Facebook is engaged in a giant crime spree to steal ad money. A battle over speech in Australia shows what top executives really think of the rule of law.
I work in the tech industry and have wondered if these ad stats are just made up. Probably not made up of whole cloth, but using vague terminology like "impression" to generate whatever numbers they feel like generating. Kind of crazy that Google and Facebook are two of the world's most valuable businesses and their customer value may not actually exist. No one really knows. I can't think of a single time, even with hindsight, where I have been influenced by a Google or Facebook ad. They, of course, say that I can't remember an occasion where their ads have worked as evidence of how great their ads are... operating at a subliminal level. That is how you play tennis without a net. It cannot be falsified. The only thing they offer customers are impression numbers, possibly made up impression numbers. The entire industry is based on faith alone... if advertisers stop having faith that these ads are actually making a difference, the value of the whole thing falls to zero.
Didn't you know 90% of statistics are made up? The saying about statistics is that you can make them say whatever you want. The only way for businesses to know how effective the ads are, and therefore if they're worth paying for, is to find out from their customers how they heard about them, whether through surveys, codes in the ads (like in podcasts), or some other method. Without doing that, they're just blindly spending money based on nothing more than the marketing prominence of these platforms. So if they're tracking how their different marketing strategies are working on their own, they'll still have a pretty good idea of how effective it is regardless of whether they're being lied to. Of course, if they're being lied to about how many people are being reached, then even if it's still an effective advertising outlet they're overpaying for it. In other words, they should be taking it upon themselves to be monitoring the effectiveness of using Google and Facebook advertising to see if it's making them money, but they have no way of knowing whether the business those ads brings in is due to the reach claimed by Facebook and Google or simply because they're getting a good response to their ads even with a smaller number of users.
Yeah, I think the question is magnitude. I'm sure putting out AdWords in front of millions of people has some impact, but are people paying $10 million in advertising for $10,000 in new, incremental sales which would not have occured but for the ads.... There is kind of two way conspiracy here though. Google and fb are probably massively overcharging, but marketers are probably not anxious to go tell management that none of their marketing campaigns are worth the cost.
To add to this, I saw a very apropos billboard today that reminded me of this. It said 83% of consumers say they read billboards. I assume it was advertising for billboard advertising, though I suppose it could be some sort of satire, though the real irony was that it was probably only the second or third billboard I'd looked at in an hour plus of driving. But hey, even if I only look at 0.5% of billboards, I guess I technically "read billboards" and therefore fall into that 83%. Just another great example of statistics manipulation.
This describes well what a good friend working in social media marketing has told me about the industry. There is little to no transparency about how these numbers are computed. For example, I recall a time when Facebook would autoplay every video ad on the news feed. Every user would swipe right past it, but that half second was enough to include it as an nebulous impression. To call that an adview is deceptive if not fraudulent.
To make matters worse, she worked for a non profit that did not have the resources to independently verify the effectiveness of their outreach campaigns. Smaller companies and are at an inherent disadvantage when the two big ad players in town are goosing their numbers. My friend would work within that context, but it was difficult for her to make any sort of informed decision.
Regarding Logitech, I've long had issue with them, or more specifically, their products, software, and service. For at least the past decade, I've periodically (at least once every 2-3 years) looked for alternatives to some of their products and have always come up empty-handed. I have and use a Logitech keyboard, mouse, and Harmony remote, but they all have issues. My opinion of them, and a friend of mine holds the same opinion, is that they're great at making products that are superficially great, with innovative and useful features and well-designed from an ergonomic standpoint, but the quality is terrible. EVERY...SINGLE...ONE of their products I've owned has had issues. I bought three of the same model mouse and every one had the same *exact* problems, but I just deal with it because it's the only mouse with the features it has. I broke one of the keycaps on my keyboard so it won't stay on, and Logitech doesn't offer replacements. The only way to "fix" it is to buy a whole new keyboard.
When selecting which Harmony remote to get, I decided against one of the touchscreen versions because there were tons of reviews for all of them saying the touchscreen fails or breaks, and the model I chose, I did so partly due to the fact it could control Hue lights. After several months of ownership, however, Logitech released an update that broke that functionality, and it took them *months* to fix it, during which time they didn't even bother to provide the ability to downgrade to the previous, unbroken version. I also had to get a replacement remote after a couple years because one of the buttons stopped working. As the person who wrote about them mentioned, their software is atrocious, and I've often just dealt with things not working right so I wouldn't have to deal with it because it's just that bad. And the latest issue is that after not using it for a while, when I hooked it up in a new place with a different setup, I couldn't get it to work.
But every time I've looked, I haven't found anything close. Mostly just a bunch of cheap remotes that are huge and don't have the same features or abilities. I keep hoping that with newer technology there will be something, but there never is. Logitech has been strangling this product segment for years. I just always assumed it was because of their dominance and name recognition, which deterred any other companies from bothering, but it wouldn't surprise me if there's anti-competitive practices at play as well.
And I've noticed similar issues all across the market, in numerous product categories, where it's hard to find something within certain criteria either because there's very few companies, and often only one, making that type of product, at least at a certain level (i.e. not counting the countless cheap, Chinese versions) or there's a (still usually quite small) handful that are all making products very similar to each other. The lack of competition brought on by monopolies doesn't just affect prices and quality, it affects variety.
