Can the Internet be made more social? This is a question with which Joe Kraus, director of product management at Google, constantly has to grapple. He believes every killer app on the web — instant messaging, e-mail, blogging, photo-sharing — has succeeded because it helps people connect with one another. For Kraus, this means the Internet has an inherently social character, but it can be enhanced further — an area he continues to explore through Google initiatives such as Open Social and Friend Connect. Wharton legal studies professor Kevin Werbach spoke with Kraus recently about the increasing socialization of the Internet. Kraus will speak about social computing at the Supernova conference in San Francisco on June 16.
Werbach: Hi, this is Kevin Werbach, professor at the Wharton school and organizer of the Supernova Conference. I’m speaking today with Joe Kraus, director of product management at Google and truth be told, a former high school mate of mine many years ago. So, good to talk to you Joe.
Kraus: Good talking to you. That was too long ago.
Werbach: It’s kind of scary, isn’t it?
Werbach: So, you’ve done some interesting things with your career. You’re one of the founders of Excite, the early web portal, and then started a company called Jot that was doing wikis and collaboration software, which got bought by Google. And then, to those of us on the outside, it seemed like you kind of disappeared into Google for a while and now you’ve come back out. So, tell us what you’ve been working on.
Kraus: For the last year I’ve been working Google’s social initiatives, which are really kind of in a couple of main buckets. But, probably the largest one is the notion of, “How do you make the whole web social?”
So, the killer apps that have really worked on the web have always been about connecting people to one another. So, whether it is instant messaging and e-mail as communications to connect people to one another, whether it’s photo-sharing as a way to connect people to one another through photos, or blogging as a way to connect people to one another through the words, people have always been social and the killer apps that have really succeeded on the web have always been social.
And really, the primary set of initiatives at Google that I work on are, “How do you make the whole web a more social place?” A lot of that is divided into a couple of buckets. One is this thing called OpenSocial, which we can talk more about. And another is a product which we’ve recently launched called Friend Connect. But, that’s just the general rubric of how do you make the web a more social place.
Kevin: But, it sounds like, as you’ve said, the web is already pretty social. Why do you need something to layer on top of what’s already there?
Kraus: Well, I think, the notion is that if you look at the way social activity happens today, a lot of it is concentrated in social networks — MySpace, Facebook, and depending on what your geography is, there’s usually a dominant social network in your geography. It is a global phenomenon, with hundreds and hundreds of millions of users in these social networks.
What’s interesting about it is that it’s kind of an odd concept at a bigger level, which is that I have to go to a particular site in order to be social with my friends. It reminds me a lot of the early days of user generated content. Before people realized that was a scheme, they thought it was a set of sites. I went to Wikipedia to do user generated content. Or I went to a shopping review site or a product review site to do user generated content.
And then slowly but surely, people recognized that user generated content isn’t a site, it is a concept that can be applied to almost every site. And so today, almost anywhere I go on the web, I have the ability to rate something, leave a comment, or write a review. From the New York Times to Amazon, I have an ability to now contribute to the conversation going on across the web.
I think, social is kind of a similar theme. Today, people think of social as social networks – a set of sites that I go to where I establish relationships with friends. And it’s in the context of those sites that I do stuff with them.
Our view at Google is that’s a transitory phase in the development of the whole social web, and that those friend relationships that you create on these sites should be usable and portable and allow you to get benefit no matter where you go on the web.
You can imagine a scenario where when I go to eBay and I am buying something… Today, eBay uses user generated content as a way to do reputations. So, I can see, “What are the reviews of this seller?” But, it would also be nice to see, “Well, have any of my friends or friends of friends actually reviewed that particular seller?”
Today, I go to a lot of specialty sites. So, I’m in to trail hiking, and I go to sites like protrails.com. It would be nice if I could actually have social interactions with people on those sites without those sites having to become a social network.
So, the idea is, how do you take these relationships that you’ve built in these pilot sites and make them useful across the web. And I think, the transition that we’re going through of social being something that you do in sites to something that you do across the web is very similar (or feels similar at least to me) to the way we looked at user generated content maybe six or seven years ago.
