Nieman Foundation at Harvard
In the blink of a digital era, The Washington Post’s Arc publishing platform has sprinted from an experiment to a full-on strategic business.
Arc is now used by more than 30 clients operating more than 100 sites on four continents. It’s not the industry standard, but it’s not too early to call it an industry standard. But its ambitions are still nowhere near met. Now the Post is moving Arc into a new phase, talking of a connective effect that could impact the face of the business formerly known as “newspapering.”
Arc wants to be more than a technology stack — it wants to be a network.
“Arc is reaching a critical mass of most of the advertising markets in the United States, the major markets,” Shailesh Prakash, chief product and information officer for the Post, told me recently, listing off cities where it has customers — New York, Los Angeles, Chicago, Boston, Philadelphia, Dallas, Washington.
How do publishers traditionally make money? Two ways: from advertisers and from readers. Arc has plans to be a player in both for news sites around the country and around the world.
First, ads: The Post will begin testing an ad network based around Arc clients in 2019. “We’ve got this technology we call Zeus, which basically does a more effective header bidding. Does things like autoplay for video, refreshing of ads. And I think our sales team is fairly good on the programmatic side to figure out how to get more, to squeeze money.”
Can the Post convince publishers in all these different markets to let the Post power some of their advertising?Prakash thinks so. “It’s good for us because The Washington Post now has a wider network to sell to, and it’s good for them because we think we can raise their CPMs. Certainly, we raise our own CPMs over here with our technology. So can we do that for others?”
He says Zeus increased the Post’s CPM ad rates by 30 percent. What drives that growth? Speed, viewability, and optimizing programmatic bidding, says Prakash.
Next, readers: The Post’s new ambition is to become a key part of its customers’ digital consumer revenue too — that is, the digital subscription businesses of its customers. At the end of this month, starting with the New Zealand network of sites NZME, the Post will launch a paywall product Prakash calls “a CMS for subscription.”
“Everybody has a paywall, but there’s a lot of back and forth that goes on every time you want to try a particular offer,” he said. “Let’s say Mother’s Day is coming, and you want to be a retailer — put some kind of an offer out there. There’s a lot of friction between the business team that wants to do and the design team and the engineering team that has to implement it.”
At this point, the Post plans to market this offer-making technology to Arc clients. As it develops a more sophisticated propensity modeling — intended to parallel the work of such paywall-tech incumbents like Piano Media — the Post and Arc could get more fully into the digital subscription business. It’s not hard to imagine networked, perhaps bundled, digital subscription offers, though that’s not in immediate plans. The subscriptions product will be in beta this fall and “will become generally available early winter,” says Prakash.
If this story sounds familiar, it’s because this isn’t the first time a Jeff Bezos-led company has tried to pull this trick — becoming the infrastructural underpinnings of an industry by offering products both good enough and easy enough that they fade into the (highly profitable) background. That’s the path Amazon Web Services took to becoming a profit-producing machine. And AWS followed the same path Arc wants to head down: a technology stack built first for internal use (running Amazon’s servers, running the Post’s digital publishing), developed for a “first and best customer,” and then licensed out to the world.
As the daily newspaper world continues to reel — from print ad loss, uneven digital moneymaking, and intensifying cost pressures — the Post wants to make Arc central to the business strategies of the trade.
Post publisher Fred Ryan positions the networking moves this way: “We’re not pitching ourselves as a replacement. We’re pitching it as an upgrade. It is something that can do more than the system that many publications now have in place.” Ryan characterizes Arc’s journey: “We’ve looked at Arc as a viable business for quite some time, but I think we’re at the stage now that we could say that it is a thriving business. There’s incredible demand from not just legacy news organizations, but from digital native organizations and brands as well.”
The opportunity is to connect up the new holy trinity of digital publishing: technology that can improve the publishing process, digital advertising optimization, and digital subscription development. And Bezos’ Post wants to be at the very center of it — globally.
Will it work? Success won’t be easy, or likely quickly proven. The landscape is littered with failed and ho-hum ad networks. Publishers will have to be convinced that the digital subscription tools of The Washington Post are applicable to their own much more local businesses.
But Arc continues to have success making that pitch for other parts of publishers’ tech stack. It’s set for major growth: Prakash says that by Q1 2019, Arc “will power over 400 websites and serve over 10 billion page views per month as ongoing implementations go live.” Beyond the U.S., it currently serves top-three news and information sites in Argentina, France, Canada, New Zealand and Spain. In September, Arc announced both The Dallas Morning News and Spain’s influential PRISA Noticias, publisher of El País, as customers.
Why are publishers adopting Arc? Major chains and titles — including Bonnier Co., Tronc, Advance Local, The Boston Globe, The Philadelphia Inquirer, Canada’s Globe and Mail, Latin America’s fast-growing InfoBae, and the New Zealand Herald — cite both the economics and the Post’s deeper understanding of the news business in the digital age as drivers. Prakash says the typical savings a single metro paper might derive from switching to Arc is typically around 15 to 20 percent, but that “there are other places that they’ve worked on where we believe the savings are even closer to 60, 70 percent, simply because they have so many vendors.” (That reduction in vendor count is something publishers often cite — both for the cost savings and just less time dealing with multiple companies and focusing on integrating their various products.)
