The Unified Field Theory of Google

A Techwise Conversation with Google+ designer Joseph Smarr

Loading the podcast player...

Steven Cherry:

Hi, this is Steven Cherry for IEEE Spectrum’s “Techwise Conversations.”

They say you only get one chance at a first impression, but, logically speaking, you only get one chance to make a second impression, too. Google’s earlier forays into social networking haven’t been well received, but Google’s newest is making a very good impression indeed. Analysts, journalists, and the public at large all seem to like Google+. That is, those who can get on—the service is still in an invitation-only mode.

The software isn’t perfect—for example, as my guest last week, Danny Sullivan from Search Engine Land, pointed out, Google’s +1 button—more or less the equivalent of the Facebook “Like” button—doesn’t sync up with Google+.

People are looking for other comparisons to Facebook as well, but surely that misses the point. This isn’t iPhone versus Android and who has the fastest processor or more pixels or a bigger app store. Google and Facebook have very different philosophies.

Google wants you to use the whole Web. The more you do, the more you need its search engine and YouTube and Blogger and Picasa and all the other sites and the ads they show you. Facebook, on the other hand, wants you to use, well, Facebook.

In short, Google wants to play off its superior knowledge of the world and how you fit into it. Facebook wants to play off its superior knowledge of you, and how everyone else fits into your world.

This is the clash we tried to lay out for you in last month’s special report on social networking. That was before even this initial beta release of Google+. Now that we’ve seen the real thing, we have some questions.

And who better to answer them than Joseph Smarr? He’s a software engineer at Google who helped design and code Google+. Previously, he was the chief technology officer for Plaxo, which bills itself as the “smart address book.” He joins us by phone from Google’s campus in Mountain View, California.

Joseph, welcome to the podcast.

Joseph Smarr: Thanks, great to chat with you.

Steven Cherry: Joseph, for Spectrum’s special report on social networking, my producer Ariel Bleicher wrote a story about the start-up social network Diaspora, which she called the “anti-Facebook.” Ariel quoted you as saying that social networks today are “broken and should be fixed.” What’s broken about today’s social networks, and how does Google+ fix them?

Joseph Smarr: Well, you know, social networking has become one of the primary ways that people communicate with one other and stay in touch. In addition to e-mail and calls and things like that, it’s becoming increasingly important. And yet as an industry, it’s still very immature. Partly, it’s just very young, right? There hasn’t been social networking for more than a handful of years, and perhaps not surprisingly, it’s kind of repeating the history of a lot of previous communication mechanisms, which is to say, it starts out with a bunch of small islands that don’t really talk to each other, and then ultimately we find ways through open standards and common protocols to bridge the islands so that people can continue to use and invent new services and still have them all talk to each other. I was sort of joking to Ariel at one point that there’s a reason why people invented the at symbol for e-mail. It really was an invention, because it used to be, you could only e-mail other people on the same server as you, and in the early days of the Internet, that wasn’t so bad. But once there started being lots of people at different computers, it was sort of like, “Well, gee, I’d really like to be able to send an e-mail to somebody else on a different server,” and so they invented the at symbol so you could put the domain name on. We don’t really have the at symbol for social networking yet, right? So if you’re like me, you’re on Facebook, you’re on LinkedIn, you’re on Twitter, there’s a bunch of smaller social networks you’re using, and now Google as well, and if you want to use multiple sites, or if your friends happen to be using different sites, you basically end up having to create yet another profile and account and fill out your photo and set a password and find your friends over and over again everywhere you go. And if you think about all the great innovation that is sure to come over the next years and decades, it just seems to me like that can’t possibly be the way of the future.

And instead, I personally, and a lot of other people, have been involved in trying to figure out what are those underlying building blocks that would let you have an identity that you could use across multiple services, have you build and maintain a set of relationships that grow and change over time, and share back and forth. I mean, people take it for granted that if I use Gmail and you use Yahoo! mail, we can communicate. And indeed, it would seem pretty broken if you couldn’t—but that’s basically where we’re at with social networking right now, and I personally think we can do a lot better.

Steven Cherry: So that sounds like a terrific dream for the long term. But doesn’t Google+ just sort of compound the problem in the short term?

