Problems & Solutions: Analyzing The EU’s Antitrust Charges Against Google Over Shopping Search
The EU has raised four concerns about how Google has dominated shopping search. A close look at each charge and possible fixes that might emerge.
After years of investigation, we finally have a formal EU antitrust complaint against Google in search. It’s narrow — all about shopping search. Here’s a look at the specific charges and what they might mean for the search engine.
The EU is issuing a “Statement Of Objections” to Google outlining its concerns, where it will back these up with what will be pages and pages of evidence. The public won’t have access to this, but there’s an excellent chance it will leak. Until then, all the public gets is a short fact sheet that outlines the EU’s four concerns, when it comes to shopping search.
I’m going to address each of the concerns below, as well as potential solutions. Be aware that this is a long article. If you really care about the issues, from the perspective of someone who has tracked the search engine industry for nearly 20 years, read on.
It’s very important to understand that with all of these concerns, Google can reject them as invalid, offering up its own evidence of why they are wrong. The EU might agree with some or all of those arguments. So, the solution to everything below might be that the charges are ultimately dropped and Google doesn’t need to change anything.
Possible, but not probable. On to the allegations….
Problem: Google Gives Itself Preferred Prominence
The first charge is this:
Google systematically positions and prominently displays its comparison shopping service in its general search results pages, irrespective of its merits. This conduct started in 2008.
There’s no doubt that Google does favor its shopping search results. Here’s an illustration of how that looks:
The search above was done on Google UK, for “dvd players” (US ads are showing because I’m in the US and the VPN I’m using to pretend I’m in the UK doesn’t override this, for some strange reason). Ahead of Google’s regular web listings is a special box, listings that are pulled out of Google Shopping.
Technically, these listings are all ads — referred to as “Product Listing Ads” or PLAs. In the first page of results, no listings from any competing shopping search engine appear. If Google hadn’t added this box, then it’s likely that its own shopping results wouldn’t be so visible.
Case closed! No. There are a gazillion caveats to consider, some of which I’ll explain more below. Perhaps the most important, in showing its shopping results in this way, Google operates just like any other general purpose search engine with subject-specific search listings, also called “vertical search.”
Search engines operated this way before Google existed (yes, there was a time like this). Even as Microsoft, which is a major backer of the EU action, does this with its own Bing search engine. Consider a search for “pictures of flowers,” as shown below:
That’s a search done on Bing, where Bing has prominently pushed its own image search results ahead of everything else, including other image search engines that exist. Few would seriously argue that it would be better for Bing’s users if it provided text-only links to places with flowers. The same is true when you talk about other vertical search results such as news, travel, local and shopping.
For background about this behavior both by Google and as an industry practice, see
- Does The FairSearch White Paper On Google Being Anticompetitive Hold Up?
- The Incredible Stupidity Of Investigating Google For Acting Like A Search Engine
- Dear Congress: It’s Not OK Not To Know How Search Engines Work, Either
- Bing’s Travel Search & Kayak Favoritism Angers No One, While Google’s Gets Headline Attention From WSJ
In general, you want a search engine — Google or Bing — to display vertical search results when relevant. It’s better for the user. That’s one reason the FTC cleared Google in its own antitrust investigation in 2013. But the twist with Google is its massive marketshare in Europe. It has 90% or more of the general search share in most European countries. With such a monopoly, it might have to act differently than others, especially if its behavior is deemed to harm competition.
That’s what the EU seems to feel. Another twist is that the EU seems to want to regulate the quality of Google’s results. That’s what the whole “irrespective of its merits” part is. The idea here is that other comparison shopping search engines might have better results and that Google potentially should be obligated to show them because of this.
In the US, that wouldn’t fly. There have been several cases where Google’s been found to have a First Amendment right to list whatever it wants, quality aside. The EU could be different. More important, it’s just bad for Google, which prides itself on quality, to be accused of showing results that aren’t the best.
Solution: Include Others As Prominently
Google had proposed a solution to the concerns of other shopping search engines in February 2014. This was to show listings from competing shopping search engines alongside its own:
That seemed reasonable to me but was objected to by competitors, with trade group FairSearch even calling it “worse than nothing.” The proposed settlement eventually collapsed. A reshaping of the proposal above might separate the artificial “Alternatives” barrier between Google’s own product listings and those leading to competing shopping search engines.