Werbach: That’s probably a good segue way to talk about OpenSocial. Can you explain what exactly that means and how it implements some of the things you’re talking about?
Kraus: Sure. OpenSocial itself is really just a specification. It is a way of allowing developers to write applications which run inside of, typically, social networks.
Now, a little over a year ago, Facebook created a platform to allow third parties to write applications that ran inside Facebook. And the idea was very simple. The idea is that inside of social networks, you want people to be able to do things with their friends.
So, typical interaction was, you go around and you make friends with a bunch of people. And then you need to do things with them. Pretty quickly you can get bored. And pretty quickly, the company itself that runs the social network runs out of an ability to write things (because they don’t have enough people) for their users to do.
And so the idea is, how do you allow third parties, who have new ideas for interactions between friends, to get those applications, those interactions, into a social network? OpenSocial really is a specification that allows for third party developers to write applications to a single set of APIs, and get it distributed in any social network that runs OpenSocial.
The world was shaping up (three, maybe nine months ago) to be a world that would be akin to not having Windows, Mac, and Linux where a developer has to decide, “What operating system do I write my application for? I have to choose among three.” It was shaping up to be a world where almost every social network was going to create their own proprietary platform, and the developer would have to choose among 17 or 18 or 20 different platforms to write for.
It was going to be a very difficult situation for developers, because they couldn’t get the application in as many places as they wanted. And all the social networks had a problem, which is they couldn’t get as many interactions or applications for their users to use, and they would have to fight for developer attention.
OpenSocial is an open community standard that allows for, instead of being 20 fragmented platforms, a single platform that developers can write to and get their application distributed. So, MySpace runs OpenSocial. Hi5 runs OpenSocial. Linkedin will be running OpenSocial several months from now. Friendster. In Europe, folks like Hive.
Social networks around the world, now totaling about 275 million users are running OpenSocial, which really means that developers can write their application once and with very little modification have it running in all these places. So, that’s OpenSocial. It’s a way for third party developers to get their applications running in social networks.
Now, a product we launched in (kind of a preview release) about two and a half to three weeks ago is something called Friend Connect. So, Friend Connect is really the other side of that. It’s more about, “How do you make the whole web social?” Whereas OpenSocial is about, “How do you get distribution for developers on social networks?”
Let me give you an example. You go to IngridMichaelson.com. She’s an independent musician. Probably most people don’t know her music from the radio, but if you watch Grey’s Anatomy or you’ve seen Old Navy commercials you’ve heard her music. And on IngridMichaelson.com, using Friend Connect, Ingrid’s webmaster can run the iLife music sharing application.
What that allows for is now, as a user, I can go to IngridMichaelson.com and instead of their just being static content, because the webmaster has put the iLife application there, I can listen to her music. I can rate and review her music. I can share my commentary with my friends. I can say whether I want to go to her concerts. And all of that information, if I so choose, gets shared with my friends from social networks that I happen to belong to.
So, the user experience… I go to IngridMichaelson.com. I can log in. And then, I can link in my social networks. I can link in my Orchid social network. I can link in my Hi5 social network. And so, suddenly, my friends from those social networks are now with me on IngridMichaelson.com and sharing the experience that I have in terms of seeing what music I like of her’s, what concerts I want to go to.
So, it’s taking those friend relationships that I built in these silos, and making it useful across the web. That’s really what Friend Connect is about.
Werbach: But now, Facebook isn’t part of this? Either of these?
Kraus: That is correct.
Werbach: The cynical view is basically this is Google attempting to route around/open up the thing that Facebook users [inaudible] this business model.
Kraus: Yeah. I mean obviously, I don’t see it that way. I see it as, users have established these relationships on these social networks and they want to use them in more places. People have spent time categorizing and classifying their relationships with other people – be it coworkers, friends, or family. What Friend Connect is really about is allowing those relationships to be more useful in more places.
Google doesn’t store any social graph data on it’s servers. We merely allow third parties like IngridMichaelson.com or portals or bloggers to basically allow their users to use that data to make their sites better.
Werbach: How do you think this plays out? What is determinative in terms of which social platform is the one that ultimately wins out?
