This is the really problematic assumption the EU made:
*'Comparison shopping services' are fundamentally different from general purpose search engines.
If we take the Comission's POV, yes Google entered a new market and used their dominant position in an existing market to unfairly compete, breaking anti trust law (though wether that's for the benefit of the consumer or not is a fair question as well).
But I don't think they are different markets. If when you search for a web page google shows you the best result, why shouldn't it search for the best/cheapest online retailer? It's searching for the most relevant 'thing' on both cases.
> But I don't think they are different markets. If when you search for a web page google shows you the best result, why shouldn't it search for the best/cheapest online retailer? It's searching for the most relevant 'thing' on both cases.
But in this case this is not what happened.
What Google did was showing their product before every other shopping comparison product despite the rank
That makes no sense. That would be like forcing Google to show Bing results inside Google searches. Product comparison is a search category—just like image based search. It's not a different vertical.
From the EU Commission's perspective, they're different categories and one of the reasons for the case in the first case was due to complaints from EU-based comparison shopping companies who claimed Google was unfairly suppressing their results in general Google search. From the perspective of a European comparison-shopping-only company, it is unfair competition from a company that spans a broader domain, hence the narrower definition used by the commission. For an antitrust case such as this one, it is a key point that Google has such high market share in web search, above 70% in EU searches. For such a dominant network, it's easy to leverage that dominance in other domains, but that's against the EU's antitrust laws.
"due to complaints from EU-based comparison shopping companies who claimed Google was unfairly suppressing their results in general Google search"
Replace "comparison shopping" with "image search" or "text result answer".
Heck, replace it with "marketing spam company".
It's nonsense. This is why the US definition involves "does this harm consumers", not "does this harm businesses".
Businesses are harmed anytime anyone competes.
This is one of the reasons that the west tends to view the European approach to antitrust as "making it up as they go along" and "targeted at non-european competitors". Because they pretty much never go after european companies, and the rulings seem completely arbitrary depending on the whims of the current commissioners.
Completely arbitrary is probably an exaggeration. I wrote a paper on this case for a Digital Cultural Policy class over a year ago so I'm a bit fuzzy on the details, but the previous commissioner almost settled the case and thought he had reached an agreement with Google that would pass in the Commission, but it failed to even reach a vote it was so poorly received.
If European businesses catch the ear of the Commission, it's usually to the tune of, "we're being cut off and consumers are thereby being cheated of the best deal." Sure, I omitted that from the previous comment, but the companies complaining know how to make their case, so it's never entirely protectionism in favor of European companies. I don't know if it would be possible to dispel the notion that it is always somewhat protectionism, so I won't go there.
The logic of the case made sense to me, and it follows a protocol that spans several commissioners. There are differences of approach depending on who's sitting in the chair, of course, but that's true of any government.
To return to your "Replace 'comparison shopping' with..." That's not a fair substitution. I see why you make it, I mean, they're tabs in the Google interface so the seem easily interchangeable. But underneath, there isn't a composite result returned on those searches with results from different vertical search engines like there is in shopping results (save "video" where YouTube and others show up), and there isn't the same opportunity for monetization from the search types you mentioned, so it's a monopoly powered search advantage that damages no one as presented on Google's results.
The US authorities are a bad standard in my view. They seem to be defanged puppies of the corporate overlords that bribe them into submission in many cases.
"The west!" How did I almost miss this gem?!? You make it sound as though the Iron Curtain didn't fall under Reagan's might but instead engulfed The Continent! Reminds me of Brick Tamland referring to Flyover Country as the Middle East :)
> But I don't think they are different markets. If when you search for a web page google shows you the best result, why shouldn't it search for the best/cheapest online retailer? It's searching for the most relevant 'thing' on both cases.
It doesn't though, does it? It certainly doesn't find the cheapest item on amazon when I tried a few examples.
Shopping doesn't crawl the web in any way. Retailers submit their stock data feed to it and can then pay to have specific products show up in the embedded box in search terms.
Which is where the anti-trust issue obviously comes in. All retailers can submit their stock feeds to all comparison engines - but only Google Shopping gives you the opportunity to flow through to a pinned spot at the top of the world's largest search engine too.
But I don't think they are different markets. If when you search for a web page google shows you the best result, why shouldn't it search for the best/cheapest online retailer? It's searching for the most relevant 'thing' on both cases.
I don't think that would be a problem, but it's not what happened. Google has a distinct 'shopping' product, and retailers have to pay to place their product in there. Almost by definition, this means that ranking is broken, and Google is essentially running its own comparison site where retailers pay for listing.
I'm inclined to agree, at least a bit. They might be reasonably different markets now, but they should not be in the future.
I think all the following searches:
- web search for "location X"
- map search for "location X"
- current weather in location X
- businesses of type Y near location X
- cheapest offers for item Z near location X
are all examples of something conceptually the same. I want to be able to make those queries over one interface, and with minimum (i.e. zero) steps between query results and actual information I'm looking for.
As it is, when I search for "cheapest item Z near X", getting links to price comparison services makes about as much sense as if I googled for something and the results were links to search results of Bing, Yahoo and DDG.
--
I appreciate how today those markets may seem different, and I also don't want to see one company dominating most on-line activity, but I also want more integration and interoperability. There needs to be a balance here.
> I want to be able to make those queries over one interface, and with minimum (i.e. zero) steps between query results and actual information I'm looking for.
Especially with computers, why would that require that they are a single product?
It should be technically outright trivial to have specialized search providers that feed into a common user interface, where earch provider can be swapped out as you prefer.
Isn't that a bit like saying that you don't want ten kinds of wall sockets in your house, therefore, your electricity should be supplied by the company that sells you all your appliances, so you only need to have one type of sockets?
This has been happening for the past two days in those Google/EU threads. Please help me.
How does "and I also don't want to see one company dominating most on-line activity, but I also want more integration and interoperability" imply that I want one company to handle all of this?
I try to explicitly state in each comment that I don't want that one-company solution. In fact, I very much want what you described - more interoperability through standardization, not through monopoly. Apparently though my comments aren't clear, given that I've been get exactly the same response for days now. Surely I must be making them hard to read somehow.
