How we used design research to launch The New York Times en Español

I’m on a plane to Mexico City, a place that I got to know last year, and a place that I now love. And I’m thrilled that as of today the people who told me last year “I’d really love to read The Times in Spanish” now can.

A year ago I led a design research field project to explore how The Times could best share its journalism with the Spanish-speaking world. What we learned from that in-depth research — as well as from subsequent exploration — led directly to today’s launch of The New York Times en Español.

Ken Doctor has a great writeup of our product approach: I’d love to think that we have indeed “cracked a new code for product development, one that’s faster, smarter and cheaper.” In essence, we are using coordinated research efforts to understand our new readers, and build the best product and editorial experience possible.

Our field work in Mexico City last year laid the foundation for the work that followed. The time “in the field” was a chance to learn directly from news readers and generate insights that would support our product and editorial strategies. We also followed our field research with ongoing, iterative quantitative research. From my perspective as a design researcher, this has been a remarkable opportunity to incorporate ethnographic research methods and data science to uncover rich and actionable insights.

Here’s an overview of that process.

We learned in-context.

When our team arrived in Mexico City last year, we wanted to get situated and quickly learn as much as we could. The first few days we were on the ground, we split up and conducted short intercept interviews where we actually stopped people on the street as they went about their daily lives to ask them what they were reading, what news sources they trusted, and what they were doing with their phones (mostly WhatsApp, as it turns out). The intercepts were a great way to get the team acclimated to the city and to help everyone feel involved in the research process. We also quickly picked up on some themes that helped inform the more in-depth research that followed, such as people reporting triangulating between multiple news sources and the importance of radio.

We went deep.

We next conducted a series of one-on-one interviews, meeting with research participants in their homes, their offices, and even joining for their commutes. During these focused conversations we were able to explore what people were doing, loving, and needing when it came to news. We also shared initial prototypes for what would eventually become our new site, and gained helpful feedback that we used to quickly iterate — in the field — and refine our product approach as we went.

As researchers, we are always on the lookout for what we call “latent needs,” meaning something that could make someone’s life better, but which they might not flat-out ask for. One-on-one, contextual interviews are useful for helping us uncover these needs. During the first days of interviewing, we heard several participants say something along the line of: “In Mexico the news is only about things that worry you.” We took that to heart and started exploring what that meant to them. We learned that our participants were hungry for news that celebrated culture, that provided a sense of escape and enrichment, and that would be “fun for the weekend.” In response, we prototyped a sort of Sunday magazine feature, with stories about travel, home, love and culture, and shared that with the next participants. People loved it. That prototype eventually became what we are calling “Reposado.”

Alongside the one-on-one conversations, we also hosted a series of group interview sessions, where we heard from an array of news readers about their daily lives, their jobs, their relationships, and how news and media plays a part. During these sessions we shared the evolved prototypes and then continued — daily — to refine our approach.

We talked to experts.

We wanted to know more about the overall media landscape to understand how The Times did, and could, play a part. We met with journalism professors, reporters, broadcasters, and executives; they gave us invaluable perspective. In particular, they helped us learn about big picture changes to the news ecosystem, and how their roles were changing.

We collaborated.

Cross-cultural research works best with support from a local partner. We collaborated with a stellar Mexico-based research team to help us with session moderation, simultaneous interview translation, and cultural translation. More than just an add-on, this local team was fully integrated into our process from the very beginning, ensuring that our insights truly spoke to reality. I trusted that we really “got it” because of the informed support of our local team.

We used data in smart ways.

Nothing beats qualitative research when it comes to certain types of understanding. Emotion and intention, for example, really come through during personal interaction. But for understanding behavior on a broad scale, data sure comes in useful. Over a period of several months we experimented with posting articles to Facebook in Spanish and our team’s data scientist analyzed what people were reading and sharing to understand more about our Spanish-speaking audience. This research into behaviors helped us know what kinds of articles people love, and helps our editorial team make choices about how to curate our selection of articles to translate.(And the types of stories that we now collect in the Reposado section performed particularly well.)

We didn’t take ourselves too seriously. (An anecdote on observation and data.)

One thing we noticed during our field work: there are so many dogs in Mexico City. We saw huge numbers of really fancy, purebred dogs. And we thought — huh! People really like dogs here! It would never have been something that we could have understood from data, because we never would have known to look.

So when The Times’ quiz about the Westminster dog show was posted on the main New York Times site, we translated it to Spanish and shared it as a promoted post on our Facebook page and got a ton of traffic. People loved it. Our data scientist reported that Facebook’s algorithm gave this post a very high “relevance rating” for its targeted audience based on the positive feedback it received. While our editorial strategy is not dogs-only, this was a fun validation for us about the power of contextual observation + data science.

What’s next…

We have only started. Now that we have a validated product and editorial approach, we have a dedicated team in Mexico City staffed with talented journalists from across Latin America doing original reporting in Spanish as well as continuing to translate selected articles from our primary news report. We are continuing to learn from our readers and refine our approach and we look forward to building new relationships with new readers in a new (to us) market.