Hi Matt - regarding NVIDIA / ARM, I encourage you to learn more about Jim Keller. Long time hardware designer for Apple who recently started a company that's focused on Graph (not Graphic) cards for AI applications. He did a podcast recently that delved into the near history of chip design and deals with some anti-competitive issues. https://www.youtube.com/watch?v=G4hL5Om4IJ4
I'm an Australian software developer and want to weigh in and provide a perspective on the News Media Bargaining Code that you discuss Facebook/Google fighting. I don't believe that the legislation has any positive impact on monopolisation at all.
First of all, I am no fan of any of the entities involved, and as a disclaimer think that some of the requirements in the legislation to share algorithm updates selectively are incredibly flawed and problematic. However those disclaimers aside, the anti-Facebook popular sentiment current is allowing the Australian Liberal Party Government to get away with overreaching legislation that is designed to prop up a different monopoly in News Corp.
The intent of the initial bill as written can be seen by the caveats that you mentioned briefly. Disallowing public broadcasters to be paid for their work makes no sense unless you are trying to hobble public broadcasters even further. The government has cut the Australian public broadcasters budgets significantly, and now through this legislation ensures that only their private sector competitors are able to receive any income from this? News Corp currently controls approximately 70% of Australian media, and over half a million Australians signed a petition for an inquiry into this situation and this inquiry is currently ongoing.
An important distinction can also be made between how Google and Facebook use news media in their platforms. Google products can show a version of the content that has been scraped from the news site on their own site through snippets or AMP. This prevents the original news content owner being able to monetise this content, and as such potentially should be paid for as Google is instead able to monetise. Facebook however shows only the content that the news media, or any website, has asked for them to surface with meta tags on the page, designed for sharing!
The current backlash against Facebook has meant that commentators are ignoring the fact that this legislation can be used to decide exactly what platforms are required to pay or enter these deals with publishers.
Finally, the most egregious issue with the bill is how it rewards monopoly behaviour by News Corp and does it's best to entrench it. Companies that are not eligible for bargaining under this legislation would have to compete with an ever widening moat, especially in the case where their competitors are provided with updates to ranking algorithms and they are not, on top of the financial remuneration.
Surely the solution to the problem of a monopoly in one area, cannot be the entrenchment of monopoly in another area? That sounds like the government picking winners, in this case the very Liberal party friendly News Corp media.
"The intent of the initial bill as written can be seen by the caveats that you mentioned briefly. Disallowing public broadcasters to be paid for their work makes no sense unless you are trying to hobble public broadcasters even further. "
The only reason News Corp (or any commercial media) would support non-government funding for the ABC & SBS, is so they can then run the argument that the government should cease all public funding for the now "independently" funded public broadcasters.
There's been a long history of News Corp (and Nine Media - the other side of the Australian news duopoly, and headed by a former Liberal Party Treasurer) demanding public funding to the ABC & SBS be reduced, and their services deliberately constrained to stop them, in their parlance, consuming all the oxygen in the room for news content.
As per usual with radical theocratic capitalists, News & Nine have long advocated the position that public broadcasting is the theft of a profit opportunity from the private sector, by the government.
Much as I despise Facebook, and use plugins to remove Facebook's tracking url additions from links on FB, and content blockers to disable the "share" button trackers on 3rd party sites, I'm still having trouble understanding the logic of why anyone would support this legislation - from the simple perspective, of why Facebook should pay The Guardian, when I put a link to a page on The Guardian's website on Facebook, which requires a person to then go to The Guardian's website to read the article.
It's not The Guardian generating the Facebook traffic (and subsequent advertising revenue / metric) of reading my post (which is my comment on the article), it's me. And it's not The Guardian which facilitates the person getting to The Guardian's website to read the article, it's Facebook - so what has The Guardian actually contributed to the process? Yes, they wrote the original article, but surely, their monetisation is supposed to come from the consumption of the article itself, on their website, from their advertisers.
By The Guardian's logic in this case - they should pay me, to place an advertorial preview / review for my art exhibition, because *handwaving* writing about my art exhibition provides them with reader-engaging content they can sell ads against?
Further to this, IF the goal of this legislation was to hobble the power of an abusive monopoly (which noone in Australia believes it is - ask around here, you'll find we're mostly very cynical, given the current government is a News / Nine puppet, that this is about subsidising the shareholders of News & Nine, and not much else)...
...then the place to direct efforts, whether it's Facebook, Twitter, Apple, or Google is in breaking the places where these businesses integrate horizontally and vertically, by forcing them to openly document the exchange points of their ecosystems, so that any competitor can provide a drop-in replacement for any part of their business.
Facebook is abusing their monopoly on advertising by falsifying their distribution numbers? Then Facebook should be required to restructure their metrics engine as a separate service, so that any 3rd party metrics vendor can plug into an advertiser's page, and see the exact same data Facebook can see.
Facebook abusing their customers by forcing an unwanted algorithmic feed upon them, which has been shown to radicalise people - then you force Facebook's back end database to openly document its entire structure, so that 3rd parties can make drop in replacements for Facebook's website & apps, that allow for people to have the feed they want (everyone they follow, in reverse chronological order).
Monopolies are broken not by taking away the spoils of monopoly (fines and taxes), but by taking away the synergies of lock-in that make the whole greater than the sum of its parts, for whomever holds all the parts.