One issue for users is that Google’s listings aren’t for its actual shopping search engine. They’re for merchants that sell products directly. The shopping search engines don’t, so they’d be running ads that lead to their own listings of merchants. That’s two different things mixed together, but it could work. However, it might only work if some of the real estate was always guaranteed to include some of the competitors.
Solution: Ballot Box To Pick Your Own Shopping Search
Possibly, the EU could demand that Google operate a “ballot box” system, where Google users have to pick which shopping search engine they want to show up within Google’s regular search results. This would be similar to how consumers in the EU used to be asked to pick which browser they wanted, in the wake of the Microsoft antitrust case.
The difficulty is that consumers aren’t always logged into Google, so they might find themselves being prompted to choose in cases where Google can’t “remember” the choice they made. Maybe the requirement could happen only for logged in users. In any case, it would be fairly dramatic.
Solution: Close Google Shopping
Another solution might be that Google would simply close Google Shopping in the EU. For example, this is the home page of Google Shopping UK:
There’s a good chance few Google users ever go to the Google Shopping UK site itself, in order to do a search or browse products. Heck, even for the much more mature Google Shopping site in the US, behavior is more likely that people just search on Google itself and get shopping listings (which are ads) mixed into the main results.
If Google closes Google Shopping as a standalone service, then the listings in its main results might be viewed more like what they are: ads. This is exactly as happened when Bing closed its own shopping search engine in 2013. Google would then no longer “promoting” its “shopping results.” It would simply be displaying shopping-related ads, similar with how it shows all types of ads without objections.
Google would have to make two further small changes, neither of which would likely be that impactful:
Some elements of the current shopping box, where shopping ads display, would have to change as highlighted above. The “Shop for….” text above the ads is a link that leads people into Google Shopping to explore more, if they should click. That would have to go. Actually, I see it gone already when I use Google UK, but that might be some strangeness to using Google UK from the US. Below some ads are links to where the product has been reviewed with Google Shopping. Those would have to go, as well. Both are probably no great loss, likely drawing relatively few clicks.
Do that, and Google might be largely done with every issue the EU has about Google Shopping. Just kill it. That’s what Google did with Google News in Spain last December. Google users don’t seem to have cared much, likely because plenty of them still get news content directly inserted within their main search results. A standalone news site isn’t necessary for that.
Note that this would be harder for Google to pull off in local, where the industry trend for local search is that users seem to like getting dedicated pages about particular businesses, with user reviews. You know, like Yelp. But Google largely killed its shopping search engine over two years ago, when it shifted all merchants to buying ads and relied more on display of those ads within its main results. Officially killing it, at least in Europe, might have no major impact, since it would continue to show ads.
Google Doesn’t Penalize Itself
The second charge is this:
Google does not apply to its own comparison shopping service the system of penalties, which it applies to other comparison shopping services on the basis of defined parameters, and which can lead to the lowering of the rank in which they appear in Google’s general search results pages.
That goes right to what started this all. A tiny shopping search engine in the UK called Foundem got penalized by Google. Practically no one used it. I know — I lived in the UK for over a decade when all this happened. No one knew much about Foundem; it wasn’t a major competitor to anyone.
Regardless, Foundem got penalized by Google for things Google considered to be spam. Unlike most sites hit by spam, Foundem fought back with a unique argument — Google was simply trying to block a competitor. This overlooks the fact that Google regularly has included links and sends huge amounts of traffic to much larger and real competitors all the time. But that argument was perfect for other competitors concerned about Google (for some good reasons) to build it into a bigger complaint.
It’s certainly fair that Google doesn’t subject product pages from within Google Shopping to a penalty. But that’s also because Google doesn’t include them in Google at all. For example, here’s the Google Shopping page for the Xbox One. That page has no chance of showing up in Google’s main web search results because Google has blocked all product pages like it from being included. In contrast, this page from competing shopping search engine Nextag is included in Google’s web search results and has a chance to rank in a way that Google can’t.
