Google is increasingly divided over supporting the news industry, either through direct funding, or with a set of initiatives directly aimed at improving publishers’ bottom line.
At the company headquarters in Mountain View, the lingering question is: should we continue to help an industry that: (a) tends to consider us as another teller for subsidies, (b) has shown little ability to use the full extent of the toolbox we built, (c) has repeatedly turned against us by lobbying Brussels?
Right now, no one is happy with the recent turn of events. At the core of the resentment is Brussels recently adopted Copyright Directive and more specifically, two dispositions: Article 13 and Article 11.
Article 13 is essentially targeted at YouTube and similar platforms that harbor user generated contents. According to the new legislation, online services are immediately liable for any material uploaded by users that infringes copyright. The argument is that YouTube is loaded with content for which fees owed to copyright holders are rarely paid in full.
It would take a dozen of Monday Notes to expose the full extent of the arguments on both sides. The general idea is YouTube is too lax when it comes to enforcing copyrights. It is true that the abundance of content available on YouTube is staggering, to the point of being suspicious. If you are penniless, you don’t need to subscribe to a streaming service to fill your hard drive with hundreds of hours of good quality tunes, ranging from complete albums to full concerts. Does YouTube redistribute due fees on this content? At the very least, it has deployed a comprehensive arsenal to do so. When a copyright infringement is suspected, YouTube is generally swift to take down the content. Around 2007–2009, after a notorious lawsuit from Viacom that found scores of illegal pieces, YouTube implemented a system called Content ID that checks content against a database of copyrights and automatically collect payments.
In an op-ed published on November 12 by the Financial Times, YouTube CEO Susan Wojcicki emphasises the efficiency of the system:
“More than 98 percent of copyright management on YouTube takes place through Content ID. To date, we have used the system to pay rights holders more than €2.5 billion for third party use of their content. We believe Content ID provides the best solution for managing rights on a global scale”.
She also suggests that the sheer size of YouTube, with 400 hours of video uploaded every minute, makes the task of enforcing copyrights nearly impossible. More broadly, hundreds of companies that deal with UGC could be affected by a narrow interpretation of the directive: blogging platforms (including Medium), dating sites, audio-sharing systems like SoundCloud, open-source repositories, review-powered sites (cooking, lodging, travel), etc. In fact, nearly every publisher is susceptible to host copyrighted content regardless of the precaution it takes. That is the argument invoked by a group of internet architects — including luminaries such as Tim Berners-Lee and Vint Cerf, who wrote an open letter (PDF here) that basically states that Article 13 is camping down the world wide web:
“By requiring Internet platforms to perform automatic filtering of all the content that their users upload, Article 13 takes an unprecedented step towards the transformation of the Internet from an open platform for sharing and innovation, into a tool for the automated surveillance, and control of its users”.
Should Article 13 be applied in a restrictive manner, scores of lectures and tutorials containing fragments of copyrighted material would become technically illegal. The formidable richness of the internet is definitely at stake here. Should a YouTube channel about the techniques of cinematography be taken down because it dissects a series of unlicensed shots by David Fincher? What about this lecture about space exploration that didn’t ask NASA for permission to show clips of the ISS?
The digital news industry is marginally concerned by Article 13, even if it contributed to its adoption.
But for another piece of this controversial legislation, European media furiously lobbied Brussels for a result that could backfire spectacularly.
Article 11 contains a disposition under which the smallest fragment of a press publication should open the right to financial compensation. In the crosshair, the snippets of an article that appear in a search engine results page (SERP); since Google is the dominant player in the search engine sector, it is fair to say that it is the main target.
Publishers who favor this legislation — most of them are in Europe — defend the following rationale: Google is making tons of money by displaying links, and snippet in their search page, and they want a piece of it.
This posture disregards four significant factors:
1. Traffic of news sites depends heavily on Google. It varies widely from publication to another, but the search engine is by far the dominant referral of traffic: at the end of last year, Google search alone was accounting for 38 percent of the visits sent to a sample of 250,000 mobile and desktop sites monitored by Shareaholic. Adding up the traffic sent by Google News, for many news outlets, the reliance on Google is above 50 percent, far outpacing the benefit of social networks (Facebook, mostly) whose share is dropping at a fast pace; at the end of last year, social accounted for 18 percent in visit referrals.
Twitter and Facebook have declined in their share of traffic sent to news sites.
Facebook traffic to publishers is down so much (nearly 40%) that according to Chartbeat, “a user is now more likely to find your content through your mobile website or app than from Facebook.”
Google Search on mobile has grown more than 2x, helping guide users to stories on publishers’ owned and operated channels.
Direct mobile traffic to publishers’ websites and apps has also steadily grown by more than 30 percent.