Thanks, that's a useful picture, seeing both sides. One thing to watch with Australia federal govt is that the regulators and monitoring agencies, ACCC, Productivity Commission, ANAO and to a lesser extent ASIC, do great investigation, as far as their budget allows. Many of their report recommendations have languished for years due to lobbying, as is the case.
The part that this piece is very coy about, and the legislative motives are very coy about, is what it is that Facebook does with the news that they allegedly need permission for. Everyone always just talks about Facebook 'using' news, or, as you put it, 'distributing' news. But what does that mean? According to the bill, it means: linking to news. We should be clear about that.
Now, proponents of the bill argue that Facebook is abusing its market power. Fair enough: If Facebook had to get a license for linking to news articles, then you could perhaps make a case that they keep prices down by (ab)using their market power. In reality, however, they do not have to get permission for linking to a news article at all. So how can they abuse their market power? There is no market! The price is zero because there's nothing to trade, not because Facebook has a dominant position.
The EU basically acknowledged as much. Their approach is to protect small snippets (which might, according to some, include even the title of a news piece) under a novel neighbouring right in the copyright acts of EU member states. By doing that, they create a market for those snippets. The Australian approach entirely skips this step. Everyone is talking about an abuse of market power on a market that, in reality, simply does not exist. If A lacks a legal means to prevent B from doing X, then B will pay exactly zero dollars for doing X, no matter if B dominates the market or not. So it is very strange to hear people talk about Facebook not paying for 'using' news when all they do is something that everyone is allowed to do without permission and at no cost.
The ads issue (explained in the Srinivasan paper) is an entirely separate animal, which we should vigorously tackle using the tools of anti-competition laws. It has nothing to with linking to news articles, though. Publishers get their revenue from ads on their own site. If Google et al rig the ads market and rip off publishers, you have a different issue -- it is wholly unrelated to the link itself. Accordingly, it affects news publishers (and every other content provider!) equally no matter where the traffic is coming from.
"Facebook uses a specific definition for an external impression: the number of times a unit was on screen. This is technically defined as 1x1 pixel of a post or piece of content in a user's viewport, for >0 than seconds. (We don't reveal this definition externally.)"
You can't come close to an admission that they were defrauding their paying customers all along than this. Facebook finally disclosed this inane definition in the fall of 2019. It appears that my complaining internally may have had some effect.
I think the entire premise of the Australian law is misguided, as it misjudges the nature of Google and Facebook's duopoly. I argue that it is specifically targeted advertising – and the invasive tracking that it demands – that form the true monopolistic secret sauce of both Google and Facebook.
First, I think the new law should be very concerning to anti-monopoly advocates. It crosses into the murky territory of trying to legally define a specific class of business and to bolster it at the expense of another class of business. This should be considered a strong legislative anti-pattern, a band-aid that risks choosing monopolistic winners versus addressing structural root causes. Furthermore, I agree with analyses that laws in this pattern could be extremely damaging to both the internet and society at large.
I've read recent thinkpieces on this topic that present the argument that Google and Facebook's monopoly "stole" something from traditional news outlets. This feels like magical thinking, and ignores the strong argument that the internet's alteration of information distribution is a deep root cause of news media's struggles. Print-era news media has permanently lost its /scarcity-based monopoly/ on both content and advertising, and must now compete with all-comers across the globe. This isn't an unalloyed good, as we're discovering the lost value in the focus those resources provided. However, I also think there's no sensible or desirable way to restore that world gone past - to put Humpty-Dumpty back together again.
I do, however, think there's hope to strike a new balance. Targeted advertising and invasive tracking are the technological moats upon which the Google / Facebook duopoly rests, not the readily linkable nature of the internet. It's worth recalling that prior to this duopoloy, there were a quite a number of independent ad networks, with different levels of use and explicit non-use of tracking tech. The big, successful tracking firms largely consolidated, perhaps culminating in doubleclick.net's purchase by Google. The non-tracking networks largely died out as tracking-based advertising sucked the air out of the room.
What if instead legislation attempted to eliminate targeted advertising? This would destroy the great technological moats of both ad giants in a way that doesn't need to call out specific business categories. Organic advertising by popular web outlets, news or otherwise, begins to look viable again. I'll argue that this change would help foster an increased number of independent ad networks to connect ad buyers with web properties. It's also quite possible that this change may not be complete, but I'll wager the worst case for eliminating tracked advertising is "necessary but not sufficient."
I'm with you. Just eliminate targeted advertising. Do it for privacy reasons. Or turn it on it's head and make it strictly opt-in instead of opt-out. Another issue is that no typical person really understands how all this works. No one realizes how valuable the data is that they are giving away for free (well, you can get an idea from google and facebook market cap). Everyone is scared of big brother government. Google and Facebook know EVERYTHING about me. More then I do myself. Third I have worked with Google Ads for years. It has become harder to use and understand over the years. They keep pushing you towards their automated solutions which are a complete black box. And guess how you measure results. With another google product Analytics. Every year I make the joke resulting in nervous laughter is that Google can literally tell us almost any numbers in adwords and analytics if they wanted to. We just have to trust on good faith that they don't. Also who clicks on these digital ads? I never have nor anyone I know. Anecdotal I know. But still. We have gone full circle to where this is no different then general "branding" ads on TV or in a newspaper. I suspect the "targeting" is way less effective then they will ever let on. But no one reads newspapers or watches commercials anymore so this is what we have. I don't know if Facebook and Google turn the dial in their favor. But it would be unbelievably easy for them to do. I suspect a lot of these big tech issues (and goverment conspircy issues) could be solved with real forced radical transparency. But how you accomplish this I have no idea.
