First, the backstory: YouTube Instant wasn't built in-house. It's a side-project written by a Stanford undergrad and former Facebook intern named Feross Aboukhadijeh. (Aboukhadijeh has just been hired by YouTube thanks to his work on YouTube Instant, even though he's still in college.)
The utility of YouTube Instant has been immediately apparent not only to Google but to a huge number of users. As Aboukhadijeh told TechCrunch:
YouTube Instant hit one million visitors on Monday, just 10 days since I launched it! The site also has 28,000 Facebook Likes, 20,000 Tweets, and 14,000 Stumbles â€" quite impressive numbers for only two weeks since launch! (I've been busy with the first week of classes at Stanford, so I didn't announce this until now.)This is exactly why companies document their APIs: so outside developers can think of nifty ways to built on top of a tool like YouTube, which is so powerful and polymorphic that Google's own engineers could never hope to think of all its possible uses in-house.
YouTube Instant has been on the homepage of CNN, Fox News, VentureBeat, CNET, Hacker News, All Things D, Engadget, and PC Magazine. The site went viral on a scale that I never could have predicted.
Also noteworthy: I just updated the site with some awesome new features: autoplay, keyboard controls, shareable URLs, and a sexy new design. And I'm not going to stop there. I'm currently working on adding awesome Facebook integration, on-the-fly playlist creation, and custom auto-updating AdSense units.
But not every third-party tool gets acquired by the platform. Twitter just acquired a one-man developer named Loren Brichter, whose iPhone app, Tweetie, won an Apple Design Award and accrued several million users and majority marketshare on the iOS. In short: it takes a lot of evidence of success and talent for the platform to buy a third-party developer.
So why is Google buying Aboukhadijeh? Perhaps because he is thinking like a music fan. Music is YouTube's least-discussed strength, and with YouTube Instant in place, the site has become a perfect song-finding engine. YouTube's database of videos includes such an incredible plethora of songs -- album versions, remixes, personal renditions, and so on -- that it dwarfs the selection on Apple's (AAPL) iTunes, Rhapsody and Grooveshark (although Grooveshark does have some alternate versions of songs). In short, if I want to find a chopped 'n screwed version of Miley Cyrus' "Party in the USA," YouTube is where I go. And YouTube Instant makes it frictionless and fast.
YouTube apparently hasn't built this kind of thing internally, or perhaps hasn't coded it as elegantly as Aboukhadijeh has. Despite Google's surplus of engineering talent, that's not surprising; YouTube doesn't have much incentive to pursue a music strategy when video advertising is so valuable (and they have enough on their plates figuring out video monetization strategies.)
It's unclear whether YouTube will use Aboukhadijeh's software as a kick-start for some new project, or whether they just think he's a capable programmer with potential worth exploring. But in one fell swoop, he has helped YouTube rise head-and-shoulder above other song-finding tools, making it an instant winner for users. Whether or not he can make Google any money remains to be seen.