Could anyone help me discover what's wrong with that particular comment, and in general in what way my comments could be improved so that they more clearly transmit my intent?
> They might be reasonably different markets now, but they should not be in the future.
I think that's what made me think you thought it needed to be a one-company solution.
Manufacturers of vacuum cleaners and computers are not in the same market, in the sense that is relevant here, even though both plug into a wall socket. A market is not defined by a common interface, but by whether or not you compete for the same customers. A flight search service is as much a substitute for a mobile phone price comparison service as a vacuum cleaner is for a computer, therefore, those products are not in competition, therefore, they are not in the same market, they are merely technically in some ways similar products with partially overlapping interfaces.
Thanks. The definition of "the same market" you provided sounds very sensible, and so I retract my belief that those searches should be "the same market" (which is what is relevant to tge EU ruling). I still think they fit the same interface, in the "wall socket" sense.
Basically, the issue here is comparable to Net Neutrality. Google Search is a utility.
It treats most sites the same, but provides an advantage for some that cooperate with it or are owned by it, just like Zero Rating happens in the ISP situation.
And, just like in the ISP situation, users (at least in the EU) can easily switch at basically zero cost (in fact, switching ISPs actually gives users a bonus usually, many ISPs give you actual cash for switching to them).
Now, the question is why Google should be able to change the ordering in the search. I think they shouldn’t.
What should be done is that Google replaces, or expands, the existing search result for Google Shopping, for example, with the rich content they used to show at top.
Same with all other of their content.
In a next step, they should allow competitors via an API to also provide embedded content. A user can then scroll through the search results, and below the amazon result is a horizontally scrolling list of results from Amazon, the Google Maps and the Here Maps results both have miniature maps, etc.
This allows rich content, easier finding of results, it ensures a fair market, and it ensures that users can easier find out where Google’s content is sourced from (very often the Google Knowledge Graph rich text snippets quote wikipedia without attribution)
This is, basically, just applying the principles of Net Neutrality to Google.
You still don’t get it. Why does no one of the americans ever understand it? I’ve heard almost 40 people (all US-americans) make that exact argument today.
How does a website owner "switch" to Bing? They have to use Google. They have to be findable on Google. Tell me how I get access to users if I don’t have my site findable on Google, please. Tell me how I can get the same placement in Google results as Google’s own pages.
It’s not about users being able to switch. That’s entirely irrelevant for this discussion. It’s as relevant as being able to replace the website logo with a browser addon.
You apparently don't get it. The user being able to switch is material to the question of Google being monopolistic. If Google is not monopolistic, antitrust regulations shouldn't apply.
When the consumer has the free choice to switch, there is no justifiable cause to violate the liberty of the company. Though apparently the EU disagrees.
Perhaps it is not very know what the article 102 of the TFEU says:
> Any abuse by one or more undertakings of a dominant position within the internal market or in a substantial part of it shall be prohibited as incompatible with the internal market in so far as it may affect trade between Member States."
Such abuse may, in particular, consist in:
(a) directly or indirectly imposing unfair purchase or selling prices or other unfair trading conditions;
(b) limiting production, markets or technical development to the prejudice of consumers;
(c) applying dissimilar conditions to equivalent transactions with other trading parties, thereby placing them at a competitive disadvantage;
(d) making the conclusion of contracts subject to acceptance by the other parties of supplementary obligations which, by their nature or according to commercial usage, have no connection with the subject of such contracts.
And there is no mention of monopoly, what it is used is "dominant position".
> Now, the question is why Google should be able to change the ordering in the search. I think they shouldn’t.
This is such an odd argument. It assumes that search is a largely solved problem.
But search is not solved, and the web (and human knowledge) is an evolving corpus.
> A user can then scroll through the search results, and below the amazon result is a horizontally scrolling list of results from Amazon, the Google Maps and the Here Maps results both have miniature maps, etc.
And this is even weirder. Who actually wants this?
> This is such an odd argument. It assumes that search is a largely solved problem.
No, it’s pretty simple. The search ordering algorithm should get as inputs only your query, and the linked page.
Currently, Google manually special-cases a handful of pages. That should simply end.
> And this is even weirder. Who actually wants this?
As I explained in another comment, I live in a city where Here Maps is vastly superior to Google Maps (Google Maps has maps material from ~2009, and, until 2 weeks ago, their satellite imagery was from 2004, so they frequently route you to non-existing roads, or so on, and they have no transit integration and bad traffic data, while Here Maps has data for roads and satellite from this week, and has transit integration), yet everyone uses Google Maps, simply because it is already preinstalled and shows up first in Google Search, despite worse results.
So, for at least a city of 280k people, Here Maps should probably show first. Or at least be equally integrated as Google Maps is.
Because currently the worse product is winning just because it is bundled.
> Take travel sites: why shouldn’t Priceline sue Google for featuring ads for hotel booking sites above its own results?
If they're treating people equally, then that's not a problem. Priceline can buy adverts like anyone else.
> Why should Google be able to make any money at all?
Google should make more money by being better at hotel bookings and better at product aggregation and better at each of their other offerings. They shouldn't make money purely because people already use google and google put their (possibly sub-par) offerings right at the top. This would make it much harder for anyone else in the field to come in and offer a better thing, because a high proportion of people will never go down even lower on the first page of results.
A very good example of this is Google Maps vs. Here Maps.
I live in a city where Google Maps has, until one month ago, had basically no traffic data, satellite imagery was from 2004, and even today still supports no transit.
Here maps has up-to-date imagery of highest quality, has traffic and transit data.
Yet, many people use Google Maps, even if that provides worse results, sends them on roads that don’t exist anymore, etc.
When one recommends they switch, they say "but Google Maps is preinstalled, installing Here Maps would be so much effort, and it doesn’t work with Ok Google".
Google Product Listing Ads are the advertisements for specific products that show up in a carousel at the top of search results that include product pictures, titles, and prices.
Product Listing Ads are like AdWords (text ads) in that they are advertisements placed by a particular merchant for a price. They are graphical ads in that they include a picture, but the picture must show a specific product that's available for sale.
Any merchant can sign up and place Product Listing Ads at the top of search results, just like how merchants can sign up to buy AdWords.