Joseph Smarr: Well, it’s interesting. When I was at Plaxo, we were the sort of small to medium-size player in this industry that was helping people connect up their contacts and stay in touch. And so that’s what originally got me interested in these open standards, ’cause we sort of said, “Well, hey, we want our users to be able to tell us all the places they keep information and just be able to go get it,” and it turned out it wasn’t very easy to do. And so I got involved in a lot of these sort of grassroots efforts, like OpenID, OAuth, Portable Contacts, things like that you may have heard of, to try to help build these. But one of the things we realized is it’s actually pretty hard to change the status quo just from the ground up. And we got some of the bigger companies to participate in some of these technologies, but it became increasingly clear to me that until you had someone like a Google who could come in and be a very strong competitor in the market in its own right but also have the right kind of long-term vision for how the industry should evolve, that it was going to be harder to have the same impact. So, you know, Google+ in a way is sort of on an arc towards that. Right now, yes, it’s another social network that integrates well with Google. But what I hope is that as this sort of helps show people that there is still a lot of room for innovation in the industry and that they’re going to want to either use multiple services themselves or at least stay in touch with people who are on different services, that’s going to then really help sort of lay the groundwork for helping the whole industry evolve in a more productive way towards more open standards and more user-controlled data portability.

Steven Cherry: And so Google+ is built on some open standards, as much as possible?

Joseph Smarr: It is. We haven’t released our platform side of things yet. So right now, it’s sort of mainly just the service, although it’s at least putting a lot more information into your Google account, but you’re already then connected through lots of open standards. So anybody with a Google account can use their Google account as an open ID, and it’s linked to your profile. You can get all your contact information out in Portable Contacts. You can get, at least…for Buzz right now, which has the APIs…all the activities come out, and activity streams and over PubSubHubbub and all that, so really very sort of federated Web–friendly. And I look forward to us trying to do similar things with Google+ as that starts to see more of a platform side.

Steven Cherry: It’s hard not to notice that by contrast, Facebook turned off something called Facebook Friend Exporter, which was a Chrome browser extension that let you export some Facebook contact information. That’s just the sort of separate islands of social networking that you were talking about, I guess.

Joseph Smarr: That’s right. I mean, this is a complicated issue where there’s, I think, some legitimate room for disagreement, but the sort of basic facts are that on Facebook—and they’re certainly not the only one—you can share, say, contact info with your friends, but you can’t necessarily as a user get that info out via APIs. And even if you can, it’s sort of only with the services that that social network chooses to want to work with. And so, it’s—I think—pretty confusing and frustrating for a lot of users to know when they can and can’t and where they can and can’t take the info that they’ve already got and that they put in to share with one another and use it in some other context. And clearly, in this case, you see that whenever there’s a violation of the policy, they’re pretty quick about snuffing it out, even though clearly there must be a lot of demand for those. That’s why people keep building these workarounds. We had our own version that we built at Plaxo a few years ago that would help you import your contact information out of Facebook into your address book, which again a lot of people wanted; this was before it became a common feature on smartphones to do that, and that also got shut down famously, getting Robert Scoble’s account suspended in the process, because he was an early customer for us and had so many Facebook contacts that it hit some rate limits inside of their servers, so it was kind of a humorous version of that same anecdote we’ve seen replayed a million times.

Steven Cherry: So, just stepping back for a minute, and you know, there’s kind of a different feel to Google+, and it hadn’t occurred to me, but maybe it’s due to this attempt to be more open and less of an island. When you log into Facebook, you really feel like you’ve entered, you know, Facebook Land, and similarly for LinkedIn, say. But Google+ feels, I don’t know—just sort of more like an open space. Does that make sense, and was that intentional?

Joseph Smarr: Well, I think so, in the sense that one of the things we really set out to do with Google+ was make it easy to do both public sharing and private sharing from the same space. It’s kind of interesting, I think, that different social networks today tend to really be focused on a particular type of sharing, and they sort of only really work well for that. So Twitter is a great broadcast mechanism, but there’s no—really any good way to have private group conversations, and I think similarly, Facebook is very good for sharing with your friends, but you can’t really have a public presence there unless you make a business page. And that always seemed weird to me, because when I go to post something, I sort of—you think about the real world, you decide when you have something to say, how publicly you want to say it, and we really wanted it to work that way when you share, too. So the share box was one of the things I focused on—you get to type in your message, and you get to choose your audience, which can be public, or it can be a single Circle or multiple Circles, or it can be individuals. And so it really gives you, I think, that flexibility to one day share something in public and the next day share something in private and have that all kind of commingle, and I think that contributes to the feeling of, you know, there’s this clear energy and vibrancy bubbling over going on in Google+ right now. And I think a lot of that is that it’s so easy to share and to share both publicly and privately that you’re just seeing a lot of that energy happening. I think another thing that’s contributing also is just the UI is very real-time, so as people comment and as people post, your Stream just keeps updating, and I think that also creates this feeling of like, “Wow, there’s so much going on here.”