Kraus: I really think that openness, over the long term, wins the day. Any platform for running applications that is open and community driven and has flexible and open licenses and a flexible and open development process is going to win. And I think any system that allows a user to use their friend relationships and merge those friend relationships across multiple social networks ultimately wins.
I mean, I think, in the end what this is really about is, users have created value. They’ve put data into these places. They want to be able to successfully use that data in more places across the web. I really think, in the end, this is a question about users having control over how they use the information around who their friends are, and making it available in more places with user control.
So I think, it plays out as — the web will ultimately become more social. Users will be able to, on any site, use friend relationships they’ve established from anywhere. That’s where I think, it goes. It may be a bumpy road to get there, but I do think that’s ultimately where it has to go.
Werbach: Yeah, there’s a lot of subtlety though in this question of what kind of control users what and need, and who can provide it to them. So, I’m curious about your experiences or thoughts in terms of how this evolves. How much granularity and control needs to be pushed out directly to users to set up all these preferences in some sort of independent way verses what gets provided (at least as a default) by some intermediary.
Kraus: Yeah, that’s always the tricky balance. And I have a feeling it will be gotten right most of the time, and wrong some of the time, and then it’ll iterate and be fixed. I mean, this is one these classic kinds of user conversations that back in ’94 we were talking about. We had all these stories about worrying about credit card transactions on the web. And now with the cycle we went through, about a year and half of people being really fearful, and then a common set of practices came about.
Or if you look at the way e-mail usage was. We all use to get jokes forwarded from people as they first came on e-mail. And then kind of a social pattern took hold. I think, it’s a combination of creating a set of patterns about what expectations users have in terms of how much control they want.
My view is that users should have complete control over what they do. Complete control over what information gets shared with whom, and you really do the best you can to put the user in the driver seat. Obviously, there are questions about how to make that an user-interface that everyone can understand. But, I think, if you start with the rubric of “the user should have complete control over who sees what information about them and who sees what information about their friend relationships,” then I think, you’re starting in the right place.
Werbach: Are there things that are not social, or that people don’t want to be social? Or do you think that this is going to be something that pervades every kind of experience that we have on the web?
Kraus: I think, it will be available in every experience that you have on the web. It may not be something that you use.
I think, the truth is there’s this blend. Social networks are usually about people you know and connecting with people you know. Community sites have usually been about connecting with people you might not know, but share an interest with. And I think, what you’re seeing is a spectrum. You’re going to see this merging.
I mean, if you look at IngridMichaelson.com, a lot of the social activity that’s going there is not between people that know one and another and happen to share an interest in Ingrid Michaelson. It’s more on the community end of the spectrum, which is people who share an interest in IngridMichaelson.com who are connecting though her website and creating some kind of relationship with one and another. Fandom, if you will, is kind of the notion of that relationship as opposed to friendship.
But, you also have elements of kind of the typical social experience where of my 300 friends on Hi5, let’s say, a few of them, after seeing that I’ve been at IngridMichaelson, might check it out and also become fans, who also happen to be friends.
So I think the web has kind of a spectrum of experiences. From the individual (just me and the website interacting) though community, which is that strangers end up connecting through mutual fandom of something – be it trail running or Ingrid Michelson – all the way though to what today is the social sense of I have my friends with me as I go around on the web.
So, I think, the truth is that community and social experience will be available nearly everywhere. You don’t have to take advantage of it, just like with user generated content. I use The New York Times, or I use Amazon. I benefit from a lot of that user generated content, even if I don’t myself create it.
Werbach: One of the things I find a lot of people on the outside don’t understand about Google is the notion of “don’t be evil” and they way that Google thinks about business in a different way than most companies. This whole thing we’ve been talking about – the social computing – is one of those areas where I feel like there’s a lot of confusion and some degree of mistrust.
People say, “Google’s getting into this because they want to just do more targeted advertising, invade people’s privacy, and make more money.” I don’t think that’s the core of it, but can you talk from your perspective about how it serves Google’s interests as a company to be building out the kind of social platforms that you’re talking about?
Kraus: Yeah, I understand the notion that a company of Google’s size often gets perceived as having a grand master plan of the way the world is going to work over the next ten years. I can understand how people might project something diabolical.