The Post has built a staff of about 110 directly supporting Arc, about 100 of them engineers. That contingent makes up about a third of the Post’s technology staff of 300, doubled percent since Bezos bought the Post and heavily reinvested heavily in both its content and technology. In early summer, says Prakash, the Post authorized 50 new hires; 15 of those are now on board.
Arc Jedi warriors with our master, the Yoda of SaaS. Great day today planning the future of publishing with the greatest innovator of our time. #innovate-we-must #disrupt-we-will @WapoEngineering @wapoarc @WashPostLife pic.twitter.com/6fxre1DW0f
— Shailesh Prakash (@sprakash007) September 15, 2018
Prakash, a former Microsoft executive, began development on Arc six years ago (before Bezos’ purchase), but it emerged fairly slowly into the public eye. Today, it has its own P&L and is “slightly profitable,” he says.
Arc now contains 17 increasingly integrated modules. These power the newer essentials of the news trade, from newsletter creation to native app creation to recommendation engines to A/B testing to audience analytics and beyond.
Beyond economics, Arc’s market timing fits in with a common understanding arrived at as the news industry’s disruption moves beyond its adolescence: In-source or out-source. For years now, newspaper publishers have adopted that philosophy when it comes to Old Iron: Either make your press, production, and distribution facilities a business unto itself — taking in outside work, including from your print competitors — or outsource the job. If they outsource that work, they save expenses, even if their payables obligation and inflexibility increase. That’s a transition many publishers have made.
Arc then fits into that thinking’s maturation into digital. Talk to people at newspaper conferences (or even at such digital-only ones like the recently concluded Online News Association confab in Austin) and they’ve come to a shared belief: License the best cloud-based publishing platform you can afford.
That’s why Arc isn’t the only player in this new game of outsourcing.
WordPress is developing a cloud-based content management system on WordPress.com (with partners including Spirited Media and the News Revenue Hub) aimed at the needs of small-to-medium local news sites. The system would incorporate a variety of innovations developed by leading news publishers. The idea, as with Arc: provide integrated services to publishers and drive continued development more quickly and inexpensively than can be done on individually hosted instances, whether of WordPress or other platforms.
WordPress already powers many smaller news publishing sites, while its enterprise group, WordPress VIP, serves larger clients, including News Corp., Penske Media, Hachette, USA Today Sports, and VentureBeat.
Acting on similar principles, digital news veteran Merrill Brown has recently launched The News Project Inc.
“What we’re building is ‘news business in a box,’ Brown says. “It costs too much to launch and operate news sites. We are building a low-cost technology solution to encourage investment and improvement of existing news site technology and approaches to sustainability.” Brown’s new company aims to put in that box “must-have features for professional publishing — plus a suite of carefully integrated tools for monetization, content creation, distribution, customer engagement, and business-critical insights.” Brown says the company will focus primarily on small newsrooms and startups providing in-depth coverage of vertical topics and underserved communities and regions.
Brown says his for-profit company expects to announce its first major customer this week, and others as soon as next month.
Then, there are the stirrings out of the revivified Los Angeles Times.
These are just the latest efforts at contemporary content management and digital platform building.
The industry has seen a lot of consolidation — a common feature of any declining industry — in the traditional content management veterans serving publishers. Newscycle (having absorbed DTI, Atex, Saxotech, and MediaSpan) and CCI are among the dominant traditional players in that business, and they, too, have been making a transition to cloud-based systems while continuing to enable print output. MediaSpectrum works the ad tech space for publishers.
Lee Enterprises’ Town News platform serves numerous local newspapers and TV stations. In the local TV trade, Lakana provides a platform solutions for several chains.
On the pure digital side, Vox Media’s Chorus recently announced a new initiative to license its tech. The Kinja platform, developed by Gawker and then expanded by Gizmodo Media Group, once seemed primed for licensing, but its future is clouded as parent Univision decides on the sale of the websites it powers.
It’s significant that the Post is now talking openly — for the first time — about its wider network ambitions. When I first began writing about Arc three years ago, the Post indicated it had no such plans. If it had done so early on, it might have scared away early customers.
As we approach 2019, though, the world has changed. Daily publishers have lost additional billions in revenue; cost cutting is down to the essentials. Fears of a dominant, Amazon-like Washington Post pale alongside the existential worries consuming newspaper company operators.
Could a bigger Arc play create a mountain of consumer readership behavioral data? More data, well used, can drive more revenue. How much of that revenue would go to the Post; how much to the Post’s Arc customers? Within just that one question, we could enumerate at least a dozen others.
There’s also a more mundane concern. How good will Arc be as a vendor, especially considering Arc’s hyper-growth? Customers I’ve talked with generally give the product good marks — and recent signees have done their reference checks. Some note hiccups in customer service, but nothing much worse than what they’d expect with any vendor. (Even the Post newsroom itself — while appreciating many of the problem-solving tools now built into Arc, including Websked — have grumbled about the time it has taken to create a new frontend system for the creation and editing of content. That system, known as Ellipsis, is now rolling out at the Post.)
Then there’s the eggs-in-one-basket concern. Sure, it’s great to eliminate multiple vendors — but doing so increases reliance on the new single one.
Can the Post, some customers wonder confidentially, continue to ramp up its customer support staff to match the speed at which it is signing up new clients? Prakash, of course, says it can, and that Bezos has provided funding to do that.
Prakash says Arc’s expansion got a thumbs-up recently from Bezos himself when he visited the company 10 days ago. “He thinks that Arc’s story is similar to AWS,” Prakash says.