Can you clarify what you think is problematic about Product Listing Ads? Can you clarify your stance on AdWords? What do you think Google should do differently?
Product Listing Ads are already open to all merchants, just like AdWords. They show up at the top because they are advertisements, and showing up at the top is precisely what the merchant is buying. Why should product ads be treated any differently than AdWords? I don't see why one should be allowed and the other prohibited.
Or are you just saying that you don't think Google should be allowed to offer any advertising whatsoever alongside their free search?
I'm asking these questions because the criticism of Google's actions in this space does not seem sensible or consistent to me. If you agree that Google should be allowed to show advertisements in its free search, then I don't see what objection there can be to Product Listing Ads, or why someone would object to PLAs and not AdWords, given that both ad platforms are open to everyone. If they were specifically prohibiting their competitors from placing ads, or raking their ads poorly, that would be one thing, but what they seem to be accused of is, as best as I can understand it, offering ads at all. The logic is tortured.
(This current behavior seems to be what the EU investigation objects to; I didn't see a mention in the press release about Google having already stopped the practices they objected to, so I assume they object to the current search behavior. It may have worked differently in the past, like pre 2012.)
>Google should make more money by being better at hotel bookings and better at product aggregation and better at each of their other offerings. They shouldn't make money purely because people already use google and google put their (possibly sub-par) offerings right at the top. This would make it much harder for anyone else in the field to come in and offer a better thing, because a high proportion of people will never go down even lower on the first page of results.
So, a Unix Philosophy view of anti-trust law? Do one thing and do it well?
I object to using the term "virtuous cycle" to describe positive feedback. Actually I object to the term "positive feedback" when talking to non-engineers because it can be misconstrued as imposing a judgement on it. But calling it virtuous? I suppose from the business point of view it's a really good thing. "reinforcing cycle" IMHO is more accurate.
I hate the terms "virtuous cycle" and "vicious cycle" with passion. Just name the damn things correctly (positive/negative feedback loop), or at least refer to their behaviour (reinforcing / dampening).
And in general, I think basics of feedback loops should be taught in primary school. I am scared by the fact so many regular people don't comprehend these basic concepts (which is often evident in discussions, when you see people missing obvious consequences of their ideas because they don't get the concept of feedback).
"Virtuous" and "vicious" cycles both refer to positive feedback, and differ only in value judgement they imply, so arguably the value judgement is the entire point of using these terms. I can't think of a common use term for negative feedback off the top of my head.
The author fails to include and rebut a crucial point in the EU commision's original article.
He equates the monopolistic wrongdoing with showing adds. The EU instead says that it was Google showing adds for its own service.
- First, Google dominated search and not shopping.
- Then, Google used search to unfairly compete with shopping by creating its own service.
The author leaves this point out, and I think it was central to the EU's decision.
If some other company had created a comparison shopping service, and then paid Google to show its service inside Google search results, this would be more similar to the circumstances of the article's argument.
There's something there, indeed. But I feel it's really hard to define it exhaustively.
The product carousel shows only paid ads, so it doesn't really matter that it's part of Google shopping. Would it be ok if the actual "Google shopping" didn't exist, and it were officially just a carousel of ads?
I'm also quite certain that they internally charge a business unit such as Google Shopping for the placement. Would it be ok if they allowed other comparison shopping sites to bid for the carousel spot?
It's complicated by the fact that for most products, Google does actually provide a much better experience than competitors. Take flight search, where the Google interface is just so much better than any competitor. I get that it can be in the consumers' long-term interest to preserve competition even if it means hobbling the best product, but the magnitudes involved here currently make it a really difficult decision.
> I'm also quite certain that they internally charge a business unit such as Google Shopping for the placement. Would it be ok if they allowed other comparison shopping sites to bid for the carousel spot?
Obviously not. What would happen in that case? The Google Shopping unit would go to the CFO (or whoever is responsible) and tell them that they needed more funding for bidding on carousel spots, so as to increase their market share. They would make the very solid business case that any money spent there would end up back in the company's accounts, so it is actually free. So, they would get whatever money they would need in order to outbid everyone else.
As long as your expenses flow back to the entity that's funding you, there won't be any limit on funding.
The theory that you're describing holds only if ads placed through Google Shopping outperform whatever other kind of ads might appear there. If the Google Shopping ads (not actually the right term) actually perform worse, then Google will lose money, and the funding will not be unlimited.
The product results that appear in Google Search are called Product Listing Ads (PLAs), and are not a comparison shopping feature today. (I don't know how things worked in the past.) They are simply advertisements.
Individual merchants publish their product catalog into Google's Merchant Center, along with the price they're willing to pay for each product ad. It's similar to adwords, just with product information too (title, picture, price).
Consequently, it's not a comparison shopping system - within Google Search it is purely an advertising feature with more information than text ads. Furthermore, as far as I'm aware, any company can participate in it by listing their products. I also think but am not positive that the only products featured in the carousel will be paid advertisements - another way in which it is not a comparison shopping system.
Note: I'm only describing Product Listing Ads within Google Search results, not whatever else Google Shopping might be. It does appear that the main Google Shopping website also involves payment from merchants, which they disclose on the page.
I don't know what determines whether the product carousel is displayed or not. I would assume that it's shown at the top when Google has enough information to display it, and when it's outperforming other types of ads. Anecdotally, Product Listing Ads have considerably higher conversion than traditional AdWords, so I would expect that PLAs will naturally perform better and would justify the top advertising slot.
He did address this by simply stating the fact there is no actual "Google Shopping", it's just a term, like "Google Hotels" or "Google Whatever".
In the end "Google Shopping" is just an Ad Block with a nomenclature of "Google Shopping. They could simply remove this term and put the same ads for Adidas shoes at the top of search results just like do for a myriad of other products and search categories; would that be a problem?
I understand that Google probably has a different engineering platform to present shopping ads/results as opposed to webpages, but is this really any different from other kinds of scraping they perform? They also show song lyrics and snippets from Wikipedia for example. They're scraping the web and collating that info for users to search for.
One important point for everyone to keep in mind, whenever you read an article about some years-long competition investigation, is that it is talking about conduct that happened in the past. In this case, by internet standards, we're talking about conduct that happened a long, long time ago (but, in this galaxy).
