Live Blogging SXSW: Can The Real Time Web Be Realized?
I’m here at SXSW waiting for the Can the Real-Time Web Be Realized? panel to begin. I’ll be liveblogging what happens, so stay tuned.
Search engines have certainly jumped to add in ways to find real-time content. This panel features reps from three of them talking: Google, Microsoft and Collecta. Plus, you’ve got real-time guru Marshall Kirkpatrick in the mix. It won’t all be search, I’m sure, but it’s going to come up.
Our speakers, pictured above, from left to right:
- Marshall Kirkpatrick, ReadWriteWeb
- Jack Moffitt, Collecta
- Brett Slatkin, Google
- Scott Raymond, Gowalla
- Dare Obasanjo, Microsoft
Marshall to panel: What are some of your favorite use case of real time?
Jack: Communication has changed a lot from postal mail to telephone. To internet, closer to zero. More yells from the audience that people can’t hear. We’re getting more real time feedback about the session, heh. Giving latency down to zero gives new applications like instant messaging, Gowalla.
Brett: Future use cases, data interchange is interesting. Amazon has great inventory system to track all they have. For many people, especially in SF, painful to buy because you don’t support your local community. So a barcode scanner that did a feed that connected to local supply chain? You could get scale better than large corporations can do on their own and help local businesses compete better. That’s one of the promises of real time web, turning competition around. Smaller players can compete better.
Scott: Top of his mind are trending spots, tying in geodata with real time web. Where’s the hotspot right now? Then the tide will shift, so how do you track it. [This is what Gowalla does, track hot areas, I guess — I need to try it. Can’t Foursquare and Plancast and Gowalla just all get together?]
Dare: One way is make web pages real time, interesting if you can make parts of pages real time. Good example, hash tag for this conference. Type that in to a search, you get a stream but not pages. Getting more and more pages doing this rather than people would help. And now hard to hear him as we get bad speaker feedback. And I’m like in second row from front. Echo, echo, echo. It’s real time echo, and I can’t hear. Being able to actually analyze the flow of information is also interesting. Lots of patterns about good movies, friends.
Dammit, I wanted more c0ncrete examples. Nice try with the question, though, Marshall.
Marshall: One of his favorite things is something Brett said, that the system is being designed to support unplanned use cases (which if you think about that, is pretty awesome if it really works, how do you plan for things you don’t know?).
Jack: Lucky that first problem is solved, that publishers will give them the data. Second problem, still wrestling with, to figure out what the data formats for real time info will be. If you want to enable the real time web, need a way to push to say there’s data here, so come get it. PubSubHubBubGlub that no one can pronounce is one example of this. It so needs a better name — that’s me thinking that, but I’m not alone.
Scott: Right now, you can send messages from Buzz to Facebook, but you can’t replicate data for easily. I guess share at once to all of them. It’s really, really hard to hear, sorry. I guess everyone in the back can hear now. Maybe I can lip-read. What about corporate data policies that prevent integration from happening, and is it in the best interest of users. Or business deals that give some but not all of info. IE: Facebook, Twitter, Google, the data’s not yours to buy and sell for just your own benefit.
Dare: Common pain we all have today. Too long to get updates to Buzz from other services and so on. We hve to go about getting those services.
Brett: Standards are an issue. If you want to put monetization or business policies into a feed, you’re messing things up. Think that’s what he was saying, sorry. I’ve said it’s hard to hear, right?
Marshall: How about privacy?
Scott: We would love it if everyone was completely open with the world, makes things more valuable. But the user perspective, there’s marginal value in providing all your data. But it’s probably better to stay selfish with your own data. So trick is how to you find the right encouragement to help the business but let users stay in control. There’s a whole lot of work to be done but hasn’t been solved yet. Gowalla trying to give some incentives, like with gaming, so have to be willing to share to win.
Jack: Did we have these issues when phones came out. They decided to be public about numbers (actually, didn’t we all have to be public by default, because of party lines?). Lots of firehoses of data from MySpace or Twitter or Flickr. The level of support these streams have for deleting a post is bad. If you yank something off Flickr, there’s nothing in the Flickr data stream that tells others to delete this post. MySpace has some of this built in.
Brett: Lots of work to do on what do you share, how do you share. Then by addressing these problems together. The “oh shit” problem, oh I didn’t want to post that.
Marshall: Haven’t users already figured out where they want their real time stuff to be private, if they choose what to share or not at places like Facebook?
Jack: MySpace is a really different site than Facebook. In Facebook, you don’t have as much control over the user experience.
Dare: If everyone on Twitter had a private feed, there would be no business model. So Twitter’s interest for people to be less private.
Scott: Similar to Amazon, they’ve never asked if they can analyze my data, but they just do it.
Question From Audience: Lots of Twitter and Facebook bashing. What are your companies doing right?
