Inside Google’s AMP: Have Google and News Publishers Found Themselves on the Same Page?
WWGD should be a best-selling T-shirt. “What Will Google Do?” slips into many a conversation, most of them private, as the search giant’s massive workforce can be trained on most any digital opportunity and produce a big impact, if only sometimes a huge breakthrough.
Recently, the answer to the WWGD question has been AMP. AMP, as in Accelerated Mobile Pages. If you had any doubt that 2016 has become the Speed First year, AMP answers that question.
AMP’s promise is ample: a 15 to 85% improvement in news load, a speed increase that’s been proven out by publisher testers. How does Google do it? You can deeply dive into its techniques, which include stripping away of much HTML and JavaScript code and “really intelligent pre-rendering and pre-caching” of content. Presto: open-source AMP HTML, or as it’s been called, HTML without the bells and whistles.
Google’s offer is one of carrots and sticks. The big carrot: Its world-class engineers offer up this new wannabe open-source standard at no direct cost to publishers. The stick: Google clearly values speed, and will likely value AMP specifically, as it ranks news sites in its search rankings. Though Google’s dominating power as a news referrer has lessened as Facebook, Twitter and LinkedIn’s influence has grown, publishers still depend on Google for 20% and more of their traffic.
First published at Politico Media
Follow Newsonomics on Twitter @kdoctor
The future? Get fast … or die on the third page of search results. (For more on search impacts, Search Engine Watch’s Kenny Chung details those search impacts and raises good related questions.)
Though publishers can certainly be cautious about further entangling their businesses with Google, and of the apparent trade-offs they believe AMP may necessitate, many have signed up. Out of the lab, 40 publishing partners – including BuzzFeed, The Economist, Gannett, Hearst, The Huffington Post, The New York Times, and Vox, and eight members of the European Digital News Initiative – have committed to testing AMP.
Everyone sees AMP as a work in progress; it’s still in the testing stages. As it urges publishers to accept a new standard, Google’s gotten lots of feedback on the current product, and what AMP further needs.
Here Google has received better grades on responsiveness than in years past. In fact, we can see movement in what has long been a Google Knows Best culture. Why? That’s a key question given the often rocky Google–news organization relationships over the years.
In a far-ranging recent interview in Building 1900 at the Googleplex in Mountain View, Calif., I spoke with two head Googlers about AMP, its origin and placement amid Google strategy. In my experience, Googlers seldom want to acknowledge that the wider world – or, perish the thought, its competitors – affect its strategic moves, and this conversation started out along those lines, but quickly turned more candid.
“I think the web is a beautiful, beautiful thing for the world,” Dave Besbris, Google’s VP of engineering told me. Bez, as he is widely known, led work on Gmail and Google+, among other high-profile initiatives, and some interpret his work on news-media-oriented projects as a good sign.
“It [the web] really hasn’t been living up to a potential that we think it can be inside mobile applications,” Besbris explained. The rest of your mobile applications — whether it’s a news reader or a messaging app or a social network — feel very fast and zippy, and users are really accustomed to things on their phone being very, very quick.
“While the mobile web … has this incredible ubiquity, its performance hasn’t felt like it fits the rest of the phone.”
Google makes this point strenuously: AMP is meant to move article reading to an instant experience; it’s an initiative focused exclusively on those who create articles – publishers, large and small.
Article publishing shares a common set of characteristics, Richard Gingras, Google’s senior director of news and social products, said. “We looked at it and said what all computer scientists say when there’s something that’s in common, we factor it out and we optimize the heck out of it, and we share it.”
Gingras is a familiar face on the news media circuit. He jokes about Besbrich’s relative youth as he describes when he first got into the online business in the early ’80s. With Apple, Excite@Home, Salon and, going way back, Cybernetic Data Products on his resume, he said: “I’ve had this opportunity to work at every different dimension in the ecosystem.”
Both acknowledge that Google’s more direct collaboration with publishers marks a departure for the company.
Google’s New Collaborative Spirit
In fact, we can see that the development of AMP took off as Google began meeting with its European publishing partners, to form the Digital News Initiative (DNI) in the spring of 2015. That initiative grew out of what we can call the European war on Google. Google has fought numerous skirmishes on the continent for most of a decade, but more fronts had opened, and had intensified more recently, to the point of actually threatening the growth rate of Google’s European-based business. DNI grew out of that firestorm, and then AMP grew, in part, out of the DNI push.
While Google has long worked with news publishers, largely as a vendor, it’s seldom been known to be an active collaborator.