The chart below is eloquent:
2. The traffic sent by Google is so precious for publishers, that collectively, they spend millions of dollars and euros in maintaining teams in charge Search Engine Optimization. We can assume that these massive investments in SEO are primarily driven by a clever cost vs.benefit calculation.
3. Publishers themselves fostered the snippets/link sharing industry by allowing RSS Feeds. Blinded by the idea of traffic-at-all-costs, they profusely distributed their content through RSS Feeds, in many cases giving the full text of articles. In doing so, they contributed to the creation of the aggregation industry, entirely initiated by data-driven, agile digital natives. That was a major train (of many) missed by the news industry. Aggregators were crucial to the explosion of mobile applications that coalesce news content in a single and convenient place.
4. Copyright laws, especially the Berne Convention, already protect copyright holders thanks to a detailed (16,000 words!) legal arsenal.
This tune has been heard over and over. The rationale is that Google is selling ads against the snippets and very few people click on the links display in SERPs.
The first argument is only partially true: in Google Search, if you enter a typical news-oriented query like “raqqa evacuation” or “trump california fires” the result page will not carry any ads (there is no point at buying ads against such tragedies); but if you enter a consumer-related query like “best hair-dryer”, the SERP will return a few news sites, but many product review outlets and this result page will carry lots of ads. Then the simplistic view that Google makes a lot on news snippets does not hold. For Google News, the case is settled: Google does not sell ads there.
As for the second idea — the proportion of clicks sent back to publishers — it varies widely. Research suggests that, as expected, the click-through rate (CTR) is unevenly distributed: roughly speaking, top positions (for general search) carry a CTR of around 30 percent then it falls quickly to a lower single digit. Plus, the first page of results takes it nearly all (90 percent of the clicks). Hence the battle to be at the top of the first SERP.
Having said that, publishers are right to point out that Google has failed to provide hard data on this. Except for a vague and global number often quoted (a billion dollars or euros), we don’t know precisely how much value Google Search and Google News are sending back to the publishers. The search engine has all the data down to the dollar, or the euro, for every publisher simply by looking at the traffic sent, multiplied by the average revenue per page. The reason for this numbness in communicating lies in the engineering culture is a sense of self-righteousness at the top of the company. Surely, it doesn’t help its cause. Everywhere in the US, Europe, or Asia, Google has great communication teams whose hands are tied.
Until today, Google had maintained a tough “we-will-never-pay-for-snippets” stance. Except that now, the obligation is carved in European law, reigniting the fantasy of European publishers of a looming windfall.
In France, the hazy anti-Google feeling combines with a deep-rooted culture of subsidies — which contributed to severely hinder innovation in the French digital media industry. Now the prospect of yet another bonanza arouses everybody. Publishers fantasize of amounts of 50 to 60 million euros per year coming from the “link tax”. They might be disappointed.
First, Google is likely to resist. Much less for the amount of money in play than the precedent it creates in Europe and across the world, which could easily translate into billions of dollars.
Second, in practical terms, making Google pay for snippets is not easy to implement. Who should be eligible? Legacy media? Any news outlet that maintains a newsroom whatever its size? We are in for a long and laborious discussion. And that’s just for France, the gold standard of public and private subsidies. How to adapt the concept to the other 27 members of the European Union? And what about other territories that will ask for the inevitable Most Favored Nation Clause?
Hence the emergence of hard-liners who toy with the idea of killing Google News altogether and de-indexing news content in search. It’s a credible threat: Google did just that in 2014 when Spain passed similar legislation, as expected, traffic plummeted. In Germany, a comparable anti-snippet law was rendered powerless by a group of publishers who did the math and offered their content for free in exchange for the usual stream of clicks sent back to them.
More broadly, many see the EU legislation as harmful for the entire ecosystem. Last April, a group of 229 academics in the field of law and intellectual property have signed an appeal to the European Parliament to denounce what they see as “a bad piece of legislation”, to no avail.
By pushing to the “link tax”, publishers are shooting themselves in the foot three times over. One, there is a tangible risk that Google opt for the Spanish/German jurisprudence. Two, the optics will look terrible: by persisting to collect a small revenue from snippets, publishers will seem to wage a rearguard battle. Three, the news publishing world has more appealing options when it comes to working with Google at improving the economics of their ecosystem. The search giant is already investing hundreds of millions of dollars for technologies that could directly, or indirectly, benefit to the news media. So far, publishers haven’t used the full extent of it. It’s time to “think different”.
In a next Monday Note, we’ll look at some ideas for healthy cooperation between the news industry and Google. Then, we’ll talk about what the search giant could do to fight misinformation.