Companion column: The New York Times Restarts Its New Product Model, In Spanish
If 2015 was a year to pause, reflect and plan, for The New York Times (NYSE:NYT), 2016 is a year of aggressive new product development.
Behind the scenes, though, the Times assembled teams to create new products – products seen as essential to get the company back in the top-line revenue growth mode.
This week, the first of those new products rolled out. New York Times en Español, a Spanish-language Times, debuted Monday morning.
It’s a relatively small investment for the Times, in the $2 million range. It’s a big model, though. The Times plans to launch a second non-English-language product globally before year’s end, built on the same infrastructure and central team as New York Times en Español. The market can also expect one or several new topical products to hit the market this spring.
With New York Times en Español, the Times believes it has cracked a new code for product development, one that’s faster, smarter and cheaper.
NYT en Español will offer readers 10 to 15 translated-into-Spanish Times stories daily, supplemented by Spanish-language reporting and curating of other relevant media out of its new center in Mexico City. There, a staff of six, headed by Times editor Eli Lopez, and made up of Latin American journalists and producers, will run the responsive-for-mobile site. In addition to that team, the Times’ half-dozen Latin American correspondents’ work will be well-represented.
NYT en Español will find a crowded market, as the Times – a global brand – competes with all manner of news outlets. Comscore gauges the Latin American Internet news audience at 120 million, in all languages. MSN and Yahoo/ABC represent the only U.S. brands in the top 10. In December, the Times places 83rd, with 910,000 unique visitors — a fifth to a fifteenth of the top 10 sites. So, consider it near-virgin territory.
Though the Times’ strategic revenue goal here is digital subscription, the Times will not charge readers for at least a year, Stephen Dunbar-Johnson, the Times’ international president, told me Friday.
Instead, it will try to learn everything it can about engaging a new audience, one in which more than 60% of the access will be on smartphones and tablets.
How does it plan to do that? Relying heavily on analytics to guide the project forward, the Times initially aims to reach readers in the middle of the proverbial traffic funnel – not primarily those at the top. Then, in 2017 or later, it will figure out how to charge them.
“Our goal is to skip the mass audience step,” said Lydia Polgreen, the editorial lead on the new team, who also serves as deputy editor for international news. “If we get 50 million people coming to this website, that will be great, but I think our real target is to try and go out there to find who are the most engaged users who we think are going to want to come to use a couple of times a week, or every day, and really have a deeply engaged relationship with the New York Times.”
Finding them isn’t easy. That’s where the numbers people come in.
“By using fairly advanced data science, we can actually get to the right audience and get them engaged,” Paul Walborsky, the business head of the team, told me. “At the end of the day, you can’t just spend money, because you can’t translate everything, and you cannot produce five hundred pieces. You need to narrow-cast, or narrow-target, your content to the right audience.”
The Times says it remade its product development approach to become more entrepreneurial. A year in development, the new international initiative for the Times also aims at employing state-of-the-art research methodology.
Walborsky, best known in publishing for his seven years building tech news site GigaOm as publisher, until he left nine months before the site foundered and was later sold, estimates the new product’s addressable market at 80 million. That’s a group with sufficient income and education – and some digital savvy.
Facebook (NASDAQ:FB) turned out be a center of the team’s pre-launch research.
“We did a lot of A/B testing, largely using Facebook, to develop our process around translation,” Polgreen told me. “And one of the most interesting things that we found was that there’s a real variance in terms of what kind of stories people want to read in English versus, or willing to read in English versus their native language. I think there’s an element of exhibitionism or showing off. Like, ‘Hey, I want to share this story on my Facebook feed because then all my friends will think I read the New York Times in English.’ Stories that are about things that were deeply personal and intimate, people wanted to read and share in their native language. Modern Love columns, anything about culture, sex, romance; the New York Times does cover those things. Parenting. A lot of our softer feature coverage. Things that touch your personal life people tended to read and share in language.”
Watching how readers read the news of the day turned up different results.
“Hard news things tended to show less benefits in terms of translation,” said the 14-year Times veteran. “People were — there was often an almost even split in their willingness to read, which language they were willing to read it in. This is an example of the kind of approach that we took to this project. We asked a lot of open-ended questions and then did a lot of research and testing to try and figure out the answer and let that guide us to the project approach that we ultimately embraced.”
Walborsky emphasized it is mostly about understanding reader needs, rather than starting with the content.
“It’s really understanding what users care for,” he said. “How are they going to interact with the content? What is the content that they want? For that, we went down to these countries and we did panels. We did individual intercepts. Spoke to people on the street.
“In Latin America, in Mexico, we spoke to over three hundred people on the street, asked them different questions. We asked them, ‘How do you interact with news? What are the apps that you have? Have you heard of The New York Times? Is The New York Times relevant to you? What is relevant?’ We’ve asked a lot of these questions that goes beyond just the panel. That’s user-design research.
“From that information, we got a lot of feedback. They would tell us x, y, and z. Then what we would do is say, ‘Let’s test that to see if it’s real or not real.’ Then we would say, ‘They told us that they like love as a column. Let’s test that through Facebook and run some A/B tests'” to see if actually columns about love resonate with that audience.
The takeaway that rings most true: “It’s one thing when people tell you they like. The other one is what they actually do.”
Facebook has been the most important lab. “We embraced Facebook for our research for two reasons,” said Polgreen.
“It tells you two important things about somebody. Number one, they’re on the Internet and they have access to the Internet. Number two, they have joined the world’s largest global community. They’re clearly outward-looking. There are a lot of discussions about this emerging platform in this country. Shouldn’t we be on that?
“’Where should we do our research?’ we asked. We felt that Facebook was a good proxy. If you’re a Facebook user, that’s someone that The New York Times can get access to. That’s a vector for us to connect with you, for us to understand a great deal about you. To put our content in front of you and see how you respond to it. We embraced Facebook as a way that we could do a lot of research.”
The in-person stuff routinized what good journalists should be doing every day.
“It’s kind of a belt and suspenders approach, as Paul described,” said Polgreen. “We did things like go and get your shoes shined and talk to the shoe shine guys about, ‘Okay, first thing in the morning, are people mainly reading news in newspapers, or are they looking at their phones? What are they doing?’ Getting that granular information.
“I spent 10 years as a foreign correspondent, so this type of research, I call it reporting. We did things like go on people’s commutes with them on the Mexico city subway to understand how they use their phones, what they use them for, how do they share information. All of those kinds of insights were super valuable for us. They helped drive our strategy.”
All the research makes sense. But making the insights business actionable requires “validation.”
“When you layer on top of that our Facebook-based A/B testing approach, we’re able to really validate what people were telling us,” Polgreen said. “People said, ‘There’s deep mistrust of the local media. We’re looking for independent news sources.’ We’re able to put things in front of people and say, ‘How do they respond to this story about corruption in Mexico or this piece about the Argentine prosecutor that was killed? Does it resonate?’”
It’s a new circle of knowledge, as Walborsky describes it: “Market research, user-design research, and then we validate with data science.”
Now that New York Times en Español has launched, the analytics work continues every day. There’s a business team of six supporting the half-dozen dedicated Mexico City staff. Among them: a data scientist and a user design researcher, in addition to a product manager and developer.
Lydia Polgreen’s journalistic passion is palpable, as is the sense that such methodology can connect readers to journalists in ways never previously impossible.
“You should mute yourself,” she said. “We’re able to understand and put in front of an audience New York Times journalism and draw insights from their interaction with it in a way that really has never been possible before.”