Solution: Allow Google Shopping To Be Ranked Like Others
One solution to the penalty problem would be for Google to allow its Google Shopping pages into its web results, so that they could compete with others and potentially be penalized like others, if they’re deemed spammy or “thin” or otherwise lacking.
I doubt that would satisfy anyone, however. As long as the shopping ads continue to show, then the accusations that Google is favoring its shopping results (which really aren’t results but merchant ads) would continue. Ultimately, any solution here would be tied to solving the earlier objection.
Problem: Google Is Helping Its Shopping Search Growth At The Detriment Of Others
From the third charge:
As a result of Google’s systematic favouring of its subsequent comparison shopping services “Google Product Search” and “Google Shopping”, both experienced higher rates of growth, to the detriment of rival comparison shopping services.
Since we don’t have the Statement Of Objections with the evidence of this, I’ll point to a great section from a Wall St. Journal article today that outlines some of this:
Between October 2007 and October 2009, the number of unique U.K. visitors to product-comparison services including Shopzilla and Nextag dropped 41% on average, while visitors to Google’s Product Search service jumped 125%, Foundem said in its complaint, citing data from researcher comScore….
From the start of 2013 to early 2015, leading shopping and price comparison services in Germany, such as Nextag, Ciao.de and Twenga, saw their organic search visibility drop 91% on average, while visibility for Google Shopping surged 880% in the U.S. and more than tenfold in Germany, according to Searchmetrics.
Those are stats from two different services, and there’s a lot that could be unpacked about what exactly is considered Google Product Search, such as whether this is visits to that standalone site or just any appearance of shopping search results within Google.
There’s also Google pushing back on this in a blog post today, highlighting that if you consider Amazon and eBay to be shopping sites, they’re doing great while Google itself, not so much. Here’s a chart from Google about the UK situation:
Realistically, there’s likely little doubt that Google has grown its shopping search visibility by including shopping result initially as part of Universal Search in 2007 and later by making special shopping ad boxes as are shown now. It also wouldn’t be surprising if a wide-variety of changes, some aimed at fighting spam, some aimed at showing more merchants, has benefitted Google. The latter especially would have harmed the ability for shopping search competitors to have visibility within Google, regardless of the intentions.
Solution: Change The Web Search Algorithm?
Again, any solution here is likely tied to solving the first charge. If Google somehow gives competing shopping search engines more visibility within its ad space, they’d likely see growth — assuming they want to pay for it (which has been the trend for even non-Google competitors. The Google “free ride” has gotten less for everyone. See also: Facebook).
Alternatively, Google could make a change so that competing shopping sites have a greater chance of appearing in its web search results as they once had. Consider this search for “dvd players” again, this time focusing on the web search listings:
There are no competing search engines in those results, a consequence of a move Google made years ago to drive people to actual merchant sites (of which there are plenty) rather than to sites that simply make them search again for merchants. Also note the first listing: Amazon. It’s a weird hybrid, both arguably a shopping search site (because you can find so many products on it) as well as a merchant that actually sells.
Google could alter its algorithm to allow shopping search pages from competitors to have better odds of appearing in this mix. To understand how that might work, consider this:
Those are the top five search results for a search on “london hotels” on Google UK. Every link leads to a place where you can search for London hotels, rather than lead to actual hotels. A change to the algorithm for shopping like this might give competing shopping sites more visibility and traffic from Google.
Of course, above those listings is a big giant box for Google’s own hotel search service. That makes the likelihood that people will “detour” from clicks to competitors below much more likely. Expect that to be an issue if the EU decides to go after Google on the travel search front, which is hasn’t yet. But in the same manner, just giving more web search visibility might not be enough without a resolution to the first charge, about the prominent shopping search listings / ads that Google shows from its own data.
Problem: Google Stifles Innovation & Consumers
Finally, the fourth and last charge:
Google’s conduct has a negative impact on consumers and innovation. It means that users do not necessarily see the most relevant comparison shopping results in response to their queries, and that incentives to innovate from rivals are lowered as they know that however good their product, they will not benefit from the same prominence as Google’s product.
I’ve looked at plenty of shopping search results in my time. None of them are that great. They’re often crap. Sometimes Google is crap. Sometimes competitors are crap. Sometimes they all are. Sometimes they’re good.