Steven Cherry: Yeah, and yet—there is the Streams, which is very personal and almost Facebook-like element to it, and then there’s Sparks, which is, you know, much more public and sort of world-oriented, and those right now are quite distinct things.

Joseph Smarr: Well, so the idea with those—they don’t have to be. The idea with Sparks was, “If you can tell us some topics you’re interested in, we’re really good at picking up relevant content.” We do it through Google News and through video and a bunch of these different sites. And so the idea was, well, let’s surface some articles that we think you might want to share, and then not only is it interesting because you have something to come and read, but you’re likely to be able to share those, and then you can choose to share publicly or privately. And so some people choose to share Sparks stories publicly. They’re public stories, after all. But I think just as nice is, people…maybe a family member has a medical condition, they might be looking for articles on that, and choosing to share them with the family. Or if you’re an avid video gamer, you might be looking for the latest news and sharing it with your gamer friends. And so I think it actually works really well for these distinguished Circles that people have to get content that’s going to be relevant for that audience. And I think it’s a good example of how a lot of that content might have been something you weren’t going to share on previous social networks because you didn’t have the right ability to just reach the people you wanted. And now with Google+, I hope it will unlock a lot more sharing, and that Sparks will just catalyze it even further.

Steven Cherry: So maybe for me it’s just I’m not really fully set up yet. I haven’t set up the Sparks flow of information for, say, rock climbing, which is my hobby, and I have…

Joseph Smarr: You could type in rock climbing and save it as an interest, and then every time you come back, it’ll show you more interesting articles and videos and whatever on rock climbing, and you can choose whichever ones you want to share, or you can just read them for yourself.

Steven Cherry: And presumably I’ll have a Circle for my rock-climbing friends, and that’s who I’ll be sharing those with.

Joseph Smarr: That’s right. And then all your other friends who just want to hear about technology news and so forth won’t necessarily have to be inundated with all your details about rock climbing.

Steven Cherry: Speaking of the Circles, I’ve seen some incredibly complicated diagrams of what happens when you share something on Google+. How something shows up in the Streams of people in your Circles versus the public at large, and so forth, and then it’s more complicated right now, because…you can invite…you can include somebody in a circle who isn’t even on Google+ yet, and they get an e-mail instead, so it becomes incredibly confusing, and these diagrams are really quite intricate.

Joseph Smarr: It’s funny, a lot of systems that we use every day and don’t even think twice about, if you actually ask somebody to sit down and explain it in closed form, you’d start to hit a mental block, but if you just sort of use it, it’s pretty intuitive. So I think in the case of Circles, you put people in your Circles, and you share them, and they get it. It’s about as simple as that. And what’s happening is you’re basically pushing the content to them. And if they put you in a Circle as well, it shows up right in their stream; and otherwise it shows up in the section we call Incoming, so you can still see the content, you can still comment on it, and it’s a good way to discover people who are sharing with you that you haven’t added back. And we also wanted to make sure that just like e-mail, sometimes you really do want to make sure somebody sees something, and it doesn’t just get lost in the Stream. And so as a sender, now in Google+ you have control to notify a Circle about a post you think is really important. Like, I might share photos, but I might want to make sure I notify my family because they might not check the site as often, but they would really want to see that e-mail from me the way I would have e-mailed the photos in the past. And you can also put people in Circles just with their name and e-mail address, and then you can choose to e-mail them when you share things, because otherwise they wouldn’t be able to see it. And what’s nice about that is rather than having to wait for all your friends to join before you can share with them, you can just start sharing today with exactly who you want, and they’ll still get the content. And then of course if they want to sign up, they can, but they don’t have to sign up to see the information. So we really just try to gear it around what’s going to make it the easiest, most compelling way for you to start sharing, and that really felt like: I choose exactly who I share to, and I can choose if I want to really put it in their face—or just sort of drop it in the stream, and if they get to it, they get to it.

Steven Cherry: It sounds...I agree with you, we do plenty of things in real life that are hard to diagram. My morning commute can get actually pretty complex with the if-thens and the branching, and if I take this train, then I can take that train. I’m curious: How did you actually sort of test the choices that you made when it came to what gets shared with what Circle, where and how, and so forth? I guess the current invitation-only status is a big part of that test, but before that, how did you make these choices, and what data did you have?