This is a refreshingly well-reasoned article, in that it shows the actual trade-offs involved, considers both sides to be legitimate, and tries to work with each sides' strongest argument.
Compare and contrast with yesterday's discussions here, which were littered with "These EU bureaucrats are STEALING!1! from US companies" and (fewer) "Google is greatest danger since the dinosaur's extinction and needs to be broken up"
"The United States and European Union have, at least since the Reagan Administration, differed on this point: the U.S. is primarily concerned with consumer welfare, and the primary proxy is price. In other words, as long as prices do not increase — or even better, decrease — there is, by definition, no illegal behavior."
Giving it away for free was precisely the definition of illegal behavior last time:
Microsoft not being broken up wasn't about them not being guilty. It was about prosecution (under the new GWB administration) seeking a lesser penalty.
Actually, it was about courts deciding the liability scope was too wide and could harm innovation, as per your linked source.
"Although the D.C. Circuit found that it was possible to examine high-tech industries with traditional antitrust analysis, the court announced a new and permissive liability rule that repudiated the Supreme Court’s dominant rule of per se illegality for tie-ins, due to the court’s concern for the dynamic effects that a per se rule would have on innovation."
Ergo, the tie-ins were found to have happened by point of fact but not found to be illegal by re-interpretation by the courts of the illegality of tie-ins. The Justice Department then reached a settlement (the win on their side being that they wouldn't have wasted the years of prosecution).
This goes a way towards explaining why there hasn't been a similar Microsoft-scale antitrust case since the Microsoft case in the US.
Surprised nobody has interpreted search engine spidering bandwidth as a form of public good. There's a limit to the amount of spidering that can happen. It's okay if 10 or a 100 companies do it, but if everyone had to do it, the web's access pattern would be exponential and overwhelming.
There's a class of consumer who is technically willing and able to operate their own spider but accepts the lesser convenience of a web search because they recognize it would take down the internet if everyone spidered all the time.
Also surprised the article doesn't mention omnibar search. By defaulting this to G, G is 'abusing' browser dominance to reinforce search dominance.
>"the web's access pattern would be exponential and overwhelming."
It might well be overwhelming, but it wouldn't be exponential. There are a maximum of N choose 2 connections between N nodes. That number only grows quadratically as N increases, while exponential growth would outstrip quadratic, cubic or any polynomial growth for large values of N.
> The United States and European Union have, at least since the Reagan Administration, differed on this point: the U.S. is primarily concerned with consumer welfare [...] The European Commission, on the other hand, is explicitly focused on competition
What is missing here is that the US has diverged from a previously common interpretation. Between roughly 1890 (Sherman Act) and 1982 (Reagan administration antitrust guidelines) the US authorities were mainly concerned with competition too (breakup of Standard Oil: 1911; breakup of Bell System: 1984). As late as 2001 Microsoft had a serious brush with antitrust law for anticompetitive practices [1]. There's a case to be made that the US is currently just going through a lapse in enforcement.
"Between roughly 1890 (Sherman Act) and 1982 (Reagan administration antitrust guidelines) the US authorities were mainly concerned with competition too (breakup of Standard Oil: 1911; breakup of Bell System: 1984)."
This is the wrong timeline.
The spectrum shifted towards consumer welfare very early on (right after standard oil), because they found the legal definitions in the sherman act too broad to be useful (every contract is a restraint of trade).
If you read the bell cases, etc that you cite, they were already focused on consumer welfare.
This is because otherwise the bounds of antitrust are limitless.
Every competition between businesses harms a business :)
> The spectrum shifted towards consumer welfare very early on
You seem unaware that Bork's Antitrust Paradox only came out in 1978 [1]. Yes, it's not surprising that its views would appear in cases from the '80s. Here's prof. Orbach in the Washington Post on Bork's influence [2]:
> Antitrust was defined by Robert Bork. I cannot overstate his influence. [...] Antitrust [in 1960] was about protecting small businesses. [Bork] built a full framework about how antitrust should be more about economic efficiency than about helping small businesses. [...] He wrote a sentence: Congress enacted the Sherman act [ed - the main antitrust act] as a "consumer welfare prescription.” The Supreme Court adopted that sentence in 1979. That is the stated goal in antitrust today. It is a big deal. A huge deal. In antitrust, it’s operational. Robert Bork defined it.
> In antitrust, antimonopoly law, it is easy to think about why big is bad. The view is, "Big businesses are bad. We should protect small businesses. We should not have big businesses." So that’s what happened in the history of the Sherman Act, between 1880 and 1960.
"ou seem unaware that Bork's Antitrust Paradox only came out in 1978 [1]"
I'm quite aware.
" Antitrust [in 1960] was about protecting small businesses."
Citation?
I gave you a case from the late 40's that was about protecting consumer welfare.
You give me a news article about a law professor who loved robert bork's work?
Look, Bork was influential for sure.
I can and have cited you cases where, over a short period of time, the supreme court decided that the sherman act was unworkable (see all the various changes around resale price maintenance, over the years as well), and started pushing towards other theories for antitrust.
Contrary to these claims, it's not "Bork wrote a book, things changed overnight".
That's just a history rewrite.
> The incipiency theory starts from the idea that it is possible to nip restraints of trade and monopolies in the bud before they blossom to Sherman Act proportions. It underlies the Clayton Act, the Robinson-Patman Act, and the Federal Trade Commission Act. [...] The courts have used the incipiency notion as a license for almost unlimited extrapolation, reasoning from any trend toward concentration in an industry that there is an incipient lessening of competition. [...] In case after case the FTC, for example, nails down its finding that competition is injured with the testimony of competitors of the respondent that his activities and aggressiveness may or have cost them sales. [...] When the head of the Antitrust Division or the FTC reports to a congressional committee, protocol requires that he wear a suitable number of bloody scalps of businessmen at his belt.
Err, Bork was advocating for his view, so i'm not sure he's a good source?
Look, your argument amounts to:
For 60+ years, between 1920 and 1982, there was literally no legal theory development in antitrust. No other cases, no thoughts by anyone, no articles. Then suddenly, Bork came on the scene and redefined everything overnight!
