Rethinking the comments sections of news sites

David Cohn of the Alpha Group explains The Tylt strategy

11 April 2016

Hakeem Muhammad

DavidcohnNews sites and their comments sections have been at the center of many discussions in the past year or so. Some organisations such as Mic and The Verge have taken them off their site completely, others like The Guardian are studying them more in depth.

In this week's interview, we talk with David Cohn, senior director at the Alpha Group, former AJ+, about user comments, how chat apps and dark social are the next big thing and the rise of VR. He also introduces us to Tylt, a project by the Alpha Group that rethinks online comments as part of a newly formed incubator by Advance Digital.


What have you been up to in the past few months, since you left AJ+ and joined Advance Publication?

I left AJ+ about four, five months ago and I've joined Advanced Digital, which is part of Advanced Publication. I actually am part of a new unit inside AD, called the Alpha Group. What we do is in-house incubation. Advanced Digital has a number of properties, they have New Jersey's Star-Ledger, NJ.com, The Oregonian, and our job is to build things completely out-of-the-box from what they are doing, things that the larger company would not produce by itself. We then start executing them and launch MVPs [minimum viable products]. Then we hire people to take over the project and we back away. We never disappear, we're always there for them, but we want them to take over and us to go and do it all over again, start another thing. Not all of these projects will succeed, right, but hopefully some of them will, and actually disrupt the larger company in a way.


The Tylt, the first project you have been working on with the Alpha Group, has just launched and until then was kept very secret, with just one article in Advertising Age. What can you tell us about it?

The Tylt is only our first project, we are hiring for it now. Basically we are seeding it, with an idea and with the core technology and then our hope is that those who we hire learn from that and they see how people respond to the ideas. [The project] may change, they may decide to pivot and we totally want them to.

What is it about?

The Tylt is about trying to figure out how people can have more reasonable discussions and debates online. I'm sure you've heard people talking about how the comments in articles are broken, we're using that as a pain point and saying, “If that doesn't work, what's another way that people might be able to have online discussions?"

It will be very much about social media and social spaces. We are trying to figure out how to leverage social spaces to have more meaningful conversations. If you think about BuzzFeed or AJ+, they do what's called "distributed content": they make content specifically for Facebook or Snapchat or other platforms. They are not competing with Facebook but actually leveraging Facebook, trying to make content that is better [for that specific platform]. We are trying to figure out within Facebook, within Twitter, within Instagram, how do we structure conversation for them so that that conversation can be really rich and meaningful.

More specifically, how does The Tylt work?

The mechanics of it will be pretty simple at first. We will launch just on Twitter and we will use hashtags as our signal. Any article on The Tylt will have two opposing hashtags associated with them. These hashtags will spread on Twitter and people can "vote" for them on Twitter by "favouriting" or retweeting them or creating new tweets with the hashtags. In this sense, we see how these ideas flow and who are the most influential people in spreading them. Every act in social space is essentially a "vote" on The Tylt. After a set period of time (48 hours after launch) we collect the information and create a data visualisation around it. We are leveraging natural behaviour in social spaces (retweeting, favouriting, etc.) and working with them rather than trying to force people into a different space and action. Of course, we would welcome people to participate on our site, but there is still a kind of "soft power" if people engage with our hashtags on other platforms. The team we hire may take things in a different direction and that's all part of being a startup. But this is what the MVP is about.

I guess picking the right two hashtags will be the key...

You are right. I can see the editorial team spending the most time on picking the right hashtag (writing ten different versions and picking one). We want to avoid the idea that one is "positive" and the other "negative" — since most people will gravitate to the positive one and that just makes for an uninteresting poll to begin with. We are looking for "polarising" issues. And part of our challenge will be to give sympathetic arguments to both sides.


Can you tell us more about the other projects that you are working on with the Alpha Group? 

We are looking into something that will enable users to make their own videos quick and easy; kind of like the way people currently make image memes. We want to allow people to take a video and make it their own. We are looking into structured content or structured data, similar to Circa. So what interesting things do we see in structured content? We are interested in messaging and the way in which messaging might be used to distribute news and information as well. I think there are a lot of different directions that we might go in. These are all things we are talking about right now.

Some of the most relevant trends in journalism now are mobile, video and social media. Which innovative ways to combine and decline them have not yet been attempted and should be?

What I've been thinking for a while now is messaging, the role it can play. Most of the things that I've seen in messenger apps right now, I don't think they are that great.  But I do think that, by the end of 2016 or maybe early 2017, we'll see a great maturation in what people call "dark social" and in how people communicate on existing platforms.

I think that the increase in messaging is going to require a maturation of dark social and if you combine that with Slack, the way that it is becoming pervasive, and its user interface, you start to imagine a real mature way in which news and information can be distributed and can be aggregated and curated for people.