Joseph Smarr: Honestly, a lot of it came from my experiences and others’ experiences with how we’ve used communications technologies from e-mail to phone to social networking in the past, and what we felt was limiting about them, or how we could do better. That’s where a lot of the Circles stuff initially came from. Then we refined it with certain things that people tend to do: We did a bunch of usability studies where we had both mock-ups and prototypes, and we brought people in and had them play around. We would ask them questions and see what was intuitive to them, and we would try different versions. We also had a lot of...we do dog-fooding inside of Google: you know, eat your own dog food, where a lot of Google employees use our products before we release them. So we had several big rounds of dog-fooding with Google+ internally and refined our model and our messaging a lot based on that, because you sort of see what people get confused by or which features they actually use. And that definitely helped hone it, especially because a lot of this is in the UI and in the wording. Does it end up feeling confusing, or does it feel natural? And so I think we got the basic model right pretty early on, but there was a lot of tweaking of just the details of what shows up where, and how we describe it to help make it feel more intuitive, and I think that definitely helps a lot. As you say, we’re in a limited field trial now, and a lot of the reason for that is making sure that we can get feedback and work things out before we have to expose everybody to it, because we’d rather learn and iterate as quickly as possible, and then get people in it better. But on the other hand, of course, we’re eager to get more people in soon.

Steven Cherry: Right. And we’ll get back to that in a second, but let me just ask you first about the +1 button. I mentioned this point that Danny Sullivan raised last week in the podcast interview: I think his point was that right now, if I +1 something on a website, nothing in particular happens in Google+, and likewise, if I +1 something in Google+, nothing really happens on the link that I’m +1-ing. Is that right, and what’s going on with that?

Joseph Smarr: I believe that is currently right. The way that +1s work inside Google+ right now is you’re sort of +1-ing somebody’s post or their comment, and I don’t think—although I could be wrong—that if somebody shares a link and you +1 the post about that link, that it’s somehow considered to be equivalent to +1-ing that site itself. But needless to say, we’re moving fast, there’s a bunch of teams working on the stuff, so there’s going to be a lot more happening in that regard in the future. And a lot of what you’re seeing is Google’s deliberate attempt—which I think is the right call—to kind of move fast, move in parallel, ship and iterate, on a bunch of different things. It’s kind of amazing in a way that…the breadth of what’s Google’s trying to do here, where they’re trying to…you know, really, they’re not just trying to release a new service, but they’re trying to socially enable all of Google. And you see that with the common UI elements, now, where you get the share box and notifications across all sites. You see…they’re doing a lot of UI refresh, and over time, you’re going to see a lot more integrations with Circles and other bits of the common social stuff throughout other Google products. But these are all separate teams, separate code bases, separate schedules, and so it is kind of frenetic, mad dash around here to try to be as coordinated as possible but also really let teams execute on their own path. You’re going to see, as a result, lots of things rolling out in stages, and it may not all seem perfectly coherent, but there is some method to the madness, and hopefully a lot of those things will continue to be improved upon and you’ll see new changes to it.

Steven Cherry: And I think that there might be another aspect to that madness as well. I noticed that Blogger, which is this blogging platform that Google bought in 2003, is going to be rebranded into something like Google Blogs, and the Picasa name—Picasa’s a photo-sharing site a little bit like Flickr—that Google bought in 2004, that name’s going to be phased out as well. So is there any sort of “grand unification theory” going on here at Google?

Joseph Smarr: Like with all such things, it’s kind of yes and no. Honestly, I heard first about the Picasa thing through the press. I haven’t talked to anybody internally about it, so I honestly can’t tell you how true that is, or how quickly it’s happening or whatever, but there is certainly a lot of people taking a look at how can we offer a more coherent experience to our users. Google, I think, more than a lot of companies has really had this approach historically of “Let’s just let lots of teams ship products and try stuff out and see what works.” And I think they haven’t wanted to pay the tax that’s inevitable with keeping everything really coordinated and coherent. But that’s a pendulum that can swing in either direction too far, right? So I think if you wanted to make sure that everything was completely in lockstep, you’d really lose a lot of innovation and agility, but on the other extreme, if you look today at the number of Google properties that have some type of profile or some type of sharing ability or some type of notification stream or some type of friending, clearly it’s probably too much compared to what it should be, and you’re not getting enough reuse, and it’s not always clear. And so I think there is a lot of ongoing work in figuring out what the right strategy is. It certainly won’t be that there’s complete convergence. I think there are legitimate reasons why people, for example, often upload YouTube videos under pseudonymous accounts and not their real identity. That’s clearly a legitimate use case. So, sorting through all of that stuff, and how do you migrate, what do you do with the legacy users—that’s a huge ball of wax, but it’s a really important problem. And so—I’m really glad to see a lot of the teams asking these hard questions and in many cases making pretty tough decisions about branding or features or what have you to try and move forward. Because I think ultimately users really will benefit more from that if things make more sense and we’re able to move faster and the work you’ve done will help you in more places.

