Nick Bilton, in today’s The New York Times, writes about Facebook’s ability to look at what is happening on its network and predict the future. He says since more than nine million apps and services use Facebook Connect, the Open Graph developer platform, the company can essentially predict what comes next. Maybe!
In reality, what is more likely is that Facebook can turn on or turn off the traffic spigot and hence play the king maker. Bilton quotes an anonymous source as saying, “Facebook is now understanding the type of information it has about what is successful online, or what is a potential threat to Facebook.”
I would take that comment with a pound of salt. Why? Because Facebook is good at hoarding data, but is terrible at interpreting the data: putting a proper context around it and then putting it to use. If you want to know the reason for my skepticism, just look at the lameness/pointlessness of the advertisements that show up next to your stream. For a more clear example of Facebook’s lack of context, just look at the red-hot video apps such as Viddy and Socialcam. On paper they may look the same, but in reality they are as different as chalk and cheddar.
Not all videos are equal
A few months ago, Viddy, a mobile video creation and sharing app, started to use Facebook’s Open Graph to post videos to people’s timelines. For instance, if I took a video clip, edited and shared it, it would show up on Facebook timeline if you viewed the video. If you liked someone’s Viddy video, the action would show up in your news feed (depending on Facebook’s algorithm). If your friend happens to see that item on their news feed on their smartphone, they would in turn get a prompt that gives them the “option” to download the Viddy app.
To be fair, this behavior is no different from how Facebook’s open graph works for other mobile video apps, social news readers and even music-listening apps such as Spotify. The automatic sharing is what made social news readers targets of criticism, and Viddy too has received similar criticism. Nevertheless, it wasn’t long before Viddy started to show up in everyone’s Facebook feeds and in the process IT became one of the hottest apps on Facebook and on the iTunes Store.
The timing for the app couldn’t be more fortuitous. Viddy exploded within days of Facebook announcing its intentions to acquire photo-sharing service Instagram for about a billion dollars in stock and cash. Viddy benefitted from the “Instagram for video meme.”
The hype around the app increased even more when celebrity investors such as Jay-Z, Biz Stone, Shoedazzle founder Brian Lee and Shakira put money in the company. Rumors began circulating that the company was going to get $30 million in funding at a valuation of around $300 million. Sure enough, Khosla Ventures, Battery Ventures and NEA pumped in major dollars into the company.
Viddy wasn’t the only social-mobile video app benefiting from Facebook’s Open Graph largesse. Another “Instagram of video” wannabe, Socialcam, started to zoom up the charts too and in the process was able to raise a ton of money from angel investors that include Yuri Milner, Ashton Kutcher, Brian Chesky of AirBnB and Ari Emmanuel.
I have been suspicious of the sudden growth spurt in video apps. When I saw long-forgotten video platforms such as Dailymotion and Metacafe starting to show up in my news feed and soon after near the top of the Facebook app charts, I knew something was amiss. Was this a repeat of the social news reader phenomenon? When I posted messages on Facebook bemoaning the spammy-behavior of many of video apps, the comments on my feed lit up and many cursed out these apps.
I went back to the video apps and saw that Viddy has fallen of the top of the iOS charts and is now down to #48 on the iTunes store. In comparison, Socialcam is still flying high at relatively #2 spot. Viddy in fact has trended down on Appsdata’s daily active app user base as well, while Socialcam is still at top with about 41.4 million monthly active users. Socialcam’s crazy growth started on April 24th. That made me even more curious. What is Socialcam doing that is keeping its flag flying high versus Viddy, which seems to have returned to a more predictable and normal growth curve?
Michael Seibel, chief executive of Socialcam, said the Facebook changes had catapulted his company to one of the top 100 sites on the Internet. “Since then we’ve seen 10 times growth.” [The New York Times]
Something didn’t add up and it wasn’t till you see Socialcam’s popular and trending videos pages that you understand the full picture. These pages are full of videos of crazy Fox newscasters and other stuff, normally the preserve of YouTube. Many of these videos are actually from YouTube, though I am not certain how Socialcam is putting all those videos on their service. The Next Web’s Drew Olanoff went ahead and did a breakdown of how Socialcam is making it all work.
Once you have a lot of users inside of your app, it’s hard to keep them. One of the ways to do that is to make sure that there’s great content for them to find. For a user generated content app like Socialcam, you’d think that the content would come from its users. Unfortunately, that might not be the case here. It appears that Socialcam is pulling in YouTube videos to its service and passing them off as content uploaded by its users.
