Why’d You Do That? How The Times Decides to Send News Alerts

http://www.nytimes.com/2016/08/25/public-editor/whyd-you-do-that-how-the-times-decides-to-send-news-alerts.html

Version 0 of 1.

News alerts and push notifications regularly draw questions from readers. So I sat down with Eric Bishop and Michael Owen from the news desk at The Times to answer some of my questions — and hopefully yours — about how the newsroom approaches this area. Last month, The Times announced it was dividing its email alerts into two channels: Breaking News gives readers developing stories as they happen; and Top Stories gives subscribers breaking news plus special reports, investigative stories or live events in sports or politics if The Times deems them important. Soon, subscribers to mobile push notifications will get the same choice.

Liz: So tell us more about these changes. Why do this?

Michael: One of the big objections people have to alerts is that they’re not “breaking news.” Even though we no longer advertise them as only for breaking news, I think that’s still an expectation people have — that people will only be interrupted for really big stuff. But we’ve discovered that both as a way of amplifying our work and as a way to engage people, and get them into the app, there’s actually a pretty big appetite for things that are not breaking news. We’ve found that many people do want the sort of bread-and-butter New York Times stuff of investigations and major enterprise. So we wanted to be able to send those things but also give them an option if they just want to know when there’s some big disaster going on.

Eric: And so in addition to that, we also have topical channels that users can opt into. That’s Politics, Sports, New York and Business. And we also had an Olympics channel during the Games that users could opt into. On those channels, you’re also going to get a mix of breaking news from those worlds — Politics or Sports — but also get some alerts of our best enterprise coverage.

Liz: These alerts, through email or mobile, seem like they’re becoming a new home page through which a large audience enters into the vast work of The New York Times. I wonder how you try to engage this audience. What does this audience look like from your vantage point: Who are you trying to pull in, and what are you trying to achieve strategically?

Eric: The point you make about this being a new home page is exactly right for a lot of people. I’ve had friends of mine say to me they would not have sought out our app and read this story had we not pushed it to them. But as far as strategy goes, that’s a question we’re beginning to confront a little more aggressively. I think in the past the stories have been the driver of notifications — where, say, we’ve got a great story, so what should the audience be for this? Now we’re starting to think a little more about whether there are audiences that The Times wants to reach but isn’t. Is there a way to reach them through Push that satisfies our goals and also clues them in to content that they might want to read?

Michael: With increasing granularity we can choose the target for a particular story. And it really helps us as journalists to think about, from the moment a story is conceived, who is it for and how is it going to reach them? That’s a sea change to not just be broadcasting it into the dark. The ethos of The Times is never going to be “We’re writing stories for one particular audience.” The generalist blood runs through our veins. But there’s a great power in being able to say we think some people are going to be particularly enthusiastic about a particular story.

Liz: What tools do you have at your disposal now that can help you identify specific users or demographic groups and allow you to watch their patterns? For example, if you watch and see that I open up a lot of politics, or lifestyle stories, are you guys at a place where you can respond to that behavior and dish up more of what I want?

Michael: I think we’re at the very beginning of it. We can segment audiences, and for people who have opted in to particular channels for breaking news, we can target within that. There’s no way for us to say systemically, “So-and-so is going to get more music articles.” Instead, we say, we have this one Arts article we want to send an alert about. Who should we target? We’ll target people who’ve read X music articles.

Liz: And you’re doing that now?

Michael: Yes.

Liz: O.K., interesting. So let’s talk about some questions I get from readers. As the public editor, my predecessors and I receive a regular stream of complaints or questions about alerts. And I’d say they fall into two categories — either people are concerned about an alert that wasn’t sent out, or they’re concerned about an alert that was sent. I imagine it’s a complicated calculation, but can you talk about how you make these evaluations?

Michael: These decisions are made on the news desk, based on each case, for a given notification. Whoever is supervising the desk has the ultimate call, but every potential notification is deliberated on by multiple editors in every case. Sometimes those editors include the masthead [the top editors in the newsroom]. These are the same editors who are involved in what goes on the front page of the newspaper. It’s part of the same overall news judgment and process. As such, it’s susceptible to all of the classic pitfalls of news judgment. Sometimes we’re wrong, sometimes we don’t have all of the information we need to make a decision so we have to put it off, and sometimes there’s a quite a torturous conversation about whether we should send a given alert or not. And there’s no objective measure of what is newsworthy and what our audience is going to be most interested in, so we just have to make a judgment call every time.

