What Would Google Do_ - Jeff Jarvis [47]
Not every business and institution has the blessing of Google’s and About.com’s data. Sometimes, of course, it’s better to listen to people one-on-one, as Starbucks and Dell are doing with their give-us-your-ideas platforms and as countless companies do when they read blogs and forums. These methods beat focus groups and surveys, which pick people at random who may have nothing to say. It’s better to listen to the people who have a reason to talk with you. Procter & Gamble chairman and CEO A.G. Lafley said in Strategy + Business magazine that he wanted customers to be “valued not just for their money, but as a rich source of information and direction.”
Sometimes, listening itself becomes your product. Flickr listens well. The photo service founded by Caterina Fake and Stewart Butterfield and now owned by Yahoo created an incredible infrastructure to take in more than a million photos a day and enable users to organize them around captions and tags—one-word descriptions—which also enable fellow users to find them (and each other). This is all made possible, as we discussed earlier, because of Flickr’s decision to make photos public by default.
Flickr brings out not just the wisdom of the crowd but also the aesthetic of the crowd and displays that for all of us to see. Go to flickr.com/explore/interesting/7days/and press the reload button a few times or click the link there that enables you to view these images as a slideshow. I predict you won’t be able to stop. It’s mesmerizing. These are the photos Flickr has determined are interesting. How did they do that? By ranking popularity? No, that would likely lead to lots of pictures of thin young people who look good wearing very little on beaches—or, worse, to pictures of cute cats. Does Flickr do this with an army of editors? That would be the reflex of old media. But that would not scale, as they say in Silicon Valley; it would take a nation of editors to sift through the 3,000 pictures that come into Flickr every minute.
How does Flickr find interesting photos? Well, of course, they don’t. We do. As Butterfield and Fake explained it to me, Flickr determines “interestingness” in a few ways. The first and most obvious component: Flickr measures the interactions—commenting, emailing, tagging, linking—that occur around a photo. Second, they map all these actions to see which users turn out to be hubs of activity. These people are presumed to be influencers and their actions are given extra weight because the Flickr community must trust them—a logic not unlike that used by Google’s PageRank. Third, Flickr performs a reverse social analysis: If Bob and Sally are emailing and commenting on each others’ photos all the time, the system presumes they are relatives or friends; they have a social relationship built on familiarity. But if out of nowhere, Bob interacts with Jim’s picture, the system then presumes that their relationship is based on the photo, not on life. The interestingness algorithm devalues Bob and Sally’s social relationship and gives greater value to Bob and Jim’s interaction around a photo. It’s counterintuitive but sensible when you think about it.
Flickr ends up with a never-ending stream of interesting photos. Granted, being interesting is not as hard a test to pass as, say, relevance on Amazon or accuracy on Google. Still, look at Flickr’s gallery. I’ll bet you’ll agree that almost all the choices are, indeed, interesting. Flickr is algorithmically aggregating the aesthetic of the crowd. Out of that comes a better service for every user, more opportunities to build traffic and revenue, a rich relationship of trust among those users and Flickr, and even new products. All from just listening.
New Ethic
Make mistakes well
Life is a beta
Be honest
Be transparent
Collaborate
Don’t be evil
Make mistakes well
We are ashamed to make mistakes—as well we should be, yes? It’s our job to get things right, right? So when we make mistakes, our instinct is to shrink into a ball