I pressed Besbris and Gingras on the question.
They both emphasize the change in attitude.
“I think there’s two reasons,” said Besbris. “One is there has been fresh leadership in here. Sundar [Pichai, the recently named Google CEO, as a result of Google’s Alphabet restructuring] is a very strong leader who’s very interested in partnering with people and doing things the right way. He brought me in. I feel the same way. We have new people who have taken just frankly, taken different approaches than we have in the past.
“But that’s not the primary reason …. Google has approached getting the web faster … through changing the browser [Chrome].”
The company has learned, he said, that the browser approach can only take the company so far. Now, it must “reach out to people who are creating content on the web” and provide better tech at the point of creation.
Besbris said: “It’s very important to us that this is the kind of thing that Google doesn’t typically do where we work with somebody beforehand, before we announce it …. As a company we’ve been spending a bunch more time with [publishers] to really understand what their needs are.”
Gingras quickly added that’s it’s a matter of shared business objective.
“The reason is really simple. We shared a very common objective. I think that in a lot of the deliberations as the ecosystem has developed, sometimes that point gets missed. It’s very crystal clear today. That common objective is, if you look at it, is Google’s product and services, specifically Google Search, its value is based on having this open ecosystem of expression.”
Not-so-strange bedfellows, as Gingras sees it, in a fast-changing media world.
AMP in a Whirlwind of News Business Model Shifts
That competitive landscape, including but stretching beyond Google’s European travails, provides essential context for understanding the AMP innovation.
AMP may be a protocol, a change in coding that most non-technologists will strain to deeply comprehend, but interwoven in the AMP rollout is the battle landscape of the modern web. Pick your storyline du jour, and Google’s AMP foray weaves in, out and around it.
First, there’s the deep 2015 recognition that mobile is the way of the world. AMP is exclusively about mobile.
Ad blockers? AMP speeds up mobile page loading, reducing the frustrations of readers that lead to greater ad blocker adoption. Further, its simple-is-better approach is hostile to intrusive ads that drive users crazy.
Then, there’s the story that has dominated the year, the 2015-born Age of Distribution. Facebook, Apple, Snapchat and others make publishers offers they can’t refuse. All of a sudden, platform companies host larger volumes of whole content, not just links. The platform’s case: Our platforms can deliver your content in an eighth of the time that your link (or you) can. AMP aims to be free, open web equivalent in this Speed First era.
Of all the platform distribution plays, it is Facebook’s that is most worrying to Google. From time spent to mobile revenue growth, Facebook keeps on surprising Google with its ability to take away business.
Both Facebook’s Instant Articles and Apple’s new News intend to pull readers away from the ungainly, slow mobile web, mesmerizing them with hours of endless pleasurable variety within their own app. To that foray, Google’s AMP issues its own, parallel to the Guardian, cri de coeur: Open! Or the wonderful web will die.
Gingras positions Google as the publisher’s friend in the current chaos of distribution and monetization.
“We see it in the result of proprietary platforms saying, ‘Come let us host your content. We’ll make it fast for you.’ We’ve not suggesting that a publisher shouldn’t have a deal with Facebook or anyone else, fine, if there are benefits there, do them. What we’re uncomfortable with is the notion that a publisher, particularly the hundreds of thousands of publishers, not just the big ones, but also the teenie ones, have an opportunity to get their content at speed in front of audiences without having to do a deal with a platform to accomplish that. That’s a core objective. That’s why speed matters.”
The common enemy, he’d say: Those that put boundaries around content.
“How do we make sure that that vibrancy and richness of the world-wide web continues in a larger environment where we are seeing the distribution ecosystem being also populated by controlled distribution environments like Facebook or Apple News?” said Gingras.
Note the pejorative “controlled distribution environment.” That’s Google speak for other people’s apps. One man’s walled garden is another’s Garden of Eden.
As a business, Google lives and dies on the open web today; it’s the lifeblood of its search business, a business that can slowly starve as smartphone users spend more and more time inside their apps (“Why Native Apps Still Matter in the Age of Distribution“). On the horizon, Google must spot numerous threats its long run of hegemony.
There’s certainly nothing wrong with self-interest, though, and Gingras’ point that the interests of Google and publishers may now intersect in ways they hadn’t previously.
The AMP/Facebook Instant Articles war offers an exemplar of Silicon Valley’s absolutely Darwinian culture, punching and counterpunching at light speed. We see the behemoths setting the agenda for everyone else. We – readers, viewers and listeners – are at times great beneficiaries of these contests, and sometimes we’re just the collateral damage.