Eric: And given you really only have 125 characters, every word counts. So the language is often very heavily deliberated on.

Liz: Along these lines, I wanted to ask you about a question I got this week where someone asked about alerts on terrorist attacks. He noted that there was a suicide attack perpetrated by a 13-year-old that killed 50 people in Turkey: no notification. When there was a priest in his 80s in France, you sent a breaking news alert. And the reader is asking: “Why the disparity? And what efforts are being done to consider balancing unnecessary alarm with equal coverage of terrorism globally?” Do you recall anything about that decision?

Michael: I was there for the priest one. This is such an articulate form of this question. Every time we debate this, it’s like this incredible confluence of so many factors. Like: What is the news temperature of the day, what else is going on? Have we sent other alerts already? Do we anticipate that there is going to be big news later in the day? Are there other big running stories? How does this compare to the most proximate events that we’ve dealt with? And so those are the internal factors, and then: Do our journalists have direct access to what’s going on, can we make informed and reasoned evaluation about it? Sometimes it’s just not clear what’s going on. We’re really cautious, I would say, in breaking news situations exactly because we don’t want to veer into alarm. Those are really the most tense moments on the news desk.

Liz: What about when some other prominent news organization has a big story but you don’t yet have the reporting yourself? You can’t confirm the story. Is that the kind of situation you struggle with?

Michael: Yeah, it drives us crazy. Because in some cases the second you hear it you know that it’s going to be an alert but you don’t have the same level of sourcing that we would have before we would publish something.

Liz: Do you have a number, an average range per day, that you aim for in sending out alerts? A number where you think, more than that is too annoying and less than that is not enough?

Eric: We usually send to the Top Stories channel, I would say, an average of one to three a day. Which does seem to be on the low end.

Liz: Yeah, I would have thought it was more than that.

Michael: If it’s a big news day and there’s a lot of different things happening and we have sent a lot of alerts we are more conservative about sending alerts at a certain point because a) we don’t want to be bugging people but b) we don’t want to be in any way compromising our credibility and seeming to just jump on everything that happens. So it’s an incredibly delicate balance. When there’s a big multiday running story, you almost can’t push too much. We see huge engagement every time we send out an update. I mean, I don’t want to say we have a sixth sense, because it’s not magical — but we’re paying attention to a lot of signals in terms of what we alert and what we don’t.

Eric: And it’s not just per day, it’s often per hour. Did we just send something 10 minutes ago? That might influence a decision about whether to send something.

Liz: So let me ask you one last question. Can you cast into the future a bit and tell us where you think this area — news alerts, push notifications — is going?

Eric: There’s a lot that could happen. For one, I think we’ll see more personalization.

Liz: And what might that look like?

Eric: Well, right now, and Michael touched on this, we are sending notifications to targeted groups, but we’re only doing that a little bit. We could probably be doing that more if we had more sophisticated tools, more editor bandwidth to do this type of thing. We’re still trying to figure out the best way to get reader feedback about these, though so far the feedback we’ve heard has been pretty good. That will inform how aggressive we get with more targeted alerts. I think another thing we aim to explore is the capacity to make these more of a two-way communication channel, and if we can figure that out, then they aren’t just push notifications, but more units of interaction with our audience.

Liz: Hmm. What might someone who subscribes say back to you?

Michael: Historically, The New York Times was this thing we printed and sent out to everybody and was largely a one-way interaction. Now, I think both the great opportunity and the difficulty of messaging is that it gives us a chance to take this one-directional way of interacting and put it in the realm of person-to-person relationships. That’s kind of the dream that messaging promises.

Eric: And in the more immediate future, moving beyond just text, the new Apple iOS will allow photos and videos to be pushed right on the lock screen, so that could be a very rich way of communicating and delivering the news and stories to people.

***

Thanks to Michael and Eric for taking the time to talk. I hope we answered some of your questions. If we didn’t, feel free to write my office or share your comments on this article.