Add to iTunesSalim Ismail - Head of Yahoo! Brickhouse
Introduction
This week I sit down with Salim Ismail, the head of Yahoo’s Brickhouse. In this role, he is exposed to hundreds of proposals for Yahoo employees each month for products to be developed. Therefore, it was very interesting to spend some time getting his views on the future of the web.
Links
Picture of Yahoo! Brickhouse
As promised on the podcast, here is a picture of the Yahoo! Brickhouse. As you can see, the space is very cool and located right in downtown San Francisco. However, the environment resulted in some tough background noise - so I apologize for the audio quality. I hope you listen though, because I think Salim has some great insights to share.
Transcript
| 1:00 | Sean Ammirati: I have decided that Salim probably has one of the best jobs in Silicon Valley.I am going to let him tell you a little bit about what he does at Yahoo and then jump into an interesting discussion that I want to share with the listeners of ReadWrite Talk around what’s coming next in media space. Salim, if you can start just by giving us a little bit of your background and what you do here at Yahoo?Salim Ismail: Ok. I have been in the startup world for about seven years. I’m best known for having founded and worked under PubSub and did Confabb which launched last year and I was working on three other startups. Way back with my technology background, I also spent five years restructuring large companies as a managing consultant.About two years ago, I was on a panel with Bradley Horowitz and Marissa Mayer from Google. I talked a bit to Bradley and again, I was on the panel a year and a half ago. And when I was leading these startups last year, he came up and said, “Hey, why don’t you come and do this at Yahoo?” |
| 02:10 | I said, “No, no, that’s horrible. I can’t possibly do innovation property.” To do it right, you have to be offsite, and you have to be off-brand. Then he said, “Yes, we can.” So we started talking. Bradley is an amazing guy and I spent a little while investigating. The company was really serious about a project like this. And then, I jumped on board. That’s my background.Basically, what we do is the role here is to find cool ideas around Yahoo and get them built and launched. And there’s an unbelievable amount of innovation around Yahoo. I mean there are 1,200 more hacks that have come from the Hack Days. There are thousands of ideas that come from elsewhere.There are all sorts of amazing technology. We have people like Robin working there. We have all sorts of people, extraordinary mentors and technology visionaries. The problem is that Yahoo! is geared towards building and managing most of the products like Mail and Messenger and so on. |
| 03:15 | And there’s not an easy path-to-market for small ideas. And you can think of Brickhouse as building a good market channel for building small, interesting ideas. So my job is to figure out what are the cool, interesting ideas that would really make it and be different for Yahoo, and launch them.And the prototype project for which Brickhouse is based on is Pipes. So Pipes is… it took Pasha about 8-10 months to get buy-in and an investment to get and build a team. Actually, then another prototype was built to launch it. But with four guys off at a corner, look at what they did.How could we reduce that 8 month cycle down to 8 weeks? And so, we have a team of a dozen interesting developers who are from all walks of the industry. And people submit ideas, where we find them and ask them to come in. |
| 04:06 | We’ve now managed, categorized and analyzed the entire backlog of ideas at Yahoo. And we’re building up what we think is the top three or four while building up the team. And it puts in place to do the easy to replicate ones.Sean Ammirati: Great. This next question, obviously you would have about as much perspective on as anybody. You started to share yesterday and I wanted to capture this for our audience. What do you think is coming next after social networks?Salim Ismail: Let me go 50,000 feet from there. There are two fundamental rules for information processing. First is: do you care about a particular piece of information? Second is: what do you do about it? Before that, you have to find the information. |
| 04:59 | I see three phases in information from the Internet. First is information gets published or discovered whether it is a webpage or a blog post, etc. The next phase is syndication or aggregation of that information by a Yahoo! or a Google or any of the companies which will take your idea, perhaps it’ll do this or not and so on. The third is reading, processing and consumption of that piece of information.Information flows in that way and we used to publish html pages that got aggregated by search engines and the web browsers. In the Web 2.0 world, we update RSS feeds or update blogs that send out pings. They get syndicated. Other people aggregate these things through an aggregation engine like Technorati or an RSS reader, and we go through that.And that is the underlying framework for that information flow. You got layers of how this happens. We got fairly excited about email in the 80’s and all of the things that come from that. In the 90’s, we got incredibly excited about the Web, etc. |
| 06:06 | Now we are getting into the next phase. There is so much information out there that we need ways to filter it. We, largely as an industry, solved the discovery and the aggregation problems. Now we are looking at, “Okay, there’s all these crap out there. How do we filter it effectively?”There are a dozen other companies, if you think about most of the companies in the space today, they are thinking how do you filter and make sense of all the information. I think in the next few years, the industry is focused on that. For social networks, it is using your friends is an obvious way of filtering information. That’s where you first have to form the connection between your friends and form the networks, and then you can leverage that network to do the filtering for you.People are hugely a buzz about social networks. I think it’s a fad. It’s another thing that over time will become part of the infrastructure. Just like mail used to be a really big deal in the 80’s. So you have to have a dedicated line and a dedicated server and as that got built into the Internet, it became invisible. |