"But no one reads newspapers or watches commercials anymore so this is what we have."
If it didn't interrupt something else I was watching I would genuinely spend 5 minutes every so often just watching commercials for the entertainment and to be informed about current products.
Years ago when Hulu was somewhat worth watching, they allowed you to watch short commercials periodically throughout a show or one long one at the beginning. I liked that and always elected for the long one, to just get it out of the way. Of course, I'm sure most people would do that then go to the bathroom or something, so not sure how effective it was, but it was a refreshing change.
What I find amazing is there's all this resistance by most people, myself included, toward commercials, and advertisers spend all this money on them, and they interrupt shows with them in order to force people to watch them, but then people go out of their way to watch Super Bowl commercials online or even in the form of an actual show that just shows various commercials. What that tells me is that the problem, or rather solution, of reaching people isn't in making them watch commercials during their shows, but by actually making good commercials that are entertaining and make people *want* to watch them.
I don't know enough about how advertising works, especially in this context, to have much of an opinion, but I'll offer a couple counter-points just for consideration. First, while this is certainly not the best way to deal with the problem, as it's more a treatment of a symptom than the disease, and, as you pointed out, it would be much better to deal with the root causes, it's certainly better than nothing, and Australia is limited in what they can do, so they're just doing what they can. It may not be ideal, and it may even be troubling to some, but they don't have many options, and it's better than doing nothing. The fact is, something should have been done years ago, and if everybody just keeps waiting for something to be done, nothing will ever be done. Someone has to take action. And perhaps this won't even need to be a permanent thing, but will serve as a kickstart to motivate others to take action. After all, as Matt pointed out, it caused Facebook to reveal just how much power they have and turned public opinion even more against them. While I'm not one for generally believing the ends justify the means, in this case I'm inclined to think they do.
The second point is that media has always, with few exceptions, been largely ad-funded. Newspapers made money primarily through charging for print ads, though they made some through subscriptions as well. Once the Internet took off and media started to transition there, it was the same, with most of their revenue being made through ads (just often much more annoying ones) and a small part through subscriptions. If their content is being shared on a third-party site, and that site is collecting ad revenue based on that content instead of, or even in addition to, the media outlets being able to do so, then clearly that's an issue that needs to be addressed. But then, I don't really care about the media outlets, and I think they've brought a lot of the issues they're currently facing on themselves, so for me it's more about reigning in Facebook and the others, at the very least.
Thanks again for the solid reporting in this niche.
In a previous post, you ask what possible benefits blockchain tech may have in our industries and markets. While I can't think of anything directly related to my field (higher education), the recent finance committee hearings regarding the Gamestop phenomenon, Robinhood, and limiting of retail equity orders brought up the first and strongest use case for the tech. It is an extremely reliable accounting system that has the potential to clear orders at a much faster rate and with less error than current wall street clearing houses. The benefits could (in my admittedly uneducated) lead to huge improvements in the quality of information accessible to individual retail investors. Questions like, what percent of an equity is short at any given moment or who has the capital to uphold their market orders right now should be easy to derive with this tech. Food for thought.
'Facecrook' or not, you have got to respect these guys. Australia has surrendered to it in 72 hours. Let that be a lesson to other governments that think of doing the same. You have no chance.
That Facebook ads crime is only the tip of the iceberg. Online ads is probably the biggest scam since the advent of the Internet. It’s an opaque business where the few Goliaths is the house controlling all the cards, where the customers have no choice but believe whatever they claim, on fear of missing out. And the customers have no way to verify any of the claims since the technology is proprietary, complex and opaque.
Search the web for the excellent article called “The new dot com bubble is here: it’s called online advertising”.
Anecdotally, when I surf the web, I almost never use bookmarks or type a site’s address; I google instead. So when I want to visit an online store, for example, I google by store name. The first “result” is almost always the ad for the store. If I click on it, does that mean the ad was effective? Of course not. The second result is the actual top search result and the URL is exactly the same as the one for the ad. I intentionally avoid clicking the ad but I’d imagine most users would click the ad and contribute in “falsifying” the effectiveness of the ad.
Great article. Just a small correction to the NVIDIA segment. NVIDIA doesn't only use Taiwan Semiconductors to manufacture their chips. In fact, with the current generation of consumer Ampere cards (30 series), they are manufactured on Samsung's fabs. Their prosumer and commercial cards are made on TSMC's fabs however, as are all of AMD's and Apples chips. Doesn't mean there isn't a shortage, still is with nearly every fab in the world working overtime to keep up with demand, and still failing to do so.
I work in the tech industry and have wondered if these ad stats are just made up. Probably not made up of whole cloth, but using vague terminology like "impression" to generate whatever numbers they feel like generating. Kind of crazy that Google and Facebook are two of the world's most valuable businesses and their customer value may not actually exist. No one really knows. I can't think of a single time, even with hindsight, where I have been influenced by a Google or Facebook ad. They, of course, say that I can't remember an occasion where their ads have worked as evidence of how great their ads are... operating at a subliminal level. That is how you play tennis without a net. It cannot be falsified. The only thing they offer customers are impression numbers, possibly made up impression numbers. The entire industry is based on faith alone... if advertisers stop having faith that these ads are actually making a difference, the value of the whole thing falls to zero.
