analogy. It's gambling in the sense "the house always wins".
It's not a simple "both sides" issue - the scenario has multiple sides. You find a good keyword, someone finds you, you both benefit. Great!
Then a third party finds and uses your strategy, and bids higher in the ad auction. This reduces profits. As more parties join in, the profits decrease.
The thesis is that those profits don't simply go to zero, but can actually result in a net loss. In the meanwhile, those higher bids end up as profits for the auction house.
Uh, yeah the house always wins because it is zero sum. In a competitive market there is gains from trades. Rational actors can drive consumer and producer surplus to zero but once rational actors won't continue to buy at a loss.
I'm picking on the gambling analogy. It is wrong.
I agree with the general thesis of the article.
Ahh, I see. I've only dealt with zero sum games with two parties.
I was under the impression that most auction houses are not zero-sum games. The auction house makes a profit, but the buyer and seller also win.
It's often constant sum, because in ads can move sales between organisations instead of creating value.
With the automated bidder maximising ad prices for shareholder returns it can become zero sum equivalent. Every gain is matched by a loss elsewhere.
This changes if ads are creating new demand, or if consumers are getting value they otherwise wouldn't have found. But in many markets that won't be the case.
This means is often
Gambling, advertising, trading, and AI have one thing in common: if you don’t do well it’s your fault. You didn’t prompt right, or target the right audience, or you don’t have the right harness, and the only way to remedy is always to spend more money.
The only way to keep out of those traps is to have real metrics of your own and benchmarks of your own. Performance changes over time so you can never trust the current state will continue.
For a long time I’ve thought that Google ads were a tax on businesses. You have to pay it, everyone does, but the return was _what_?
Psychological elements aside, this sort of misses that goods and services require multiple inputs from different counterparties (or different markets) that have different pricing power (some of which might be close to monopolies). It’s implausible that whenever there is one monopoly seller anywhere that all added economic value (or more than all, as in the article??) has to go to that one.
If I have to necessarily buy ads from google and water from the local water company, where does the economic value go in equilibrium? Clearly not all to the water company and not all just to google.
That’s just victimhood mentality. There are tons of examples of companies printing with ads. I’m not saying that there aren’t players that play net-negative games. And those playing hurt everyone but the advertising sellers, but the game is not fundamentally flawed.
The whole ecosystem is extractive with the goal of improving quarterly earnings by removing control from the people spending the money. Online ads started out being trackable and much higher ROI than traditional media. Then they ran traditional media out of business while continously raising their own rates (opaque quality score, nebulous targeting options) and then go even further, essentially stealing publishers' content for their own walled garden models.
So glad not having to look at this damn dashboard and all its myriad of tools, interconnects and whatnots, which even claim to be able to attribute a sale to some bizarre ins and outs of the search. Of course if all cookies, tags, and pixels that you have to install beforehand work. And also validations, and confirmations, legal checks, etc, etc. It reaaally is a giant slot machine. Really, first you have to be very lucky to get an actually good connectivity of all the needed info, and then you need to see Google in its lucky day actually driving this traffic that your struggling business so much needs.
As my godfather put it once - dude, you better give $10 each customer to come and buy, rather than spend $420 to get three. It will work better.
I very much hope this all dies for good with AI, something tells me we're in for v2 where all the bullshit is injected into LLM streams, and then other local LLMs would need to adblock it. Sad story. Not the internet I got fascinated about as a kid, tempting to attribute it to age, but there was surely more awe and fun in it than this never-ending slop of irrelevancy, and I'm not even talking AI.
They do not present any data at all to support their 98% figure.
There's too little discussion out there about how scammy the online advertising is. Wish there was a reckoning.
I tried social media ads years ago for my project, and it was all bots. Wasted hundreds of dollars, so I'm afraid to try again today now that there is AI.
Still blows my mind that people pay to advertise to bots. Dead internet theory and all that
"Half the money I spend on advertising is wasted; the trouble is I don't know which half.” John Wanamaker (1838-1922)
A slot machine knows how much income the user is receiving because the slot machine hands it to them.
But in this model, the small business's revenue stream isn't (in most case) something that the ad networks can see.
I drive by an Ikea on my way to work. They just installed a video wall in view of the highway. It is huge, four stories tall and square. It gets noticed but is not intrusive. It is not animated. In daylight it just looks like a big painting. At night it isn't any brighter than the parking lot around it. Yesterday it was advertising an end table for 65$. This morning it was something with a plant in the background. I actually saw and remember that ad. I almost never see ads on websites and certainly don't remember those I do. But that Ikea wall... THAT I remember.
Online adds are a fool's errand.
Business are trees. Customers are sunlight. Advertisements are the trunks.
Incredibly wasteful and as tall as they can physically be. If trees could talk to each other and decide to limit the heights of the trunks arbitrarily there would be so much energy saved that could be put into seeds and fruit to spread the forest faster.
But the trees of course can't agree because any defector would gain advantage over the cooperating trees.
In case of the economy the government should step in and replace the whole advertising industry with a single always up to date catalog of goods and services. If comapnies couldn't get ahead by inflating marketing budgets they would be forced to compete on quality and price instead what would be beneficial for everything including the health of those companies.
It's an auction. It's not a conspiracy.
Every marketing channel has its pros and cons. Advertising is no exception.
Amazon got fined a bajillion dollars for deciding to add undisclosed reserve prices for their ad auctions. They just decided that too many auctions ended too low, so they added fake reserve prices and just chose what the price would be...
Google and Meta have also both been caught in advertising shenanigans -- I forget the details but I recall getting some (tiny) class action settlement check.
Won't somebody think about the founders??