Certainly that would have made my antitrust law classes a lot shorter, but i don't think it's reality :)
Let me try a different source for you
Here's what one of the authors of the act (Senator Hoar) had to say about it:
""... [a person] who merely by superior skill and intelligence...got the whole business because nobody could do it as well as he could was not a monopolist..(but was if) it involved something like the use of means which made it impossible for other persons to engage in fair competition.""
Oh look, a consumer welfare argument.
Also, here's another case from the supreme court in 1940 (Apex Hosiery Co. v. Leader 310 U. S. 469)
"The goal [of the sherman act] was to prevent restraints of free competition in business and commercial transactions which tended to restrict production, raise prices, or otherwise control the market to the detriment of purchasers or consumers of goods and services, all of which had come to be regarded as a special form of public injury."
(emph mine). A very clear consumer welfare argument. They also cite supreme court cases.
The investigation was concerned with Google's introduction and promotion of Google Shopping in its search results, not with whether it's a problem for Google to show product listing ads.
So would this mean that showing image results in Google's main search page is also against EU rules, since "Google Images" (images.google.com) is also a "search product" that Google is promoting over its competitors like Bing (which also offers image search capabilities)?
Why not. Google Search is financing and providing users to Google Map. How could you be competitive in the map market when you have a giant with limitless ressource ? See street view, how could you finance that ? WHat is the commercial viability of such a product ?
Why should we consider a separate "map market"? For consumers, there might just be an "information market. Sometimes an image would be the right form of information; sometimes a map; sometimes a video etc.
Trying to cleave out different verticals from a search engine would be a major disadvantage to consumers.
I’ve written this in another comment before, but it explains why we should separate them:
> I live in a city where Google Maps has, until one month ago, had basically no traffic data, satellite imagery was from 2004, and even today still supports no transit.
> Here maps has up-to-date imagery of highest quality, has traffic and transit data.
> Yet, many people use Google Maps, even if that provides worse results, sends them on roads that don’t exist anymore, etc.
> When one recommends they switch, they say "but Google Maps is preinstalled, installing Here Maps would be so much effort, and it doesn’t work with Ok Google".
Do you think this is an advantage to consumers, that they end up with the worse product?
I disagree with some of the arguments but reach a similar conclusion:
Google is a monopoly, is a bad actor and this ruling is not a useful solution.
I think the government needs to break up the cable & telco rent companies. Conversely, I blame the market for empowering Google to this level and the market needs to fix it or accept the tradeoffs. I think the price is too high, but I'm in the minority of DuckDuckGo users who limit chrome / gservice usage. Even I touch some part of it daily as it is nearly impossible not to. I'm not as dogmatic as RMS so using some useful parts sometimes is fine, but the issue is almost all core services are being leveraged by almost everyone all the time and that is dangerous if simply begause it's a single failure point
I wonder why the EC would not want a good enough solution where Google Shopping results will not take more than one row and comparison shopping services would appear on at least page two (or maybe even page one)?
68 comments
[ 0.22 ms ] story [ 132 ms ] thread*'Comparison shopping services' are fundamentally different from general purpose search engines.
If we take the Comission's POV, yes Google entered a new market and used their dominant position in an existing market to unfairly compete, breaking anti trust law (though wether that's for the benefit of the consumer or not is a fair question as well).
But I don't think they are different markets. If when you search for a web page google shows you the best result, why shouldn't it search for the best/cheapest online retailer? It's searching for the most relevant 'thing' on both cases.
But in this case this is not what happened.
What Google did was showing their product before every other shopping comparison product despite the rank
Replace "comparison shopping" with "image search" or "text result answer". Heck, replace it with "marketing spam company".
It's nonsense. This is why the US definition involves "does this harm consumers", not "does this harm businesses".
Businesses are harmed anytime anyone competes.
This is one of the reasons that the west tends to view the European approach to antitrust as "making it up as they go along" and "targeted at non-european competitors". Because they pretty much never go after european companies, and the rulings seem completely arbitrary depending on the whims of the current commissioners.
If European businesses catch the ear of the Commission, it's usually to the tune of, "we're being cut off and consumers are thereby being cheated of the best deal." Sure, I omitted that from the previous comment, but the companies complaining know how to make their case, so it's never entirely protectionism in favor of European companies. I don't know if it would be possible to dispel the notion that it is always somewhat protectionism, so I won't go there.
The logic of the case made sense to me, and it follows a protocol that spans several commissioners. There are differences of approach depending on who's sitting in the chair, of course, but that's true of any government.
To return to your "Replace 'comparison shopping' with..." That's not a fair substitution. I see why you make it, I mean, they're tabs in the Google interface so the seem easily interchangeable. But underneath, there isn't a composite result returned on those searches with results from different vertical search engines like there is in shopping results (save "video" where YouTube and others show up), and there isn't the same opportunity for monetization from the search types you mentioned, so it's a monopoly powered search advantage that damages no one as presented on Google's results.
The US authorities are a bad standard in my view. They seem to be defanged puppies of the corporate overlords that bribe them into submission in many cases.
"The west!" How did I almost miss this gem?!? You make it sound as though the Iron Curtain didn't fall under Reagan's might but instead engulfed The Continent! Reminds me of Brick Tamland referring to Flyover Country as the Middle East :)
Do you have any stat to back this claim?
> This is one of the reasons that the west
Define the "west"
It doesn't though, does it? It certainly doesn't find the cheapest item on amazon when I tried a few examples.
Which is where the anti-trust issue obviously comes in. All retailers can submit their stock feeds to all comparison engines - but only Google Shopping gives you the opportunity to flow through to a pinned spot at the top of the world's largest search engine too.
I don't think that would be a problem, but it's not what happened. Google has a distinct 'shopping' product, and retailers have to pay to place their product in there. Almost by definition, this means that ranking is broken, and Google is essentially running its own comparison site where retailers pay for listing.
I think all the following searches:
are all examples of something conceptually the same. I want to be able to make those queries over one interface, and with minimum (i.e. zero) steps between query results and actual information I'm looking for.As it is, when I search for "cheapest item Z near X", getting links to price comparison services makes about as much sense as if I googled for something and the results were links to search results of Bing, Yahoo and DDG.