Didn't you know 90% of statistics are made up? The saying about statistics is that you can make them say whatever you want. The only way for businesses to know how effective the ads are, and therefore if they're worth paying for, is to find out from their customers how they heard about them, whether through surveys, codes in the ads (like in podcasts), or some other method. Without doing that, they're just blindly spending money based on nothing more than the marketing prominence of these platforms. So if they're tracking how their different marketing strategies are working on their own, they'll still have a pretty good idea of how effective it is regardless of whether they're being lied to. Of course, if they're being lied to about how many people are being reached, then even if it's still an effective advertising outlet they're overpaying for it. In other words, they should be taking it upon themselves to be monitoring the effectiveness of using Google and Facebook advertising to see if it's making them money, but they have no way of knowing whether the business those ads brings in is due to the reach claimed by Facebook and Google or simply because they're getting a good response to their ads even with a smaller number of users.
Yeah, I think the question is magnitude. I'm sure putting out AdWords in front of millions of people has some impact, but are people paying $10 million in advertising for $10,000 in new, incremental sales which would not have occured but for the ads.... There is kind of two way conspiracy here though. Google and fb are probably massively overcharging, but marketers are probably not anxious to go tell management that none of their marketing campaigns are worth the cost.
To add to this, I saw a very apropos billboard today that reminded me of this. It said 83% of consumers say they read billboards. I assume it was advertising for billboard advertising, though I suppose it could be some sort of satire, though the real irony was that it was probably only the second or third billboard I'd looked at in an hour plus of driving. But hey, even if I only look at 0.5% of billboards, I guess I technically "read billboards" and therefore fall into that 83%. Just another great example of statistics manipulation.
This describes well what a good friend working in social media marketing has told me about the industry. There is little to no transparency about how these numbers are computed. For example, I recall a time when Facebook would autoplay every video ad on the news feed. Every user would swipe right past it, but that half second was enough to include it as an nebulous impression. To call that an adview is deceptive if not fraudulent.
To make matters worse, she worked for a non profit that did not have the resources to independently verify the effectiveness of their outreach campaigns. Smaller companies and are at an inherent disadvantage when the two big ad players in town are goosing their numbers. My friend would work within that context, but it was difficult for her to make any sort of informed decision.
Regarding Logitech, I've long had issue with them, or more specifically, their products, software, and service. For at least the past decade, I've periodically (at least once every 2-3 years) looked for alternatives to some of their products and have always come up empty-handed. I have and use a Logitech keyboard, mouse, and Harmony remote, but they all have issues. My opinion of them, and a friend of mine holds the same opinion, is that they're great at making products that are superficially great, with innovative and useful features and well-designed from an ergonomic standpoint, but the quality is terrible. EVERY...SINGLE...ONE of their products I've owned has had issues. I bought three of the same model mouse and every one had the same *exact* problems, but I just deal with it because it's the only mouse with the features it has. I broke one of the keycaps on my keyboard so it won't stay on, and Logitech doesn't offer replacements. The only way to "fix" it is to buy a whole new keyboard.
When selecting which Harmony remote to get, I decided against one of the touchscreen versions because there were tons of reviews for all of them saying the touchscreen fails or breaks, and the model I chose, I did so partly due to the fact it could control Hue lights. After several months of ownership, however, Logitech released an update that broke that functionality, and it took them *months* to fix it, during which time they didn't even bother to provide the ability to downgrade to the previous, unbroken version. I also had to get a replacement remote after a couple years because one of the buttons stopped working. As the person who wrote about them mentioned, their software is atrocious, and I've often just dealt with things not working right so I wouldn't have to deal with it because it's just that bad. And the latest issue is that after not using it for a while, when I hooked it up in a new place with a different setup, I couldn't get it to work.
But every time I've looked, I haven't found anything close. Mostly just a bunch of cheap remotes that are huge and don't have the same features or abilities. I keep hoping that with newer technology there will be something, but there never is. Logitech has been strangling this product segment for years. I just always assumed it was because of their dominance and name recognition, which deterred any other companies from bothering, but it wouldn't surprise me if there's anti-competitive practices at play as well.
And I've noticed similar issues all across the market, in numerous product categories, where it's hard to find something within certain criteria either because there's very few companies, and often only one, making that type of product, at least at a certain level (i.e. not counting the countless cheap, Chinese versions) or there's a (still usually quite small) handful that are all making products very similar to each other. The lack of competition brought on by monopolies doesn't just affect prices and quality, it affects variety.
Hi Matt - regarding NVIDIA / ARM, I encourage you to learn more about Jim Keller. Long time hardware designer for Apple who recently started a company that's focused on Graph (not Graphic) cards for AI applications. He did a podcast recently that delved into the near history of chip design and deals with some anti-competitive issues. https://www.youtube.com/watch?v=G4hL5Om4IJ4
Thanks!
I'm an Australian software developer and want to weigh in and provide a perspective on the News Media Bargaining Code that you discuss Facebook/Google fighting. I don't believe that the legislation has any positive impact on monopolisation at all.
