All likes, no clickbait

It’s not often that one of the more sought-after product heads in the world talks about the app he would build if he had a month off. Adam Mosseri, vice president, News Feed at Facebook, was in India recently and spoke to Fortune India about how he keeps users hooked to the world’s largest social network. Mosseri started his journey as a product designer at the Menlo Park, California-based company eight years ago, and these days divides his day between work and his 9-month-old son. Edited excerpts:

How do you decide what products to build?
Some of it is top down—we have an idea that we want to build something because we either think it’s good strategically or we’ve heard a lot from our users that we should build it, or it looks like an opportunity. And a lot of it is also bottom up—ideas come from teams all the time.
I like to not give my team things to build, but instead give them problems to solve. Problems come from a combination of feedback and our mission and strategy. For instance, for reactions, we got a lot of feedback about the dislike button and the super like button; we still get asked about it. We didn’t tell the teams that we want calm reactions, or there’s going to be six types of reactions. Instead we said: Figure out a way to help people express emotions beyond like.

Are there concepts that you keep coming back to while building products?
Other companies build products for individuals. At Facebook, we are always trying to think about communities. Even though your phone is a network device, it’s actually built for you as an individual. Making something for a person is great, but you also need to make sure that if everybody uses that feature, everybody benefits. One reason we are not excited about the dislike button is that although for an individual it might be useful, if everybody dislikes lots and lots of things, there could be a lot of negative energy and people are going to be less happy and won’t share.
Another is building for the next billion users. I also stress on making sure that we avoid blind spots. Most of the product teams are in the U.S. and London, while over 85% of people who use Facebook don’t live in these places. So we need to deeply understand user needs.

Is there a product other than Facebook, WhatsApp, and Instagram that you wish you had built?
I’ve not really thought of that ever! So, going back to lists [he noticed I’d numbered my questions, which led to a conversation about lists], I haven’t found a list app that I love. I’m constantly trying new ones but nothing meets my needs. I used to be a designer and also program, though I’m pretty bad at programming. So if I had a month or two,
I would love to build a to-do app, which is not super exciting, but that would be the first. And then I’d like to take time to figure out what I should really build. But the thing that’s nice about a list or a productivity app is you can do a relatively simple one in a few months.

How do you think media and publishing have evolved in the last decade?
The huge shift to mobile means that the context in which someone reads is different—you’re not reading a newspaper in the morning with tea or while you’re in the bus—and that’s led to unbundling. Instead of buying a bunch of different sections and articles, you’re now reading one article at a time either on the publisher’s app or Facebook or Twitter. More recently, video has grown a lot. All this is shifting as the nature of how people communicate and spend their time is shifting.

Lately, the words content and journalism are used interchangeably...
I like to see it this way: There’s an umbrella of—I like to use the word stories—which could be a photo of my son standing for the first time, or a link from New York Times that my mum has shared or something that you’ve published. Now within that, there are categories, which often overlap. Journalism is one important category and is actually a subset of a larger group of stories that I would call informative. You can also have a story that’s informative and entertaining. Of the umbrella, half are maybe media, be it entertainment or news or sports and the other half is personal, about you, what you might communicate with your cousin, what you might tell your cousin about me and so on. Journalism is part of a larger wealth of things that people want to hear about.

Looking ahead, what are the defining themes for News feed?
News Feed’s mission to connect people to the stories that matter to them remains consistent. I see that more than just video, people want rich and immersive formats—be it 360 degree videos or photos, virtual reality, bigger images, full screen. Also, people are consuming a wider range of content. Facebook started as a place to just share with friends and family; now people read news and watch TV shows, buy and sell. In Mauritius, there’s a group, that’s a third of the population that’s basically a for-sale group, or a Craigslist. Facebook wasn’t designed to be a for-sale group, but people use it that way, so now we’re trying to make the experience better.

What was the thinking behind your recent decision to downmark clickbait in News feed?
It really came from user feedback. We started work two years ago to identify clickbait as a pattern where people would like, click, come back, and unlike. We made changes that reduced it, but the feedback was that people were still seeing too much. We tried to understand what they really meant by clickbait and came up with this: What if we clearly define clickbait, manually mark tens of thousands of stories, and then build a classifier on top of that to further reduce it? The conversations within the team were about: Is that going to work? How do we make sure we are accurate? What if it brings down overall engagement? But we always take the long view. And if people are constantly complaining about something, it’s better to listen to them, because in the long run, that will help them use Facebook more and they will connect more and they will read more publishing content.

Have engagement levels fallen?
No, but clickbait has. A large percentage of that drop happened after we announced and before we rolled it out; it further fell when we rolled it out. We know there are publications built on this model, which is why we proactively communicated since we wanted publishers to understand what we are doing.

What are the defining themes for Facebook in India?
First and foremost, we need to better understand how Facebook works in India, what people here are interested in, and how to best meet their needs. Another big theme is around connectivity. More and more people are coming online so how we can help increase that, how can we make Facebook work better on 2G connections. We’ve started 2G Tuesdays, where for one day, everyone—across all product groups—works on a 2G connection. The third is language and we know we are behind. India is a different market in so many ways, language fragmentation being just one.

Your teams in India, especially on the product side are not very large. Are there plans to ramp up?
At the high level, we’ve said that India is the most important market and we’ve given a team the responsibility for coordinating the efforts across the company. Also, within each product team there is a dedicated sub-team that is focussed on and accountable to that product being good in India. The total number of people is growing significantly and faster than the organisation is growing.