Marshall: Rephrases, what are your specific examples of good use cases?
Brett: Favorite problem is what’s on Google Buzz right now. Thing doing, encouraging competition in the community. Until now, people like Facebook have owned your data. Buzz is encouraging competition, you can pull in feeds from other sites. Make a rich stream. Then expose it to the world without you knowing. No! He didn’t say that last sentence, you all just thought it. How do you deal with the challenge of different user groups. Today on Google Buzz you can link up sites to your profile (when it works, it’s still pretty lame with me), then next step back, if you host your own blog and that got pulled into Buzz with comments, then my comments there on Buzz, can it go back to your blog. Does that help people regain ownership they may feel about their content, get their comments back (maybe, but kind of good luck with that. it’s like saying you want to pull all the real world non web discussion back into your own living room). We don’t want to own anyone. At end of the day, competition will be the best for all of us.
Scott: Love to see more sites think about what’s the first 10 seconds of users experience.
Dare: Bing Twitter which is interesting way to zoom in on a location and see what’s being said (I think that’s Twitter on Bing Maps). A lot of work went into trying to find trending topics. Other thing interesting at Microsoft, Windows Mobile 7. He has four apps he jumps back and forth to find out what’s going on. I go to Fourquare to Twitter to Facebook. The thinking is that’s its not about going to 5 7 or 10 apps but as a person, I know Rob — here’s what Rob’s doing. Moving away from notion of status update and have people who are your friend get info to you.
Brett: But with Buzz, a criticism was we show ads. So how do you monetize that. Something like people Buzz, if you break down the barriers, lose the monetization for the brands that build the feeds, who’s going to do this. Specifically in the case of ?peoplehub peoplebuzz? dammit it’s so hard to hear.
Dare: I like to start from what’s the best experience for the customer. Doesn’t mean you have to lose the monetization. [Now moved to sit directly in front of speaker, maybe I can hear better that way]
Jack: Something like Twitter is doing, charging for their content, is that more of what you mean.
Dare: There’s no one size fits all. Some companies feel if you get content out as much as possible, that drives content. Hitwise stats show Facebook drives Foursquare the most traffic. Should they pull out unless paid? I don’t think so.
Audience member asks question on privacy:
Scott: We provided some basic degrees of control. If you go more fine grained, it becomes a user experience challenge. It’s a design problem more than anything. We’re working on it, and I definitely believe it is solvable, but I think the trick is to always be explicit with the user and show them where it is going.
Jack: I don’t think users have figured this out. Has friend with 12 year old daughter who had party at house, got busted, forgot her dad follows her. Write about your teacher, but it gets back to her. People have to figure out the potential repercussions, and I don’t think we’ve even figured these out. When our children have these experiences each day, then we can figure out the technology maybe [or it’s not a tech issue but maybe teaching kids you don’t share everything].
Brett: Not just tech issue. Part of what doing to set your preferences overall. Webfinger designed to set common privacy sharing, so you don’t have to remember and say how you want sharing to work everywhere [yeah, good luck with that, when like at Facebook, there’s so much you shared in a granular fashion]. Every provider has different settings and models, and that’s wrong. The user should just say this is what’s important to me.
Jack: I don’t think users will understand the ramifications of those decisions. If I say it’s OK to share my Flickr photos to the world, there are still the ramifications.
Brett: But that’s because your privacy settings are inconsistent. That’ll change. So you get a job, you change those settings. We need consistency first.
Scott: Skeptical, with fine grained issues, touchy feely issues, so many shades as gray. I think it will be as successful as to have cross site identity, like OpenID, which has failed.
Dare: Same, hard to be consistent, that type of solution won’t work. Facebook model is different than Twitter follower model, how do you describe all these things that are different. What I tend to suspect are two things. One, users want privacy controls. Wehn you ask them, they have some expectations. But if you do introduce those, it produces friction. People don’t go into the sharing box and say this tweet is NSFW, so I will share it with only a few people. They just don’t share it. Even if you give them the functionality, it’s too much work. They don’t do it.
Another audience question, didn’t catch it, will dive into answers.
Jack: When you have a ton of info, like this blog post appears first, then this came second, you can dedupe and fight some spam. The real descrepancy is betweent different sites. Each Twitter ID is unique, you can tell. But if you put it on Facebook, harder. And Facebook has internal duplication because not all things open [think that’s what he said].
Brett: Tech solutions trying to do. If Buzz hits Twitter, then Twitter hits Buzz, its gets worse and worse over time. Algos help, but not easy for little guys to do. Using Atom source [think he said] as original content producer can help. Also cross post extenstion from Six Apart to link feeds that are similar but still slightly different, to tell if you posted in multiple places.
Jack: Still going to be a hard thing no matter what. We have to make sure people actually do things.
Marshall: Thanks for joining us!