First of all, I am no fan of any of the entities involved, and as a disclaimer think that some of the requirements in the legislation to share algorithm updates selectively are incredibly flawed and problematic. However those disclaimers aside, the anti-Facebook popular sentiment current is allowing the Australian Liberal Party Government to get away with overreaching legislation that is designed to prop up a different monopoly in News Corp.
The intent of the initial bill as written can be seen by the caveats that you mentioned briefly. Disallowing public broadcasters to be paid for their work makes no sense unless you are trying to hobble public broadcasters even further. The government has cut the Australian public broadcasters budgets significantly, and now through this legislation ensures that only their private sector competitors are able to receive any income from this? News Corp currently controls approximately 70% of Australian media, and over half a million Australians signed a petition for an inquiry into this situation and this inquiry is currently ongoing.
An important distinction can also be made between how Google and Facebook use news media in their platforms. Google products can show a version of the content that has been scraped from the news site on their own site through snippets or AMP. This prevents the original news content owner being able to monetise this content, and as such potentially should be paid for as Google is instead able to monetise. Facebook however shows only the content that the news media, or any website, has asked for them to surface with meta tags on the page, designed for sharing!
The current backlash against Facebook has meant that commentators are ignoring the fact that this legislation can be used to decide exactly what platforms are required to pay or enter these deals with publishers.
Finally, the most egregious issue with the bill is how it rewards monopoly behaviour by News Corp and does it's best to entrench it. Companies that are not eligible for bargaining under this legislation would have to compete with an ever widening moat, especially in the case where their competitors are provided with updates to ranking algorithms and they are not, on top of the financial remuneration.
Surely the solution to the problem of a monopoly in one area, cannot be the entrenchment of monopoly in another area? That sounds like the government picking winners, in this case the very Liberal party friendly News Corp media.
"The intent of the initial bill as written can be seen by the caveats that you mentioned briefly. Disallowing public broadcasters to be paid for their work makes no sense unless you are trying to hobble public broadcasters even further. "
That's not Rupert Murdoch's doing. https://www.theguardian.com/media/2020/oct/02/news-corp-backs-push-for-abc-and-sbs-to-receive-payments-from-google-and-facebook
The only reason News Corp (or any commercial media) would support non-government funding for the ABC & SBS, is so they can then run the argument that the government should cease all public funding for the now "independently" funded public broadcasters.
There's been a long history of News Corp (and Nine Media - the other side of the Australian news duopoly, and headed by a former Liberal Party Treasurer) demanding public funding to the ABC & SBS be reduced, and their services deliberately constrained to stop them, in their parlance, consuming all the oxygen in the room for news content.
As per usual with radical theocratic capitalists, News & Nine have long advocated the position that public broadcasting is the theft of a profit opportunity from the private sector, by the government.
Much as I despise Facebook, and use plugins to remove Facebook's tracking url additions from links on FB, and content blockers to disable the "share" button trackers on 3rd party sites, I'm still having trouble understanding the logic of why anyone would support this legislation - from the simple perspective, of why Facebook should pay The Guardian, when I put a link to a page on The Guardian's website on Facebook, which requires a person to then go to The Guardian's website to read the article.
It's not The Guardian generating the Facebook traffic (and subsequent advertising revenue / metric) of reading my post (which is my comment on the article), it's me. And it's not The Guardian which facilitates the person getting to The Guardian's website to read the article, it's Facebook - so what has The Guardian actually contributed to the process? Yes, they wrote the original article, but surely, their monetisation is supposed to come from the consumption of the article itself, on their website, from their advertisers.
By The Guardian's logic in this case - they should pay me, to place an advertorial preview / review for my art exhibition, because *handwaving* writing about my art exhibition provides them with reader-engaging content they can sell ads against?
Further to this, IF the goal of this legislation was to hobble the power of an abusive monopoly (which noone in Australia believes it is - ask around here, you'll find we're mostly very cynical, given the current government is a News / Nine puppet, that this is about subsidising the shareholders of News & Nine, and not much else)...
...then the place to direct efforts, whether it's Facebook, Twitter, Apple, or Google is in breaking the places where these businesses integrate horizontally and vertically, by forcing them to openly document the exchange points of their ecosystems, so that any competitor can provide a drop-in replacement for any part of their business.
Facebook is abusing their monopoly on advertising by falsifying their distribution numbers? Then Facebook should be required to restructure their metrics engine as a separate service, so that any 3rd party metrics vendor can plug into an advertiser's page, and see the exact same data Facebook can see.
Facebook abusing their customers by forcing an unwanted algorithmic feed upon them, which has been shown to radicalise people - then you force Facebook's back end database to openly document its entire structure, so that 3rd parties can make drop in replacements for Facebook's website & apps, that allow for people to have the feed they want (everyone they follow, in reverse chronological order).
Monopolies are broken not by taking away the spoils of monopoly (fines and taxes), but by taking away the synergies of lock-in that make the whole greater than the sum of its parts, for whomever holds all the parts.
It is said that Sheryl has been a backer of VP Harris for a long time and in some media it is said they are close.
Thanks, that's a useful picture, seeing both sides. One thing to watch with Australia federal govt is that the regulators and monitoring agencies, ACCC, Productivity Commission, ANAO and to a lesser extent ASIC, do great investigation, as far as their budget allows. Many of their report recommendations have languished for years due to lobbying, as is the case.