The way I stumbled onto this is that to view a Socialcam video in your Facebook Newsfeed, you have to sign up for the app, and it’s safe to say that these non-app users account for some of those 36M. It’s a practice similar to that of social readers used by Washington Post and it’s pretty annoying. I’m also an app user and have social sharing off, so none of my video views or comments end up my timeline. However, when I accessed a video through Socialcam’s site via Facebook, social sharing magically turned back on, thus sharing this in the Newsfeed of my friends.
Olanoff’s assessment is spot on. In addition, Socialcam is buying traffic (and downloads) from the likes of Tapjoy and FreeAppaDay, thus boosting its total app numbers.
Download the @FreeAppADay.com Store App and wish for top rated paid apps to become FREE for a day [Socialcam iTunes page description]
Socialcam isn’t alone. If you look at the top ten video apps on Facebook, they are essentially repackaging videos from somewhere else. What I am amazed with is that YouTube and Facebook would allow such behavior. No word from Facebook just yet, but when I asked YouTube, a spokesperson replied:
While we don’t comment on individual cases, we take any violation of our open API’s Terms of Service seriously and take action against known abusers.
When I asked Michael Seibel, CEO of Socialcam, how can they be showing YouTube videos when they are an app for creating videos, he said that more people watch videos than they create videos. And what they are doing is offering people a chance to see videos even when they don’t have friends on Socialcam or videos from friends. He did say that Socialcam has seen about 3 million videos uploaded to the service and are currently seeing between 1,500 to 2,500 videos being uploaded to the service every hour.
When I argued that their approach doesn’t really add up to be “Instagram of video,” Seibel said that it was a label chosen by the media and not by the company. “We have two clear purposes: help you watch videos created and watch by your friends and also create amazing videos.”
A chink in Facebook’s armor?
This is a nice arbitrage scheme that exposes Facebook’s shortcomings. And it has to do with how Facebook’s Open Graph works.
My understanding is that there is a certain amount of inventory in Facebook’s news feed that is devoted to different categories such as news articles, photos, videos, games and music. They try and create a balance within those categories and depending on the click-throughs, comments, likes and shares a certain item gets, that is surfaced in a person’s feed. Now if there is a video with a header – two hot girls kissing or Lil Wayne is gay — how do you think it will do versus your dog running around on a beach? That magic formula has worked for the likes of Socialcam, but to call it Instagram of video is just wrong. Like I said earlier, Viddy is from Venus and Socialcam is from Mars.
The King Maker
The saga of these very different video apps is being played on a much bigger stage. I bet you I am not the only one who has noticed the “Facebook-the-king-maker theme.” I wondered: had Facebook given a deliberate boost to all video apps just to reinforce its value of its platform in the mind of skittish investors? The idea that they could drive app-downloads across platforms – Facebook, Android, iOS and HTML5 – was definitely a way to blunt any questions around company’s questionable (and lagging) mobile strategy. The launch of Facebook Appstore was yet another brilliant PR move to divert attention away from Facebook’s shortcomings.
The last word on this saga comes from Fred Wilson of Union Square Ventures who writes on his blog:
SEO and Facebook timeline integration is “best practice” on the Internet. You should do both. They can be great free acquisition channels. But they are not great retention channels. Because easy come easy go.
26 thoughts on “Facebook giveth, Facebook taketh: A curious case of video apps”
Anyone else see the irony in YouTube getting their content stolen after that is how they built their site in the first place?
Yup, it isn’t lost on me at all 😉
They aren’t hosting the videos, it’s a simple embed, totally kosher.
Great analysis, this reminds me of Facebook 1.0 when people were throwing around sheep and inviting you to man their police station in Cityville. Facebook being the great gatekeeper (think mobile phone carriers), they can dial popularity up or down at will. I’m stunned that socialcam et al get such monstrous valuations – good video is hard to produce and none of these guys have any revenue model… besides being bought by Facebook!
> If you want to know the reason for my skepticism, just look at the lameness/pointlessness of the advertisements that show up next to your stream.
I think that’s more an ad inventory problem than a testament to their technological ability to target ads.
All right @erickerr, I will give you that. But still the ads are lame.
This is a super insightful article – thanks Om.
Meanwhile the volume on social readers (Guardian, Washington Post, etc.) has dropped so low in the news feed algorithm, I don’t think I’ve seen a “xxx read a story on yyy” piece in days. And I’ve been scrolling.
Same here NewsWhip.
That’s because people want to share what they like not just what they read, the social readers tell everyone I’ve read something before I’ve had chance to do the actual reading.
The biggest current threat to facebook is this autosharing feature they introduced. In my experience, the privacy hurdles over the past few years led to people posting less often and the newsfeed suffered as a result. If their data indicates that same trend on a mass scale then I understand why they introduced the autosharing, but the result is my feed just feels like one of those old email addresses that I hardly check because it’s mostly just junk in there.