--
I appreciate how today those markets may seem different, and I also don't want to see one company dominating most on-line activity, but I also want more integration and interoperability. There needs to be a balance here.
Especially with computers, why would that require that they are a single product?
It should be technically outright trivial to have specialized search providers that feed into a common user interface, where earch provider can be swapped out as you prefer.
Isn't that a bit like saying that you don't want ten kinds of wall sockets in your house, therefore, your electricity should be supplied by the company that sells you all your appliances, so you only need to have one type of sockets?
How does "and I also don't want to see one company dominating most on-line activity, but I also want more integration and interoperability" imply that I want one company to handle all of this?
I try to explicitly state in each comment that I don't want that one-company solution. In fact, I very much want what you described - more interoperability through standardization, not through monopoly. Apparently though my comments aren't clear, given that I've been get exactly the same response for days now. Surely I must be making them hard to read somehow.
Could anyone help me discover what's wrong with that particular comment, and in general in what way my comments could be improved so that they more clearly transmit my intent?
> They might be reasonably different markets now, but they should not be in the future.
I think that's what made me think you thought it needed to be a one-company solution.
Manufacturers of vacuum cleaners and computers are not in the same market, in the sense that is relevant here, even though both plug into a wall socket. A market is not defined by a common interface, but by whether or not you compete for the same customers. A flight search service is as much a substitute for a mobile phone price comparison service as a vacuum cleaner is for a computer, therefore, those products are not in competition, therefore, they are not in the same market, they are merely technically in some ways similar products with partially overlapping interfaces.
https://www.google.com/shopping
It treats most sites the same, but provides an advantage for some that cooperate with it or are owned by it, just like Zero Rating happens in the ISP situation.
And, just like in the ISP situation, users (at least in the EU) can easily switch at basically zero cost (in fact, switching ISPs actually gives users a bonus usually, many ISPs give you actual cash for switching to them).
Now, the question is why Google should be able to change the ordering in the search. I think they shouldn’t.
What should be done is that Google replaces, or expands, the existing search result for Google Shopping, for example, with the rich content they used to show at top.
Same with all other of their content.
In a next step, they should allow competitors via an API to also provide embedded content. A user can then scroll through the search results, and below the amazon result is a horizontally scrolling list of results from Amazon, the Google Maps and the Here Maps results both have miniature maps, etc.
This allows rich content, easier finding of results, it ensures a fair market, and it ensures that users can easier find out where Google’s content is sourced from (very often the Google Knowledge Graph rich text snippets quote wikipedia without attribution)
This is, basically, just applying the principles of Net Neutrality to Google.
I understand this way of thinking but the only trouble I have with this is that unlike ISPs I can switch to bing in a heartbeat.
How does a website owner "switch" to Bing? They have to use Google. They have to be findable on Google. Tell me how I get access to users if I don’t have my site findable on Google, please. Tell me how I can get the same placement in Google results as Google’s own pages.
It’s not about users being able to switch. That’s entirely irrelevant for this discussion. It’s as relevant as being able to replace the website logo with a browser addon.
When the consumer has the free choice to switch, there is no justifiable cause to violate the liberty of the company. Though apparently the EU disagrees.
> Any abuse by one or more undertakings of a dominant position within the internal market or in a substantial part of it shall be prohibited as incompatible with the internal market in so far as it may affect trade between Member States."
Such abuse may, in particular, consist in:
And there is no mention of monopoly, what it is used is "dominant position".This is such an odd argument. It assumes that search is a largely solved problem.
But search is not solved, and the web (and human knowledge) is an evolving corpus.
> A user can then scroll through the search results, and below the amazon result is a horizontally scrolling list of results from Amazon, the Google Maps and the Here Maps results both have miniature maps, etc.
And this is even weirder. Who actually wants this?
No, it’s pretty simple. The search ordering algorithm should get as inputs only your query, and the linked page.
Currently, Google manually special-cases a handful of pages. That should simply end.
> And this is even weirder. Who actually wants this?
As I explained in another comment, I live in a city where Here Maps is vastly superior to Google Maps (Google Maps has maps material from ~2009, and, until 2 weeks ago, their satellite imagery was from 2004, so they frequently route you to non-existing roads, or so on, and they have no transit integration and bad traffic data, while Here Maps has data for roads and satellite from this week, and has transit integration), yet everyone uses Google Maps, simply because it is already preinstalled and shows up first in Google Search, despite worse results.
So, for at least a city of 280k people, Here Maps should probably show first. Or at least be equally integrated as Google Maps is.
Because currently the worse product is winning just because it is bundled.
If they're treating people equally, then that's not a problem. Priceline can buy adverts like anyone else.
> Why should Google be able to make any money at all?
Google should make more money by being better at hotel bookings and better at product aggregation and better at each of their other offerings. They shouldn't make money purely because people already use google and google put their (possibly sub-par) offerings right at the top. This would make it much harder for anyone else in the field to come in and offer a better thing, because a high proportion of people will never go down even lower on the first page of results.
I live in a city where Google Maps has, until one month ago, had basically no traffic data, satellite imagery was from 2004, and even today still supports no transit.
Here maps has up-to-date imagery of highest quality, has traffic and transit data.
Yet, many people use Google Maps, even if that provides worse results, sends them on roads that don’t exist anymore, etc.
When one recommends they switch, they say "but Google Maps is preinstalled, installing Here Maps would be so much effort, and it doesn’t work with Ok Google".
This is literally the effect shown IRL.
Product Listing Ads are like AdWords (text ads) in that they are advertisements placed by a particular merchant for a price. They are graphical ads in that they include a picture, but the picture must show a specific product that's available for sale.
Any merchant can sign up and place Product Listing Ads at the top of search results, just like how merchants can sign up to buy AdWords.
Can you clarify what you think is problematic about Product Listing Ads? Can you clarify your stance on AdWords? What do you think Google should do differently?
Product Listing Ads are already open to all merchants, just like AdWords. They show up at the top because they are advertisements, and showing up at the top is precisely what the merchant is buying. Why should product ads be treated any differently than AdWords? I don't see why one should be allowed and the other prohibited.
Or are you just saying that you don't think Google should be allowed to offer any advertising whatsoever alongside their free search?