The part that this piece is very coy about, and the legislative motives are very coy about, is what it is that Facebook does with the news that they allegedly need permission for. Everyone always just talks about Facebook 'using' news, or, as you put it, 'distributing' news. But what does that mean? According to the bill, it means: linking to news. We should be clear about that.
Now, proponents of the bill argue that Facebook is abusing its market power. Fair enough: If Facebook had to get a license for linking to news articles, then you could perhaps make a case that they keep prices down by (ab)using their market power. In reality, however, they do not have to get permission for linking to a news article at all. So how can they abuse their market power? There is no market! The price is zero because there's nothing to trade, not because Facebook has a dominant position.
The EU basically acknowledged as much. Their approach is to protect small snippets (which might, according to some, include even the title of a news piece) under a novel neighbouring right in the copyright acts of EU member states. By doing that, they create a market for those snippets. The Australian approach entirely skips this step. Everyone is talking about an abuse of market power on a market that, in reality, simply does not exist. If A lacks a legal means to prevent B from doing X, then B will pay exactly zero dollars for doing X, no matter if B dominates the market or not. So it is very strange to hear people talk about Facebook not paying for 'using' news when all they do is something that everyone is allowed to do without permission and at no cost.
The ads issue (explained in the Srinivasan paper) is an entirely separate animal, which we should vigorously tackle using the tools of anti-competition laws. It has nothing to with linking to news articles, though. Publishers get their revenue from ads on their own site. If Google et al rig the ads market and rip off publishers, you have a different issue -- it is wholly unrelated to the link itself. Accordingly, it affects news publishers (and every other content provider!) equally no matter where the traffic is coming from.
As I describe in my talk on Facebook at 34:40 (https://youtu.be/iYJQSfQGDEE?t=2080), the internal definition of an ad impression is:
"Facebook uses a specific definition for an external impression: the number of times a unit was on screen. This is technically defined as 1x1 pixel of a post or piece of content in a user's viewport, for >0 than seconds. (We don't reveal this definition externally.)"
You can't come close to an admission that they were defrauding their paying customers all along than this. Facebook finally disclosed this inane definition in the fall of 2019. It appears that my complaining internally may have had some effect.
I like this blog but trying to prop up a post based on a good faith reading of this Australian law has me questioning your opinions generally
I think the entire premise of the Australian law is misguided, as it misjudges the nature of Google and Facebook's duopoly. I argue that it is specifically targeted advertising – and the invasive tracking that it demands – that form the true monopolistic secret sauce of both Google and Facebook.
First, I think the new law should be very concerning to anti-monopoly advocates. It crosses into the murky territory of trying to legally define a specific class of business and to bolster it at the expense of another class of business. This should be considered a strong legislative anti-pattern, a band-aid that risks choosing monopolistic winners versus addressing structural root causes. Furthermore, I agree with analyses that laws in this pattern could be extremely damaging to both the internet and society at large.
I've read recent thinkpieces on this topic that present the argument that Google and Facebook's monopoly "stole" something from traditional news outlets. This feels like magical thinking, and ignores the strong argument that the internet's alteration of information distribution is a deep root cause of news media's struggles. Print-era news media has permanently lost its /scarcity-based monopoly/ on both content and advertising, and must now compete with all-comers across the globe. This isn't an unalloyed good, as we're discovering the lost value in the focus those resources provided. However, I also think there's no sensible or desirable way to restore that world gone past - to put Humpty-Dumpty back together again.
I do, however, think there's hope to strike a new balance. Targeted advertising and invasive tracking are the technological moats upon which the Google / Facebook duopoly rests, not the readily linkable nature of the internet. It's worth recalling that prior to this duopoloy, there were a quite a number of independent ad networks, with different levels of use and explicit non-use of tracking tech. The big, successful tracking firms largely consolidated, perhaps culminating in doubleclick.net's purchase by Google. The non-tracking networks largely died out as tracking-based advertising sucked the air out of the room.
What if instead legislation attempted to eliminate targeted advertising? This would destroy the great technological moats of both ad giants in a way that doesn't need to call out specific business categories. Organic advertising by popular web outlets, news or otherwise, begins to look viable again. I'll argue that this change would help foster an increased number of independent ad networks to connect ad buyers with web properties. It's also quite possible that this change may not be complete, but I'll wager the worst case for eliminating tracked advertising is "necessary but not sufficient."
I'm with you. Just eliminate targeted advertising. Do it for privacy reasons. Or turn it on it's head and make it strictly opt-in instead of opt-out. Another issue is that no typical person really understands how all this works. No one realizes how valuable the data is that they are giving away for free (well, you can get an idea from google and facebook market cap). Everyone is scared of big brother government. Google and Facebook know EVERYTHING about me. More then I do myself. Third I have worked with Google Ads for years. It has become harder to use and understand over the years. They keep pushing you towards their automated solutions which are a complete black box. And guess how you measure results. With another google product Analytics. Every year I make the joke resulting in nervous laughter is that Google can literally tell us almost any numbers in adwords and analytics if they wanted to. We just have to trust on good faith that they don't. Also who clicks on these digital ads? I never have nor anyone I know. Anecdotal I know. But still. We have gone full circle to where this is no different then general "branding" ads on TV or in a newspaper. I suspect the "targeting" is way less effective then they will ever let on. But no one reads newspapers or watches commercials anymore so this is what we have. I don't know if Facebook and Google turn the dial in their favor. But it would be unbelievably easy for them to do. I suspect a lot of these big tech issues (and goverment conspircy issues) could be solved with real forced radical transparency. But how you accomplish this I have no idea.