I’m going to be plenty interested in the actual evidence, if it ever emerges, that the EU can prove that Google is consistently more crappy than its competitors. Moreover, it also seems hard to argue that the results aren’t the “most relevant” when the consumers themselves clearly haven’t been abandoning Google for other services.
Just to give a quick visual comparison, here’s Kelkoo UK, a major UK shopping search engine, with results for “Xbox One” shown below:
Here’s the same for Google Shopping UK:
I didn’t cherry-pick these results. I just searched for a product that popped into my head. Kelkoo gave me some console listings but also a lot of accessories. I wanted an Xbox One — not accessories. Google was on target with just consoles. Of course, some people might have wanted the accessories that Kelkoo offered. But the bottom line is that there’s no huge difference in these two players. Do another search, Kelkoo might be a bit better. Do another, Google might be.
Shopping search quality is different than local or travel search, by the way. I have a huge degree of sympathy for companies like Yelp and TripAdvisor, who have build wonderful ecosystems rich with quality reviews that can often easily be demonstrated to be better than Google Local reviews. But the EU is addressing shopping search here, and shopping search overall just hasn’t been that great.
Potentially, if the shopping search competitors had been getting more traffic, they’d be making more money and reinvesting in themselves, to the ultimate benefit of consumers. But my experience in watching the comparison shopping sites over the years is that most of the time seems to have been spent on figuring out how to get merchants to pay more to be listed, not by making a great user experience.
The most powerful argument for curtailing Google’s shopping search growth is probably that if you don’t have other competitors, Google could potentially ratchet up the cost to run product ads on its own site. In turn, that could potentially translate into higher prices for consumers.
Potentially, sure. But then again, Amazon has completely opted-out of Google product ads since the program began charging, to my knowledge. It seems to be doing just fine. In turn, Amazon puts pressure on the entire marketplace.
Solution: Fix Everything Above?
The only solution here seems to be related to all the other charges above. If the competing shopping sites feel they’re getting more visibility in Google, or that Google isn’t gaming things to help itself over them, then maybe we’ll get all this innovation that’s apparently been held back over the years.
Maybe. Then again Microsoft’s Bing, which invested hugely in search in the United States and lacked nothing in terms of cash to do so, decided the best way to invest in product search was to kill it and replace it all with ads. Just like Google did.
Personally, I’m not expecting the EU action today will ultimately bring any great benefit to shopping search. But I’ve been pessimistic about the space for years. Just use Amazon.
So what’s going to happen? Google now has ten weeks to respond to all these charges. I’d bet it will push back on everything. Then I’d expect we’re going to have round after round of negotiations, maybe an actual legal fight and in the end, the shopping space will have changed so much from where this all began that none of it will matter.
I certainly don’t expect some type of “Microsoft moment” where Google collapses. Ultimately, I’d bet we’d see some small changes applied to shopping ads presentation.
The EU has also said this:
The Commission continues its ongoing formal investigation under EU antitrust rules of other aspects of Google’s behaviour in the EEA, including the favourable treatment by Google in its general search results of other specialised search services, and concerns with regard to copying of rivals’ web content (known as ‘scraping’), advertising exclusivity and undue restrictions on advertisers.
If the EU ever does issue objections for other vertical search services, in particular local and travel, I’d expect a stronger case, especially in terms of stopping the scraping activity. Google has agreed to such things already in the US, as well as opening up stuff for advertisers. Those are relatively easy to solve.
More tricky is if there’s a demand to give local and travel search more visibility. With these cases, Google isn’t showing just ads. It’s showing actual content pages, because it has evolved into a content player, and it might stand to lose notable traffic if pressed here.
Then there’s Android. That investigation has formally opened, and that’s where the real action is. The search allegations are hard to press, when there are often good reasons for users to the things Google has done.
The “it’s for the users” argument holds less when Google exercises terms to force its apps and search onto Android devices that want to be called “Android.” Here, I wouldn’t be surprised if we ultimately saw some type of ballot box choice emerging where Android users have to pick their search engine and that Google gets limits on how much it can restrict handset makers from including rival products and services.