I'm asking these questions because the criticism of Google's actions in this space does not seem sensible or consistent to me. If you agree that Google should be allowed to show advertisements in its free search, then I don't see what objection there can be to Product Listing Ads, or why someone would object to PLAs and not AdWords, given that both ad platforms are open to everyone. If they were specifically prohibiting their competitors from placing ads, or raking their ads poorly, that would be one thing, but what they seem to be accused of is, as best as I can understand it, offering ads at all. The logic is tortured.
(This current behavior seems to be what the EU investigation objects to; I didn't see a mention in the press release about Google having already stopped the practices they objected to, so I assume they object to the current search behavior. It may have worked differently in the past, like pre 2012.)
So, a Unix Philosophy view of anti-trust law? Do one thing and do it well?
And in general, I think basics of feedback loops should be taught in primary school. I am scared by the fact so many regular people don't comprehend these basic concepts (which is often evident in discussions, when you see people missing obvious consequences of their ideas because they don't get the concept of feedback).
"self-defeating"?
Ah, so that's what they're called.
He equates the monopolistic wrongdoing with showing adds. The EU instead says that it was Google showing adds for its own service.
- First, Google dominated search and not shopping.
- Then, Google used search to unfairly compete with shopping by creating its own service.
The author leaves this point out, and I think it was central to the EU's decision.
If some other company had created a comparison shopping service, and then paid Google to show its service inside Google search results, this would be more similar to the circumstances of the article's argument.
The product carousel shows only paid ads, so it doesn't really matter that it's part of Google shopping. Would it be ok if the actual "Google shopping" didn't exist, and it were officially just a carousel of ads?
I'm also quite certain that they internally charge a business unit such as Google Shopping for the placement. Would it be ok if they allowed other comparison shopping sites to bid for the carousel spot?
It's complicated by the fact that for most products, Google does actually provide a much better experience than competitors. Take flight search, where the Google interface is just so much better than any competitor. I get that it can be in the consumers' long-term interest to preserve competition even if it means hobbling the best product, but the magnitudes involved here currently make it a really difficult decision.
Obviously not. What would happen in that case? The Google Shopping unit would go to the CFO (or whoever is responsible) and tell them that they needed more funding for bidding on carousel spots, so as to increase their market share. They would make the very solid business case that any money spent there would end up back in the company's accounts, so it is actually free. So, they would get whatever money they would need in order to outbid everyone else.
As long as your expenses flow back to the entity that's funding you, there won't be any limit on funding.
The product results that appear in Google Search are called Product Listing Ads (PLAs), and are not a comparison shopping feature today. (I don't know how things worked in the past.) They are simply advertisements.
Individual merchants publish their product catalog into Google's Merchant Center, along with the price they're willing to pay for each product ad. It's similar to adwords, just with product information too (title, picture, price).
Consequently, it's not a comparison shopping system - within Google Search it is purely an advertising feature with more information than text ads. Furthermore, as far as I'm aware, any company can participate in it by listing their products. I also think but am not positive that the only products featured in the carousel will be paid advertisements - another way in which it is not a comparison shopping system.
Note: I'm only describing Product Listing Ads within Google Search results, not whatever else Google Shopping might be. It does appear that the main Google Shopping website also involves payment from merchants, which they disclose on the page.
I don't know what determines whether the product carousel is displayed or not. I would assume that it's shown at the top when Google has enough information to display it, and when it's outperforming other types of ads. Anecdotally, Product Listing Ads have considerably higher conversion than traditional AdWords, so I would expect that PLAs will naturally perform better and would justify the top advertising slot.
You're allowed to have a monopoly, as long as you gain it fairly and defend it fairly, and don't use it to give other products an unfair advantage.
So then the question simply becomes: is Google Shopping a separate service or is it search.
The author argues that Google Shopping is search, the EU argues it's a separate service. There's merit to both arguments IMO.
In the end "Google Shopping" is just an Ad Block with a nomenclature of "Google Shopping. They could simply remove this term and put the same ads for Adidas shoes at the top of search results just like do for a myriad of other products and search categories; would that be a problem?
I understand that Google probably has a different engineering platform to present shopping ads/results as opposed to webpages, but is this really any different from other kinds of scraping they perform? They also show song lyrics and snippets from Wikipedia for example. They're scraping the web and collating that info for users to search for.
I think the author made a good case, good read.
When Google shopping was introduced it was a search product, actually, and was not "just an ad." It only became an ad in 2012: https://en.wikipedia.org/wiki/Google_Shopping#Change_to_paid...
Compare and contrast with yesterday's discussions here, which were littered with "These EU bureaucrats are STEALING!1! from US companies" and (fewer) "Google is greatest danger since the dinosaur's extinction and needs to be broken up"
Giving it away for free was precisely the definition of illegal behavior last time:
https://en.wikipedia.org/wiki/United_States_v._Microsoft_Cor....
https://en.wikipedia.org/wiki/United_States_v._Microsoft_Cor...
Microsoft not being broken up wasn't about them not being guilty. It was about prosecution (under the new GWB administration) seeking a lesser penalty.
"Although the D.C. Circuit found that it was possible to examine high-tech industries with traditional antitrust analysis, the court announced a new and permissive liability rule that repudiated the Supreme Court’s dominant rule of per se illegality for tie-ins, due to the court’s concern for the dynamic effects that a per se rule would have on innovation."
Ergo, the tie-ins were found to have happened by point of fact but not found to be illegal by re-interpretation by the courts of the illegality of tie-ins. The Justice Department then reached a settlement (the win on their side being that they wouldn't have wasted the years of prosecution).
This goes a way towards explaining why there hasn't been a similar Microsoft-scale antitrust case since the Microsoft case in the US.
There's a class of consumer who is technically willing and able to operate their own spider but accepts the lesser convenience of a web search because they recognize it would take down the internet if everyone spidered all the time.
Also surprised the article doesn't mention omnibar search. By defaulting this to G, G is 'abusing' browser dominance to reinforce search dominance.
It might well be overwhelming, but it wouldn't be exponential. There are a maximum of N choose 2 connections between N nodes. That number only grows quadratically as N increases, while exponential growth would outstrip quadratic, cubic or any polynomial growth for large values of N.
