AMP (and Google’s news carousel) speed out of beta
Are you ready for Google’s Circle Game? Not as in Joni Mitchell’s epic song, and more importantly not as in the Google Circles that were a feature of the now-defunct Google+. But once again the "carousel of time" has brought us back to another circle-shaped Google innovation that is meant to shore up its position in the news game against big Silicon Valley competitors like Facebook and Apple: The launch of Google AMP and its sidekick, the Carousel.
You’ve started seeing more carousels if you do a mobile Google search — that dominating, graphics-heavy swipeable selection of news stories that breaks the left and right margins of the screen. It has now become the face of news searching on the smartphone and the tablet for Google users.
The carousels have begun populating and will be fully operational as Google’s AMP protocol comes out of beta (POLITICO: “What are they thinking? Inside Google’s AMP”) and fully into the public eye, its official launch now aimed at midweek.
Its launch has been accelerated, and in that speed-up, we can see the company’s mounting competitive pressures. What may look like a mere presentation change actually conveys a major strategic push by Google to compete with its increasingly troublesome nemesis up the Valley.
As Facebook launched Instant Articles last year (and opened it up to all publishers last week), the Googleplex has been abuzz with the existential Facebook semi-threat. Could Facebook and Apple and Snapchat take away substantial Google traffic – the nectar of its search-ad business – by moving billions of Internet minutes off the open web’s search platforms and containing them on new “closed platform” discovery apps?
The threat isn’t an immediate one only because of the $23B net income posted by Google for 2015. But it is aimed directly at the ecosystem that Google has profited from more than anyone else on the planet.
Facebook’s relative advantage (Fortune: “Facebook, Google and the battle for mobile intent”)in mobile ad growth, for instance, portends a troubling trend line for Google.
AMP, or Accelerated Mobile Pages, is one answer to that threat. As Facebook pitched its Instant Articles to publishers last year, it focused on one point: speed. If a publisher’s page might take eight seconds to open up from a FB link, it could open instantly, hosted on IA. Latency is killing you, Facebook told publishers.
Publishers have found themselves seduced by the logic, but they also found another way to apply it. As European publishers met last year with Google, through the Digital News Initiative (POLITICO: “Google to launch $150M news initiative with publishers”), the new partners collaborated on what became AMP.
AMP makes article delivery near-instantaneous on mobile devices, cutting down load time by as much as 85%. What Google is saying to publishers: You don’t need Facebook to obtain world-class speed. Just AMP-ify your pages.
As AMP takes off — you can see the demo at g.co/ampdemo — some “several hundred” publishers have been working with Google to do just that, Richard Gingras, Google’s senior director of news and social products, told me. Further, more than 1,000 domains are on board, having opted into the open-source code. Finally, dominant platform Wordpress is integrated.
Unsurprisingly, it’s the largest news publishers who have first adopted it, and done the work to get ready for it. That’s as usual a matter of both resources and awareness, with national sites engaged and most local news companies sitting on the sidelines.
One major national chain now works with just-born start-up Relay Media on AMP-ificiation. Dave Gehring, late of the Guardian and still working as a consultant to the Digital News Initiative, saw a business opportunity out of his work on AMP. So Relay Media has just been born. His pitch, as CEO, is simple: We’ll AMP-ify your content, adding code, to get you ready for the new standard.
Relay now heads to market with a couple of different business models — one built on revenue share and another on licenses — as publisher awareness of AMP grows.
What’s the big payoff for publishers to invest their own resources or buy services to AMP-ify? That’s uncertain. Some publishers hope for a 10-20% lift in engagement, given the new speediness; Google’s making no public claims.
It’s the connection between the new carousel launch and the AMP launch that provides the other “incentive.”
Only AMP-ified stories will qualify for the eye-attracting carousel; others will continue to run as text, below or above the carousel. Gingras says that the GOOG algorithms will continue to judge the value of the stories, both within the carousel, and those placed above or below it. Clearly, though, publishers see that if they are not represented in the carousel, their relative Google search traffic could decrease. Some, including Atlantic.com General Manager Kim Lau, have noted that on-site carousels haven’t always produced more engagement.
In fact, if we take a long view, if most news publishers move to the new AMP protocol, might this be a zero-sum gain? At this point, the speed gain is welcomed by all, but the business benefit highly unclear.