Take this newfound issue of speed.
As Scott Gilbertson pointed out, in a good Ars Technica piece exploring the edges of AMP, speed is nothing new as a digital issue. “The Web, while certainly improved from the days of 14.4k modems, has never been as fast as we want it to be, which is to say that the Web has never been instantaneous.” Indeed, we’ve lived with various versions of the World Wide Wait for two decades now, and solutions such as AMP or Facebook’s IA are just a couple of ways to address our impatience. Increasing network speeds, decreasing network latency and speeding up mobile browsers are among the others that Gilbertson notes.
Facebook smartly made speed an issue of the year, and the marketplace has followed, just as a politician might seize on debt reduction or immigration, even as those issues have long simmered.
All that said, AMP, in its demos, is devilishly and seductively fast. So publishers test, collaborate – and ask for more.
Why Publishers Are Cautious
I talked with several publishers highly experienced in the art and sciences, of these deals. Many have become the ultimate pragmatists, understanding that every one of these partner deals involves trade-offs.
They love the speed, of course, and know that it will help their businesses (though how much is a key question), but they have to consider the negative impacts.
Publishers have isolated at least four.
AMP “simplifies” advertising, and if that means less impactful, interactive advertising, it could reduce effectiveness. It’s not the nonsensical takeover ads that are targeted here; sophisticated branded content campaigns can involve a complexity that may tangle with AMP’s simplicity. That’s an issue, as it the integration of AMP with ad platforms other than Google’s own.
As one savvy publisher points out, the AMP move is really a twofer strategy. The allure of fast-loading publisher pages mates with Google’s strengthening of its big advertising platform business.
“The AMP format in effect makes Google the gatekeeper on ad formats and analytics scripts. This will allow it to prevent heavy ads and intrusive scripts from being deployed on mobile, both of which are vital if the rise of ad blockers is to be stemmed. It may not work, but it’s a great move.”
Analytics present a second concern. AMP relentlessly reduces tag clutter, and, at this point, that means fewer tracking analytics, just as publishers are investing in such data science.
Then, given the ascendance of paywalls, how AMP will play with them is a paramount point, especially for those publishers who offer much more than one-size-fits-all subscription offers. With reader revenue the probable lifeline of news publishing into the future, the question looms large. Richard Gingras says it’s on his list and that’s he optimistic about it. “There’s some things that we can do to really support paywalls.”
Then, finally, there’s the resource question.
The still-open AMP question: How much work will publishers have to do? Given near-universal resource constraints, that’s a big issue. Big publishers have already devoted good resources to speed improvement this year. AMP builds on that idea, but some early testers are quite concerned about the new workload it may require.
Of course, AMP isn’t the only resource question. Each of the new big distribution deals – Facebook’s, Apple’s, Snapchat’s and more to come – require some new level of work. Every platform, with Google at the head of the list, would like to be the standard, but rather than standardization, the year has seem more splintering of distribution tech.
Gingras’ solution: make AMP the industry standard feeding Facebook, among other platforms.
“What I expect we’ll see months down the road is if I’m sharing an article on a publisher’s site, that publisher is going to want me to share the AMP version of that article, because that AMP version of that article will surface very fast in Facebook, for that matter, everywhere. Get that speed advantage and allow them to get their content out there with that business model intact, fully integrated. It’s their business model.”
Not unexpectedly, he offers up AMP as a labor saver for any publisher, big or small.
“I had this conversation with [New York Times CEO] Mark Thompson. They’ve got a big engineering staff. They too can’t scale to the demands of the environment we’re in. They’ve got to have a great iOS app. They’ve got to have a great Android app. They’ve got to maintain their CMS systems. They’ve got to do clever new things.
“One objective of what we’re doing is how can we reduce those efforts such that organizations, even theirs, because Mark will say, ‘We can’t keep up.’ Then how can a small organization that has two engineers, or four, keep up?”
There’s a logic of standardization to be sure, but it’s no accident that involves more roads leading back to Mountain View. Ironically, news publishers’ long reliance on Google has lessened, given the business importance of Facebook, Twitter, Apple and others, in providing audience and new ways to make money. For the first time in a long time, publishers actually see a deal-bettering competition for their partnership from several platforms – and they won’t easily trade that in.
Lastly, there’s the sheer speed at which AMP and its cousin initiatives are moving. Each of the forays are self-acknowledged to be works in progress, beta tests that clearly try to claim turf. Given the brutal head-to-head platform competition, that makes sense – but it leaves publishers either having to wait, or bet on the come.