Also, if you think about platforms, like Facebook with its Instant Articles, as a publisher it is very good for me because I'll get good numbers. But it’s also bad for my brand because you, as users, only see it as great content on Facebook, you don't associate it with me or my brand.  


You've tackled this issue in your blog recently saying: “What you gain in speed and access to audience, you lose in branding and differentiation”. Is there something that publishers and news organisations can do to maintain their own identity? Is it important to be fast?

It is really important to the users that you be fast. If I am a user, I want this site to be super fast; if it's not fast I'm not happy with my experience. What can publishers do? There is no beautifully perfect answer here, it's a give-and-take because what you can gain in branding or what you can gain in uniqueness, you might lose in speed. Let's say I put up a paywall. With that I might gain a lot, but I've been intentionally shutting out a lot of people. I make it really slow for them to engage with me. They have to literally sign up, they have to take out their wallet, and so on.

I think that every publisher needs to figure out for themselves where they want to be on that continuum. It's not unlike going broad audience or going niche. There is no right or wrong if you want to start a publication about shoes, that could work, maybe there is an audience for that. Maybe it's very niche but it would be so deep that it would work, versus BuzzFeed, which is as broad as you can be.

As far as niche goes, there is a publisher called The Information in San Francisco. They are purely paywall, it is $400 per year to subscribe. It's not cheap, but if they get 100,000 subscribers — most of them being tech companies — that's four million dollars.

I think that, to answer your question, publishers should just be aware of the choices they are making. I'm not going to tell them what's right or wrong. They just need to know what they are choosing so that they're prepared for the consequences and not surprised.


The place where you are working is quite experimental, and this was also true for AJ+, with Al-Jazeera playing the incubator role. Do you think that news organisations should be doing more of these in-house experiments?

Yes, I do think that media companies should be investing in their own disruption. AJ+ is a perfect example of that. It started off as an experiment: they had an idea, they seeded it, they formed something, they hired people like me and other people from San Francisco. We took it over and told them what we wanted to do, and it worked.

And it is as far away from the headquarters as possible, at the opposite side of the world. In the same way, the headquarters of Advanced Digital is Jersey City in New Jersey (and I also think they have some offices in the World Trade Center), and my team, we work in a WeWork, a co-working space in San Francisco: we wanted to be separated. We don't deal with the day-to-day of the rest of the company. It is not about being in service of the existing product: nj.comThe OregonianThe New Yorker. These are all products in a larger company, they cannot come to us and say, "Hey, will you build this for us?" We'd say, "Sorry, we are busy doing our own things." Conversely, if we build something for a news publisher we can't force them to use it. I can pitch it to them, but they can tell me no, and that's totally fair.

I think that when news organisations talk about doing this disruption and having incubators, they really need to buy into this idea of keeping separate the headquarters and the team that is doing different things.


Talking about disruption, what may be the next disruptive trend in journalism in a more distant future?  

I’ve been thinking a lot about self-driving cars. In a future where Uber, Google, Toyota, and Lyft all have self-driving cars, what will be the difference [among them]? They all get me from point A to point B, they all are probably reasonably priced, so I actually think one of the differentiators in self-driving cars would be what content they'd offer. Think about the way shows are being made for Netflix and imagine if a show is made for Uber, and you can only watch it when you are in one of their cars. Or think about the possible role of news organisations in it: I might have to choose between Uber and Google and I'm going to choose Uber because I'm on episode five of the show and I need to see what happens, and I also know The New York Times has an exclusive deal with Uber and I like The New York Times. Funnily enough, that's where branding will actually come back into it: I'm a Wall Street Journal guy, so I go with Google, I'm a New York Times guy I go with Uber. I'm making this up but I don't think that's an insane proposition. I don't see that happening anytime soon, I'm not saying this is going to happen this year. That's in like five years. 

You have talked a lot about platforms. As a journalist, how do you feel about the growing role of platforms for publishers? Is it risky? Will they lose freedom? Will they give the audience only what they want to know rather than what they need?

I don't worry about that last part, because I think that the audiences are the ones to tell us what they want and need, but everything will become the same. Medium is a good example of this. I love Medium but there the writer is minimised and the date of publication is impossible to find. That's on purpose. The reason is because Medium wants you to think or to feel as if it's their content. So if I write something, it's not David Cohn, it's Medium. They are an extreme example but I think that's the biggest downside: who is actually getting the brand awareness? Who is getting the relationship with the audience that is valuable over time? The audience might not even associate with me. They associate with "I saw something cool on Facebook" or "I saw something cool on Medium". Those are the brands that get lifted.

Will platforms end up hiring journalists and content editors? Will they really become publishers themselves?

That's a good question. I think that by most definitions they would, but I think that they would be forced into it; they're like, "Fine, we'll hire some journalists, will start publishing, we'll do it". The only reason they would do it it's because they feel like an obligation so to speak, or it would be a PR move. So, could they become publishers? Absolutely. Do I see that as something they want to do? I don't think so.