November 2, 2016
Over the years I’ve started a few different social video platforms: Justin.tv, Socialcam, Twitch. Twitch became a success; the other platforms, not as much. What went wrong with JTV and Socialcam and what went right with Twitch?
At Justin.tv, we didn’t ever talk to our users. We didn’t understand why they wanted to broadcast live video or what their goals were. We actually stumbled on lots of things that could potentially have worked as a business (for example, social broadcasting, where people chatted with an audience via webcam — something that looks a lot like what YouNow is today). But, we never focused in on any of those use cases and built a big community.
At Socialcam back in 2011, we tried to build something that normal people would use to share video from their mobile phones. The biggest problem: most people don’t want to create video: normal people’s videos don’t look good, they don’t have anything exciting to capture, and it is many times harder to frame and shoot a video than it is to take a decent photo. We never identified a specific group of users who really wanted to create short-form mobile video with the product. Eventually, Socialcam was acquired by Autodesk for $60mm.
Twitch was our success case. What did we do differently? Focus on building a community around a narrow set of content creators, in this case, video game broadcasters. The single-handedly most important question my cofounder Emmett Shear asked the team was “who are our users?” In this case, Emmett decided our most important users were our broadcasters, and we subsequently spent a lot of time asking these users about their goals, needs and hopes, so that we could generate ideas for what to build. It turns out many of these broadcasters were concerned with our video quality (so we worked on stability and transcoding) and wanted to pursue a career in broadcasting full time (so we helped them make money).
At the time, the common perception in the tech industry was that Youtube had an unbreakable network effect among all user generated video. But, as it turned out, while Youtube is really good for short form video, it wasn’t perfect for gaming live streams. By focusing on a specific, new format of content, we could add features that Youtube couldn’t or wouldn’t: commercial breaks controlled by the broadcaster, live chat moderation, special emoticons, broadcaster subscriptions. Eventually, Twitch broke out, built its own community with its own strong network effect, and we sold it to Amazon for $970mm two years ago.
But despite having built three video platforms, it was only last year, by accident, that I became finely attuned to the life of a video creator — and it wasn’t even on a platform I controlled. I started using Snapchat as a way to pass the time while doing cardio on my exercise bike. Soliciting questions from the audience was (and remains) an easy way to come up with content ideas. By recording answers to aspiring entrepreneur’s questions on building product, starting companies, and general life advice, I felt like I was giving back — all while getting a good workout.
So what have I learned from all of this, and how does it apply to what we are doing at Whale?
First of all, I believe that there are many types of video formats out there yet to be discovered. Video is medium, like text. There are lots of compelling forms of text content out there: books, short stories, blog posts, Quora answers, tweets. Fifteen years ago, in the offline world, video meant either movies or TV. When video came online, at first all videos were lumped together on Youtube. A few years later you had livestreams (Justin.tv, Ustream, Livestream) which were mostly a mix of celebrity chats, events, concerts and sports. In the past couple years, there have been an explosion of video content formats:
- Disposable daily Stories on Snapchat and Instagram. These are low pressure (because they disappear) and almost like a video status update. They are constrained to ten seconds to make them easy to consume in a low attention span world.
- Video game livestreams on Twitch. These are long form, almost a kind of talk radio, where viewers’ experience isn’t diminished by starting the stream at any random point in the middle. In fact, almost all users don’t watch an entire stream. You can start from any point (whenever you happen to tune in), and watch for as long as you’d like (for 30 seconds while you wait for the bus to hours while you work on something on your laptop).
- Facebook auto-play videos. These are designed to grab your attention as your scroll past and they begin autoplaying while muted. These videos often have text captions, are visually very interesting, and immediately start with an interesting subject (no time for build up since you might just scroll past!).
Our hypothesis is that short form video answers are a new content format for video.As I became a video influencer on Snapchat, I realized that one of the most common interactions from the audience would be to send me questions. Talking to other influencers, this was very common; almost every influencer on Youtube, Snapchat and Instagram will get asked questions every day, whether they post about fitness, business, makeup, or just make comedy videos. I shouldn’t have been surprised — this is a common behavior on Twitch as well.
With my small audience size, some days I would get upwards of 200 questions, but some bigger influencers estimated that they get pinged 10,000 times a week. That volume makes it impossible to filter through and meaningfully interact with your audience. This is one of the reasons we decided to allow influencers to charge to answer questions: to allow them to filter down to the most meaningful or important questions.
The Q&A format also solves some of the problem we had with Socialcam of “I don’t know what to create.” Many people are experts in something: their profession, their passion, the experiences they have lived. If you look at popular reddit AMAs, many of them are simply regular people that answering questions about an uncommon profession. By making the content format a response to a questions, the creator gets a prompt for what to record (and at least one built in audience member!), and the viewer gets an upfront indicator of the quality and subject of the content.
We also think that not everyone wants to create video, and that’s ok. Everyone can share personal photos on Instagram or status posts on Facebook. But video is hard to create, and not everyone is comfortable recording themselves. When I started Justin.tv (the very first iteration was a 24/7 live stream of myself), people thought I was insane to want to record myself on video. Apps like Snapchat and Musically are changing that attitude slowly, and more and more people are getting used to filming themselves. At first, not everyone will be comfortable answering questions on Whale, but the number of people who will be is growing over time.
Our ultimate vision for Whale might be described as Quora for the Snapchat generation. It’s a community about exchanging information, where people share their thoughts more permanently than Snapchat, and where the audience is divided around interests. As selfie video has become part of the cultural zeitgeist, the perceived cost of production of video has actually become less than that of text: it’s easier to take a short video of yourself than sit down and write a blog post. In the future, the preferred way to answer a question might be by video.
In the end, Whale is an experiment. Here are some of the open questions we have:
- Do people like answering questions on video?
- How does it feel to have an influencer answer your question?
- Is it worth paying for?
- Can we get someone to make their first ever video?
- Can video answers be better than answers through text?
We hope you’ll join us on the journey to find out the answers. Check out Whale and give us any and all feedback on the app. You can ask me any question here.
Want to join the Whale team? We’re always on the lookout for great engineers: hit us up at jobs@justinkan.com
Curious about what goes on behind the scenes? Leo learned how to edit video so he could start a vlog about what goes on. Here’s the first episode.
Why Whale? was originally published in Justin Kan on Medium, where people are continuing the conversation by highlighting and responding to this story.