#2 is not true. Stories posted automatically via the API won't show up in friend's newsfeeds. But apps can popup a "feed form" which allows users to preview the content the app wants to post to their wall (this can include an image or embedded swf). If the users posts it then that content will show up in all their friend's newsfeeds. I think this format is actually much better for apps.
It's possible (though not very likely IMO) that that could turn out to be "much better" once app writers adapt their apps to it, and people get used to it. In the meantime, though, the viral aspect of apps was based on publishing actions people take, as a side-effect of people taking those actions, and that just isn't happening now. You do NOT see this stuff on your homepage anymore.
As I pointed out, it's not just apps. Photo tagging, friending, and a host of other Facebook actions that used to be viral aren't anymore.
The implicit news feed stories were never really a strong source of virality. Our current data is showing that the new stream is performing quite a bit better for us, but YMMV.
If someone joins a group, for example, I'm much less likely to see that, whereas before I might've looked at the group based on the title. If someone adds a friend who I know, I probably won't notice now, whereas before I might've seen it and added that friend. These kinds of things are valuable for social networking.
What kind of performance are you measuring, and why does it matter more than other kinds?
I was speaking as a Facebook application developer responding to the comment about "applications can no longer be viral". I have no idea whether it's working out for Facebook as a whole, but for our own purposes it's looking very good across a whole range of metrics, mostly growth (user acquisition and activation).
As a data point, these guys: http://apps.facebook.com/livingsocial managed to get around 2 million uniques over the past couple of days, mostly from Feed virality. Again I have no idea if it's good for Facebook as a whole (and the only people who can say for sure are people with access to that data at Facebook), but for application developers it's been great.
I'm not sure how common it is for apps to be making that much, but I know one of the top apps is pulling in $5,000+ a day, and it's by an individual developer.
I was making around $150 a day from an app with 100,000 monthly active (90% were CPA ads with the rest being socialmedia), and there are a fair amount with over 5 million monthly.
I agree. I totally understand ojbyrne's scepticism, but after a number of conversations in the last couple months I don't have any trouble believing TechCrunch's claim. The question is whether the money is sustainable.
Does it matter if the money is sustainable? I only ask because there is a good chance that the cost to develop the app is much less than 500k. I would take a month of that kind of revenue and parlay it into capital for another venture if it stopped suddenly.
Yeah, I absolutely agree. That's the attitude I see in developers. The apps that make money are often not very complex, so it's worth the effort even if the party doesn't last forever.
There's a lot of secrecy in the industry right now, because most developers see it as a fixed-sum game. Most apps are not technically difficult to duplicate, so saying publicly that your app is making lots of money is asking for others to copy it.
I think the idea of an embeddable widget that flips through all your photos (like the videos you describe that inspired the site) would be really cool. I think if a widget like this is done right you could get some good viral growth from MySpace.
I'd think about capitalizing as much as possible on the timelapse concept and stray away from being just another photo sharing site. Right now, users profiles are just a feed of images, which doesn't seem that compelling..
Sorry, bad wording on my part. I don't mean the reversed phrase should be meaningless, but that it should not mean something different than the unreversed phrase. However, it clearly does mean something different, therefore the implied relationship between smart and drunk is not bidirectional.
That being said, I missed the possibility of a 3rd underlying factor. The title is agnostic about this possibility.
Just a guess, but they may feel that having the beta label on a downloadable app tends to turn people off, whereas people don't mind beta web apps (its the norm and they won't screw up your system).
Maybe that person just got unlucky and had a bad app tester.