The truth is far more simple than that. The trust is that Google’s fate is directly tied to how good the web is. As the web goes, so goes Google. If you look at something like Google Gears, which tried to solve the problem of one deficiency of the web, which is it didn’t work offline. And the reason that Google did Google Gears is not because it was some grand plan to make more money as a result of Google Gears, but as a result of the fact that by making the web better and by making it a place where people spend more time, where people store more information, that ultimately that benefits Google indirectly.
The more time people spend online, the more likely they are to do more searching, and the more money that Google makes. So, it is very indirect, but it is also betting on the web as a platform of the next decade.
OpenSocial is like that as well. OpenSocial doesn’t have any monetization component. If you’re an OpenSocial container like MySpace, you don’t have to use AdSense. You can use whatever advertising system you want. You don’t have to have advertising. So, this is really about making the web better – making it a great platform for developers and users – and ultimately, because Google is a big part of people’s web experience, the belief that to better the web means the more time, the more (even if you just boil it down) the more searches people do… That’s ultimately what the benefit is.
We just held Google I/O, which was our annual developer conference. And the whole theme, in all the things we’re doing, we’re really about, “How do we make the web a better place for users and developers?” So, for developers – be it App Engine as a way to allow developers to have an easy development and deployment environment for applications, so you don’t have to build so much from scratch. Or Gears, in terms of making an application both work offline and having a local data store, so you can do more for end users. Or OpenSocial or all the APIs that’s Google’s providing. It’s really about, “How do you make the web a better experience?” And ultimately, we think Google benefits from that.
Werbach: So, just by way of wrapping up… We could talk more about any of these things for a while, and Joe’s going to be speaking at Supernova 2008 Conference coming up June 16-18 in San Francisco. But, let me just ask you one more question for this interview, which is…
You and I have actually both been involved in the web industry from fairly early on, which is now going on close to 15 years – hard to believe. How much do you think that things have really fundamentally changed in terms of the value propositions of the web, and the opportunity? And how much is it primarily that things have scaled up, and the user base now is so huge and so representative of society of a whole, that things that were right on before just couldn’t get the traction?
Kraus: I’m not exactly sure how to field that question, but I’ll try.
You know, there’s a saying in entrepreneurship that being early is the same as being wrong. If you’re an entrepreneur and you create a company and you’re idea’s too early… It doesn’t feel like there’s anybody who wants to buy your product or use your product. It’s hard to raise money. And you can’t really distinguish it between whether your idea was bad or whether your idea was just too early.
I think, there are a lot of examples of ideas that come back around. I know about four years ago, I heard of a couple of groups of VCs that were scanning back issues of Wired magazine to look for ideas in the 1998, 1999, 2000 timeframe that were out there, that were conceived as hot ideas, but ultimately went belly-up as companies because maybe they were too early. So, I thought that was interesting, in terms of scanning back issues to find interesting ideas that were good, but just were too early.
I think that a lot has actually changed in terms of the fundamental assumptions. I believe that if you just look early on… I mentioned the credit card example. Today, people don’t hesitate when they see the little lock icon to feel comfortable putting credit card data into a website. Needing rating and reputation, which is something very unfamiliar in the mid-90s, has become critical to the way people use the web.
Search – I remember in 1994, we would show people a search box to say, “What do you think of this interface?” And people wouldn’t know what to search for. You’d say, “Search for anything.” Eventually, they’d search for their name and see what they saw. Today, we reflexively search for anything. The notion that the web is a place not just for reading but for contributing though the concepts of user generated content…
I really think that a lot of our assumptions about how we use the web go through these cycles every three, four or five years of some new idea – be it user generated content, or putting credit cards on the web, or storing data in the web, or using web-based applications like Google Docs. And they become mainstream over the course of three to four to five years. So, I do think you have these waves that go though, and I think that says a lot about how the web continues to evolve.
Werbach: OK. As I’ve said, we could go on talking about this for a while. Joe, thank you very much for taking the time. And it’ll be exciting for all of us to see how the social web evolves in the years to come.
Kraus: Thanks Kevin.