Kudos to these apps and sites though for taking advantage of it and essentially broadcasting every pageview they get. I’ll stick to youtube for finding video content.
Great article that points to YouTube as the winner in video. I’m using the iMovie app on my iPhone: great automatic audio capture, easy quick edits and then upload to YouTube (where their new enhancements handle color corrections, etc). Hard to imagine either of these doing a better job.
Hey Om – thanks for the feedback. My name is Mike Vernal, and I lead the engineering team for Facebook Platform. To set context, nothing has changed with news feed. What has changed is that new categories of apps are sharing different types of content on Facebook. What shows up in news feed is based on user behavior and reflects many quality signals, and any updates we make are based on these signals. We aren’t making changes to influence the prospects of a specific app. There are fluctuations and competition any time a new category of apps emerge. Our focus is to make sure the right content is getting in front of people who want to see it. It’s up to the apps to provide the experience long–term that people want to come back to.
Given that you’re seeing some content in your feed that you don’t want to, we still have some more work to do. It’s still early. What you might think is spam, another person might love. We’re trying to strike the right balance for each person, focusing aggressively on cutting content users have reported as spam out of the system.
Hope that helps.
Hope that helps.
Thanks for your response and outlining the process. I have some additional questions and will follow up via your PR team.
I pointed out this on my blog a few days ago, and have a running conversation there with Socialcam’s community manager on Twitter. They are clearly trying to pass these YouTube videos off as their own. The meta description on each video that appears in your Facebook newsfeed is “Video recorded with Socialcam” even though these videos are clearly old recycled YouTube videos. Here’s my full conversation with her: http://www.loganabbott.com/post/22678504530/why-do-some-videos-on-socialcam-appear-to-be-embedded
That is a helpful link and yup, you are right when you point out those differences. I have been following and observing this pretty closely as well.
Great article Om. Seems NEA is smarter than this to invest in a company that is so obviously taking advantage of a viral loop in Facebook that will be short lived. Other companies like Slide/Zynga were built around these viral loops, but at least they were driving revenue. Viddy doesn’t make money, and the content within the app is non desirable from an advertising perspective (too much UGC as well as porn).
Also, Viddy isn’t available on the #1 mobile OS.
great handle @gossipdonkey. made my day.
The thing I don’t get is how investors are silly enough to pour huge money into these sorts of apps based only on quick traffic spikes. Viddy raised $30M from big name investors based primarily on the fact that they were getting a lot of facebook traffic which historically tends not to convert into long term users. Viddy ramped up their expenses quickly and then as soon as the fickle audience leaves for the next-cool-app, they are stuck with extra server capacity and employees that they now need to get rid of. Timeline integration is not the smartest thing to do for the long run in terms of building a company. Start-ups used to see the “Techcrunch” and “GigaOm” effect in terms of quick traffic spikes and I think we are going to see that from Timeline integration but on a much larger scale.
It surprises me how quickly investors forget the fundamentals of investing when they get caught up in the hype. Viddy had to ramp up their expenses by a huge amount because of the Facebook traffic they were getting (both in terms of servers and hiring). Historically, facebook users have been a fickle bunch and although they sign up for an app, they don’t stick around for long. This is troublesome for start-ups and they need to be careful about timeline integration. Yes, your traffic numbers skyrocket very quickly but so do your costs. The traffic numbers will plummet but the costs increased are harder to bring down. This reminds of the “Techcrunch” and “GigaOm” effect many start-ups used to talk about a few years ago when talking about traffic spikes. We are going to see the same thing with Timeline integration but the traffic spike and traffic free-fall will be much larger.
Great article, Om! About OpenGraph and Facebook Connect, yes people are integrating them like crazy. We have seen over 10k integration through our startup LoginRadius.
Thanks for the good article. I also have a lot of interest in Camera apps, and found a new app which has great filter effects: you can download and check it out from the link. recood.com
I expect a post about this kind of non-popular apps as well.
It’s really interesting article and makes me thinking what is really important things for services. From a user’s point of view, I like Recood.com among other mobile video apps: it is only focus on users’ generated videos. However, the big reason why Recood is not famous enough might be they didn’t do what Socialcam and Viddy did?
Great analysis. Some people may be misled into thinking that YouTube content embedded on Socialcam is actually from Socialcam. But I’d characterize this more as Socialcam arbitraging the fact that it leverages Open Graph to integrate with people’s timelines, and YouTube doesn’t. Presumably, YouTube doesn’t do this because its parent company relationship is holding it back. But Facebook’s timeline is patently effective at facilitating viral discovery of YouTube’s video content in ways that YouTube/Google can’t do themselves.