| 07:08 | It’s the same thing with request response and the Web in general. We used to think it was a big deal. Now it’s kind of pretty standard. I think social networks would eventually get built into the network. People like Facebook and Myspace are trying to be the network but that never succeeds over time. The market will never bear a closed environment over time. It’ll bear it for a while but then, over time we’ll all be doing that.What I see coming after social networks is when people first get on the Internet, they start communicating. Then they establish a presence in which we can do email. We all had our homepages in Geocities. Then they start transacting. As we go to the Web 2.0 world, we first have our walls for our Myspace profile, etc. |
| 08:00 | I see bloggers as a disappearing genre. I think bloggers are stratifying into a few thousand self-publishing journalism like yourselves. And then everybody else is a Facebook user or a Myspace user, which is essentially blogging. We just don’t call it that. It is structured publishing in some way or the other.I think there are two themes that will dominate for the next five or seven years. The first is how do you enable consumers to publish stuff about themselves, stuff about things they do or what they’re friends do. Facebook, Myspace, a couple of others go along this line.The next step will be: how do you help them transact? We have gone through the way transactions with Amazon, eBay, etc. I’m fascinated and curious about what eCommerce 2.0 will like. The pundits think social networking doesn’t quite do it. If you want to buy and sell stuff, your friends are probably not the people you’re going to buy and sell from. So you need something else.The ratings and reviews aggregators will try to get into it. I think that’s the interesting problem space that will dominate in the next few years. |
| 09:14 | Sean Ammirati: Ok, with a really interesting 50,000 feet point of view on this. There are a couple of areas I’d like to dive a little deeper on with you. First of all though, I’m curious, before we even get to some of these deeper things. You’ve sit on a pretty senior slab inside of Yahoo. What do you think is Yahoo’s role is inside of this 50,000 foot view and then maybe we’ll step into some of the other players that you see emerging as well.Salim Ismail: The key reason that I am with Yahoo is when you are a startup today, you’re spending all your time figuring out how to be viral. And If you can’t figure out how to be viral, you’re pretty much dead. If you do figure out how to be viral, then you’re servicing the other needs, not doing the cool things you want to do in the first place. So I think that is the case we are with most startups, plus the fact that you have a one in a hundred chance of success. |
| 10:17 | I have been in Yahoo for three weeks. First of all, I love innovation, building ideas, launching them. I love the process, the buzz that comes from that. The second is that it is rare that you get the opportunity to impact how a big company really operates, the opportunity and the privilege to help this company prepare itself for the next 10, 15 to 20 years. It is too much of a process.Third is that assets are unbelievable. I don’t know the exact number but I think we have half a billion registered users. The traffic that anyone can get on those pages is off the charts. It’s unbelievable to me when I talk to the mail guys on about how many users or mails are sent. It just boggles my mind. And the opportunity to leverage those assets is way too cool. |
| 11:07 | Given the momentum that Yahoo has, there are some organization issues and the company is kind of entering its third phase for a mature company. The things move so quickly that structure has to evolve very quickly. That’s why Brickhouse is a little bit separate from the company.The company is geared towards how we can handle half a billion users. It is a completely different mindset from, “How do we launch a really cool new product?” It is sometimes very hard to imagine the two of them. That’s why it has to be separate. Then you know when something succeeds, and then you build it.If you try to do it inside, you’re worrying about the scaling, operationalizing before you have actually tested the concept on the website. So what we’re trying to do is launch cool stuff, iterate it quickly into the marketplace. And then if it’s successful, Pipes in an example of this, launch it, get it out there and see if people use it and how people use it and then bring it back in.And that’s really the principle and the method on how we carry the vision in the workplace and how we kind of rolled out with it. |
| 12:10 | Sean Ammirati: That’s helpful context but I’m still interested in the way you’ve painted this picture of eCommerce 2.0. And I want to dig in some of the other pieces of it but before we even do that, cause I know that time short, the first thing I want to know is what’s Yahoo’ s role in eCommerce 2.0?Salim Ismail: Yahoo is extensively involved in eCommerce. We have the number one personal site in the world. We have the number one finance site in the world. The question is how do you extend all of that? How do you embrace this new way of thinking?Forget Yahoo, I think the whole industry is faced with this question of, “How do you marry eCommerce with user general content where users can self-publish?” How does any company deal with the fact that all the major brands are dealing with the change? I think that is what is coming to eCommerce. And it would probably be central. |