In Lau’s thinking, the move is a highly consequential one, even if that immediate benefit is unclear.
“Our strategy is to distribute as widely as we can,” she said. “There was no question as to whether we would participate. What Google has done is reset what mobile best practices were.”
Publishers, overall, give Google high grades for its AMP work. Typical is the sentiment of Financial Times managing director of business-to-consumer Jon Slade, who told me, “We're very pleased with how AMP has progressed in recent months …. Google have engaged with us throughout the process and AMP is so far looking like a product which works well for the consumer, and for the publisher.”
But it wasn’t always easy to achieve that balance. AMP’s blazing speed derives both from increased Google caching of news content and from stripping down code in the serving of pages. That stripping removed Java Script and other “heavier” elements — increasing speed, but sacrificing code that directly addresses lots of publishers’ business needs, and they cried out for code that could close that gap.
We can divvy up the main issues with AMP into three A’s: Analytics, Advertising and Authentication.
Analytics: Publishers increasingly want to know everything they can about content usage. Now, as AMP launches, Google has its analytics pipe in place. Publishers can use Parse.ly, Chartbeat, Omniture and other tools, integrated with the AMP protocol, to keep the data flowing.
Authentication: It’s been a top priority of the FT, the New York Times and the Wall Street Journal, among others, to be able to identify subscribers and “authenticate” them. Their increasingly nuanced paid digital subscriptions must work via AMP. That potential looked shaky, even a few months ago – requiring a fair amount of work – but is in place at launch to the satisfaction of publishers.
Advertising: Google had to integrate with a range of ad platforms and networks, including ad recommenders Taboola and Outbrain, to allow this form of monetization to work for publishers. Most of them tell me they are satisfied with version one of AMP and hope for improvement in version 2.
Chris Hendricks, McClatchy’s vice president for products, marketing and innovation, sums it neatly: “Everything’s not portable,” meaning that publishers can’t use all of their ad tricks of the trade as they test out new plays like AMP. Still, he believes, they must be tested.
Clearly, AMP-enabled advertising causes some issues in using interstitial ads and site takeovers; some of these more intrusive ads may be going away anyhow in the reader-sensitive, ad blocker era. Lau told me that AMP doesn’t currently allow the re-sizing of ads, a functionality she hopes gets added. Some publishers have also raised questions about supporting full programmatic ad capability.
In fact, in the hubbub of launch, there’s a lot of incomplete information or understanding out there about AMP. Some articles have suggested that video is adequately supported, while Google tells me that “we have integrated support players like Brightcove and Vimeo. YouTube is there as well. And finally, using amp-iframe, a publisher can embed basically any video player into their pages, even a proprietary one.”
Clearly, after a half year of intensive work led by Gingras and Dave Besbris, Google’s VP of engineering, there is much accomplished, and there are lots of loose ends. V2 beckons, and for the publishers most involved, a sense of optimism that Google will indeed come up with the necessary solutions for them. In fact, at this point, the AMP-ified carousel will only be found on Google search. Within a couple of months, it will extend to Google News itself and the news tab on Google search, Gingras told me.
As I’ve pointed out, that’s somewhat a reversal of form in the long-fraught publisher–Google relationship, which has included a fair amount of rancor and some European political and legal action. In fact, as the new platform wars have ignited over the past year, Google’s newfound ability to work with news publishers has become a whiff of fresh air.
We can’t discount the increasing competitive environment as one reason for the new spirit, but publishers hope the good cheer can be stoked. As one savvy publisher surveyed the off-platform landscape, those observations were confirmed in other interviews:
“AMP feels promising, Facebook is uncertain but there’s room for optimism. Apple News ... not so much. Lack of data back from Apple is a real issue. We have had zero data since December. The system is broken and they've not fixed it. Ad monetization is poor.”
Google intends AMP then as a multiple differentiation. It will try to gain “good guy” status with publishers. It wants to be seen as a big advocate of the free and open web. Finally, in the perfect Google-created world (Alphabet for All), AMP would become a new standard for mobile web news delivery.
Richard Gingras tells me that LinkedIn and Twitter have both signed on to it already, but there’s one company that hasn’t. “Facebook.”
“Facebook, we hope,” says Gingras, “will ultimately do the same.”