"But no one reads newspapers or watches commercials anymore so this is what we have."
If it didn't interrupt something else I was watching I would genuinely spend 5 minutes every so often just watching commercials for the entertainment and to be informed about current products.
Years ago when Hulu was somewhat worth watching, they allowed you to watch short commercials periodically throughout a show or one long one at the beginning. I liked that and always elected for the long one, to just get it out of the way. Of course, I'm sure most people would do that then go to the bathroom or something, so not sure how effective it was, but it was a refreshing change.
What I find amazing is there's all this resistance by most people, myself included, toward commercials, and advertisers spend all this money on them, and they interrupt shows with them in order to force people to watch them, but then people go out of their way to watch Super Bowl commercials online or even in the form of an actual show that just shows various commercials. What that tells me is that the problem, or rather solution, of reaching people isn't in making them watch commercials during their shows, but by actually making good commercials that are entertaining and make people *want* to watch them.
I don't know enough about how advertising works, especially in this context, to have much of an opinion, but I'll offer a couple counter-points just for consideration. First, while this is certainly not the best way to deal with the problem, as it's more a treatment of a symptom than the disease, and, as you pointed out, it would be much better to deal with the root causes, it's certainly better than nothing, and Australia is limited in what they can do, so they're just doing what they can. It may not be ideal, and it may even be troubling to some, but they don't have many options, and it's better than doing nothing. The fact is, something should have been done years ago, and if everybody just keeps waiting for something to be done, nothing will ever be done. Someone has to take action. And perhaps this won't even need to be a permanent thing, but will serve as a kickstart to motivate others to take action. After all, as Matt pointed out, it caused Facebook to reveal just how much power they have and turned public opinion even more against them. While I'm not one for generally believing the ends justify the means, in this case I'm inclined to think they do.
The second point is that media has always, with few exceptions, been largely ad-funded. Newspapers made money primarily through charging for print ads, though they made some through subscriptions as well. Once the Internet took off and media started to transition there, it was the same, with most of their revenue being made through ads (just often much more annoying ones) and a small part through subscriptions. If their content is being shared on a third-party site, and that site is collecting ad revenue based on that content instead of, or even in addition to, the media outlets being able to do so, then clearly that's an issue that needs to be addressed. But then, I don't really care about the media outlets, and I think they've brought a lot of the issues they're currently facing on themselves, so for me it's more about reigning in Facebook and the others, at the very least.
Also, the ACCC is going to do a bunch of other stuff, like taking on targeted ads. They are the leading authority in that area. https://www.accc.gov.au/focus-areas/inquiries-ongoing/digital-advertising-services-inquiry/interim-report
A possible alternative for the music streaming folks - https://torrentfreak.com/university-runs-massive-bittorrent-seedbox-to-showcase-music-streaming-app-210220/
Hi Matt,
Thanks again for the solid reporting in this niche.
In a previous post, you ask what possible benefits blockchain tech may have in our industries and markets. While I can't think of anything directly related to my field (higher education), the recent finance committee hearings regarding the Gamestop phenomenon, Robinhood, and limiting of retail equity orders brought up the first and strongest use case for the tech. It is an extremely reliable accounting system that has the potential to clear orders at a much faster rate and with less error than current wall street clearing houses. The benefits could (in my admittedly uneducated) lead to huge improvements in the quality of information accessible to individual retail investors. Questions like, what percent of an equity is short at any given moment or who has the capital to uphold their market orders right now should be easy to derive with this tech. Food for thought.
:)
'Facecrook' or not, you have got to respect these guys. Australia has surrendered to it in 72 hours. Let that be a lesson to other governments that think of doing the same. You have no chance.
That Facebook ads crime is only the tip of the iceberg. Online ads is probably the biggest scam since the advent of the Internet. It’s an opaque business where the few Goliaths is the house controlling all the cards, where the customers have no choice but believe whatever they claim, on fear of missing out. And the customers have no way to verify any of the claims since the technology is proprietary, complex and opaque.
Search the web for the excellent article called “The new dot com bubble is here: it’s called online advertising”.
Anecdotally, when I surf the web, I almost never use bookmarks or type a site’s address; I google instead. So when I want to visit an online store, for example, I google by store name. The first “result” is almost always the ad for the store. If I click on it, does that mean the ad was effective? Of course not. The second result is the actual top search result and the URL is exactly the same as the one for the ad. I intentionally avoid clicking the ad but I’d imagine most users would click the ad and contribute in “falsifying” the effectiveness of the ad.
Hi Matt,
Great article. Just a small correction to the NVIDIA segment. NVIDIA doesn't only use Taiwan Semiconductors to manufacture their chips. In fact, with the current generation of consumer Ampere cards (30 series), they are manufactured on Samsung's fabs. Their prosumer and commercial cards are made on TSMC's fabs however, as are all of AMD's and Apples chips. Doesn't mean there isn't a shortage, still is with nearly every fab in the world working overtime to keep up with demand, and still failing to do so.