What is missing here is that the US has diverged from a previously common interpretation. Between roughly 1890 (Sherman Act) and 1982 (Reagan administration antitrust guidelines) the US authorities were mainly concerned with competition too (breakup of Standard Oil: 1911; breakup of Bell System: 1984). As late as 2001 Microsoft had a serious brush with antitrust law for anticompetitive practices [1]. There's a case to be made that the US is currently just going through a lapse in enforcement.
[1] https://en.wikipedia.org/wiki/United_States_v._Microsoft_Cor....
This is the wrong timeline. The spectrum shifted towards consumer welfare very early on (right after standard oil), because they found the legal definitions in the sherman act too broad to be useful (every contract is a restraint of trade).
If you read the bell cases, etc that you cite, they were already focused on consumer welfare.
This is because otherwise the bounds of antitrust are limitless. Every competition between businesses harms a business :)
Also note: The AT&T case was settled. The original case was filed in 1949, not 1984. You can see this in the 1982 settlement, which is a modification of the final judgment from 1956: http://web.archive.org/web/20060830041121/http://members.cox...
(a lot of folks are not aware of the 1949 case, it seems :P)
You seem unaware that Bork's Antitrust Paradox only came out in 1978 [1]. Yes, it's not surprising that its views would appear in cases from the '80s. Here's prof. Orbach in the Washington Post on Bork's influence [2]:
> Antitrust was defined by Robert Bork. I cannot overstate his influence. [...] Antitrust [in 1960] was about protecting small businesses. [Bork] built a full framework about how antitrust should be more about economic efficiency than about helping small businesses. [...] He wrote a sentence: Congress enacted the Sherman act [ed - the main antitrust act] as a "consumer welfare prescription.” The Supreme Court adopted that sentence in 1979. That is the stated goal in antitrust today. It is a big deal. A huge deal. In antitrust, it’s operational. Robert Bork defined it.
> In antitrust, antimonopoly law, it is easy to think about why big is bad. The view is, "Big businesses are bad. We should protect small businesses. We should not have big businesses." So that’s what happened in the history of the Sherman Act, between 1880 and 1960.
[1] https://en.wikipedia.org/wiki/The_Antitrust_Paradox
[2] https://www.washingtonpost.com/news/wonk/wp/2012/12/20/antit...
I'm quite aware.
" Antitrust [in 1960] was about protecting small businesses."
Citation?
I gave you a case from the late 40's that was about protecting consumer welfare.
You give me a news article about a law professor who loved robert bork's work?
Look, Bork was influential for sure.
I can and have cited you cases where, over a short period of time, the supreme court decided that the sherman act was unworkable (see all the various changes around resale price maintenance, over the years as well), and started pushing towards other theories for antitrust.
Contrary to these claims, it's not "Bork wrote a book, things changed overnight". That's just a history rewrite.
Bork was the final nail in the coffin.
How about Bork himself in 1963 [1]:
> The incipiency theory starts from the idea that it is possible to nip restraints of trade and monopolies in the bud before they blossom to Sherman Act proportions. It underlies the Clayton Act, the Robinson-Patman Act, and the Federal Trade Commission Act. [...] The courts have used the incipiency notion as a license for almost unlimited extrapolation, reasoning from any trend toward concentration in an industry that there is an incipient lessening of competition. [...] In case after case the FTC, for example, nails down its finding that competition is injured with the testimony of competitors of the respondent that his activities and aggressiveness may or have cost them sales. [...] When the head of the Antitrust Division or the FTC reports to a congressional committee, protocol requires that he wear a suitable number of bloody scalps of businessmen at his belt.
[1] https://www.scribd.com/document/117500897/The-Crisis-in-Anti...
Look, your argument amounts to:
For 60+ years, between 1920 and 1982, there was literally no legal theory development in antitrust. No other cases, no thoughts by anyone, no articles. Then suddenly, Bork came on the scene and redefined everything overnight!
Certainly that would have made my antitrust law classes a lot shorter, but i don't think it's reality :)
Let me try a different source for you
Here's what one of the authors of the act (Senator Hoar) had to say about it: ""... [a person] who merely by superior skill and intelligence...got the whole business because nobody could do it as well as he could was not a monopolist..(but was if) it involved something like the use of means which made it impossible for other persons to engage in fair competition.""
Oh look, a consumer welfare argument.
Also, here's another case from the supreme court in 1940 (Apex Hosiery Co. v. Leader 310 U. S. 469) "The goal [of the sherman act] was to prevent restraints of free competition in business and commercial transactions which tended to restrict production, raise prices, or otherwise control the market to the detriment of purchasers or consumers of goods and services, all of which had come to be regarded as a special form of public injury."
(emph mine). A very clear consumer welfare argument. They also cite supreme court cases.
The author's point about how Google Shopping is an ad is really not worthwhile. When Google introduced shopping it was a search product, not an ad. https://en.wikipedia.org/wiki/Google_Shopping#Change_to_paid...
The investigation was concerned with Google's introduction and promotion of Google Shopping in its search results, not with whether it's a problem for Google to show product listing ads.
Trying to cleave out different verticals from a search engine would be a major disadvantage to consumers.
> I live in a city where Google Maps has, until one month ago, had basically no traffic data, satellite imagery was from 2004, and even today still supports no transit.
> Here maps has up-to-date imagery of highest quality, has traffic and transit data.
> Yet, many people use Google Maps, even if that provides worse results, sends them on roads that don’t exist anymore, etc.
> When one recommends they switch, they say "but Google Maps is preinstalled, installing Here Maps would be so much effort, and it doesn’t work with Ok Google".
Do you think this is an advantage to consumers, that they end up with the worse product?
Google is a monopoly, is a bad actor and this ruling is not a useful solution.
I think the government needs to break up the cable & telco rent companies. Conversely, I blame the market for empowering Google to this level and the market needs to fix it or accept the tradeoffs. I think the price is too high, but I'm in the minority of DuckDuckGo users who limit chrome / gservice usage. Even I touch some part of it daily as it is nearly impossible not to. I'm not as dogmatic as RMS so using some useful parts sometimes is fine, but the issue is almost all core services are being leveraged by almost everyone all the time and that is dangerous if simply begause it's a single failure point