Steven Cherry: So, is it fair to say that maybe people won’t get all of the benefits of Google+ unless they do sort of unify their various Google-related profiles?

Joseph Smarr: You know, I think it’s going to vary product by product. I think there’s going to be plenty of products you can keep using just fine, and I’m sure there will be other cases where we add features that only work if you’re a Google+ user with your Google+ profile, and it really is kind of calls we have to make on a case-by-case basis, based on who are the users and what are they trying to do and how much complexity is it adding to support multiple ways of doing things. But it’s something that comes up in lots of different debates, especially since we have all these products with millions of users already. And so the question isn’t just, what should the ideal end state be? but, how do we get from here to there?

Steven Cherry: All right. So let me close by asking you where we stand with Google+. It’s still invitation-only, and as you say, you’re collecting data from the beta stage now. Even with the—when people get an invitation, they have to keep checking right now. It’s like looking for a cancellation at a great restaurant or tickets for a Rolling Stones concert. If Google+…

Joseph Smarr: We’re very aware of it, and I really hope we’ll be able to—there’s just a delicate balance of the scalability of the servers and making sure that people are using the product properly and that we’re not seeing confusion or other problems with it and how quick we can open it up. Believe me when I say that we’ve worked on this for a year, and we’re really proud of it and really excited about it. So everyone on the team is supergiddy to get as many people in it as fast as we can. And I certainly hope that you’ll see the pace pick up soon. It’s so just the beginning. There’s a lot that went into building where we are today, but when I look forward—just to get back to what you started asking me about, this social-networking industry in general—what drives me and what drives a lot of the people at Google is this idea that this is still such early days for the social Web. Humans are such exquisitely complicated social creatures. Our relationships are incredibly nuanced. The way we communicate is incredibly nuanced but incredibly important to us. And software today has just been pretty brittle compared to that flexibility. It’s a grand challenge that no one company is up to solving, including us. But the more people who can be innovating, the more ways that we can be trying different interfaces and different graph models, there’s just going to be so much cool innovation to happen, not to mention making so many other Google products and products across the Web socially enabled; and so I think that there’s just a ton to do there. One of the things you’ll notice about the way we did the Circles model is by being asymmetric and by letting you put in people who don’t have Google profiles. We’ve tried to make it “federation-friendly,” you might call it, from day one. So that if you look forward to that world I was telling you about before, where we’re on different services and we’re still able to communicate, we wanted a designer product that, while it doesn’t offer all those features today, it wouldn’t be a square peg in a round hole to offer them tomorrow. That in fact it would seem really natural over time that this becomes one of many places that you can do social stuff on the Web, and that they all can integrate seamlessly. There’s just so much to do, and we want to make sure we get a lot of the core experiences right—but then we really do want to open it up and see what delightful and unexpected things users and developers do with it.

Steven Cherry: So when it comes to the “wide releases”—Hollywood likes to say—are we talking days, weeks, months, years?

Joseph Smarr: You know, I honestly couldn’t tell you. I don’t know myself. It really is something where they’re taking it day by day, they’re looking at the statistics, they’re looking at the back ends, they’re trying to make the right judgment calls. And so I think there’s a sense of urgency here, but there’s also a sense that we want to make sure we do it right by our users. It’ll be hopefully as fast as we can get it, but no faster. Sorry, that’s not a very satisfying answer, but that’s honestly the real answer.

Steven Cherry: Very good. Well, thanks a lot—I appreciate your time.

Joseph Smarr: Cool. It was fun talking to you. See you later.

Steven Cherry: We’ve been speaking with Google software engineer Joseph Smarr about how the company’s new social-networking service, Google+, could lead us out of the walled gardens of Facebook and into a more open social Web. For IEEE Spectrum’s “Techwise Conversations,” I’m Steven Cherry.

This interview was recorded 6 July 2011.
Segment producer: Ariel Bleicher; audio engineer: Francesco Ferorelli
Follow us on Twitter 

NOTE: Transcripts are created for the convenience of our readers and listeners and may not perfectly match their associated interviews and narratives. The authoritative record of IEEE Spectrum's audio programming is the audio version.