Pixable CEO Inaki Berenguer looks out at the explosion of social photo sharing and sees a vision of the 1990s, when the growth of web pages were finally tamed by smart search engines like Google (s goog). He’s now in the business of trying to apply similar intelligence to organizing social photos, helping people make sense of all the pictures that are flying back and forth within their various social networks.
The New York City-based company, which originally built photo-related products and albums, has more recently focused on tackling this problem of consuming photos online and has built out its web site and iOS apps (s aapl) to handle the flow. The site and apps provide a focused way to aggregate, view and interact with photos that have been organized by their relevance to a user. Users can see pictures organized into categories like Top of the Day, Top of the Week, Profile Pictures, Family Updates and others. About a million people are now using Pixable to view their photos and videos, the company said.
The service has until now worked primarily with Facebook, Flickr and Instagram and more recently added support for viewing videos from YouTube and Vimeo. Today, it’s adding support for Twitter, so users will be able to see pictures from Twitter, yFrog, Twitpic, Picplz and other services when their photos are shared through Twitter.
It’s a nice addition, one that becomes even more significant with Apple’s decision to integrate Twitter into iOS 5. So all those photos shared through iOS 5 will also get swept up into Pixable. It’s still a far cry from the 6 billion photos shared each month on Facebook, but it’s still significant considering two percent of tweets on Twitter are photos and a little less than one percent are videos, said Pixable.
But while the news is a nice addition for Pixable, I really like the company because it’s using some powerful intelligence and machine learning to make the whole process of photo viewing smarter. It’s showing how it can make sense of our online social relationships and deliver content to people that’s ranked by their relevance to them. That’s a growing problem on social sites, which feed news, updates and content to users but don’t often apply much deeper logic as to how to surface the best and most interesting stuff to them.
“We’re really trying to make the discovery of photos smarter,” said Berenguer. “Just like search engines did that for web pages, we’re doing it for photos. Everyone can broadcast photos but it’s very difficult to organize and consume photos.”
Pixable takes into account more than 70 signals in trying to sort out what a user wants to see. It will look at whether a picture comes from a person who went to the same school as a user, is in the same city or has appeared in pictures with a user or interacts with a user a lot. It will also look at “likes” and comments to determine if it’s a picture that a user wouldn’t want to miss.
I like this approach because it can really personalize the photo and video viewing process. Instead of sorting through a lot of pictures from people I don’t care about but only appear because they have a lot of comments, Pixable is looking deeper for content that means something to me. It does this automatically but also provides tools for users to augment their experience. So if a user sees a photo from someone they’re not interested in, they can tell Pixable to show them less of that person. And if they like photos from a person they can ask for more. They can also follow a specific person to get alerts about new uploads from that person.
Facebook has some controls but largely surfaces content by what’s getting a lot of activity. It’s a big challenge for Facebook because it’s not just dealing with photos but status updates, links, check-ins and other content. But I think that the work Pixable is doing has the potential to make a lot of social sites more personal. You still have to deal with the question of how to show interesting content that might get filtered out but would still be interesting to users. I think that too can be incorporated into systems like Pixable.
Unfortunately, the new photos from Twitter won’t get the smart ranking from Pixable. The company will just feature it as a separate feed right now, but it’s working to merge it into the regular product as it gains more data with which to score the incoming photos. It’s a challenge for Pixable, one of many for the company. It’s not storing any of the photos, but it’s constantly updating and grabbing new pictures to display from various sources, which puts constant pressure on its servers.
Pixable has some great ideas at work and is showing us what we should expect from social media services. As we get flooded with content from all these people in our networks, it’s going to be even that much more important to find ways to cut through the clutter and surface the things that matter. Pixable is doing it for photos and videos, but I’m hoping its lessons can be applied more broadly soon.