| 13:18 | Let me give you a small view what I see is happening. If you look at the walled garden of content today, whether its eBay or Monster.com or Overture, etc. They basically create and get a lot of value from the ownership of the data.As we get to the UGC (user-generated content) world where users own their own data, what happens to a Monster.com if anybody can publish a job announcement off the website? It then gets syndicated and anybody can aggregate that. The value that they believe in ownership of the data will disappear and where you will get value as a service to layer on top of them.So let’s look at eBay as an example. eBay gets its value; its market count is one of its great things. One is the ownership of all the auction data that they have. And they own that monster list of hundreds of millions of auctions per hour. The second is the payment services provided by Paypal. There’s a lot of value provided by that. And the third is the reputation system that they already arrived at. Some sort of sense of fairness in the marketplace and so on. |
| 14:21 | If you take away or if you start publishing offers to sell off their blogs, or off their Myspace profiles or Facebook profiles or whatever. What happens to eBay? You can now setup and you can compete with that system. And somebody can set up a competitor to Paypal and leverage that. Whereas today, it’s very hard to do that.So that’s one example, one angle on it. Let me step back and talk about HTML versus XML. Really, Web 2.0 is about XML, fundamentally. If you look at the evolution of HTML, 10 years ago to have good content, you have to go to AOL. You have to navigate AOL directories to get the good content. You had to pay to play as a publisher or as a subscriber and AOL tried to hold it off. |
| 15:13 | Then came cheap hosting and tools like Dreamweaver and FrontPage made it easy for anybody to publish HTML. And people like Google and Yahoo came along and indexed it all. AOL got completely disoriented. Now you compete in the HTML world not on data, because it’s owned by the individual source websites, but by the services you layer on top of it.Google indexes it, Yahoo portals it, Alexa ranks it and so on. And that’s really where services layered on top of the underlying data are where the value is gleaned in the HTML world. If you jump in the XML world, we’re still in the stone ages. You have to pay for the privilege of publishing a job announcement on Monster, or publishing a car for sale at Autotrader or selling something on eBay, or even announcing the sale of something on eBay. |
| 16:08 | So the question is why? At what point will that get dimidiated. If you think about the success of Craigslist, it’s largely because it’s free to publish. The only reason it’s successful. So if you can make it free to publish from anything, from anywhere, I think consumers will start using that.So at PubSub, we were doing something called structured blogging. And thousands of people started using it instantly just because it looked nice on their blogs and it was free to publish. And because it was structured, it was more valuable and fundable. I think what will happen in the industry in general is that you’ll have much more XML publishing.The wave of XML publishing and syndication has just taken off. The whole human web is waiting to be exposed, which is currently trapped in the walled garden of databases. And as we release all of that, I think a rising tide lifts them both that enable an entire new class of applications and services that are going out there. |
| 17:05 | Sean Ammirati: So, would you look at somebody like Indeed as a good example in the job space?Salim Ismail: Yeah. Indeed simply hired a bunch of those folks who are doing a lot on the aggregation side and then figuring out how to monetize it. Edgeio is also doing that in the listing space. You look at vFlyer, people like that are trying to enable the consumer to publish, etc.In those three panes, and I’ve got a slide somewhere that I can share with you of publishing and aggregation, etc., you’ll have products and services listed across outputs. Princeton sited Confabb as an experiment to say, “Here’s the conference industry which is underserviced. How can you create a conference portal that can operate and narrow the slice and enter the Web 2.0 world?”So we aggregate. A lot of people who publish it, when they announce a conference we aggregate all the conferences and people will consume and look at conferences. User generated conferences are incorporated. |
| 18:05 | And what I think the Web will see much more of in the next year or two is the Web 2.0 implication of all the traditional contact portals and the little vertical niche applications that exist everywhere. And you can roll them out in months whereas in previous years it would be derailed, etc. I think we’ll see a lot more of those kinds of things.Sean Ammirati: Very interesting. I’m going to shift directions a little bit because the eCommerce was interesting. Another thing you said that I want to pick up on was you made two statements, your opening remarks on this that I want to juxtapose against each other and I probably won’t get the quotes exactly right.The first thing you said is that you felt like blogging was going away and there’ll be just a few thousands of almost-publishers that are just using blogging platforms to publish. And then you also said that you felt that social networking was a fad.But I felt that your case for the blogs going away was that people are going to do that on social networks. I’m curious what comes after social networks or how to reconcile those statements. |
| 19:22 | Salim Ismail: I didn’t say social networks are going away or that blogging is. What I meant to say is that it will transform. A blog is a crude tool that can be used by geeks. It’s not easy to set-up a blog still today. It’s really geared towards text publishing. If you think about RSS, it’s really a wrapper for text.So blogging and text publishing, as we work with HTML – and HTML is again just a wrapper and presentation-ware for text - what blogs have done is create a syndication ecosystem and introduce the concept of self-publishing and user generated content, leading to Flickr and Facebook and all the other sites you have out there. |
| 20:10 | What we used to call bloggers was anybody who would self-publish. I think the concept of self-publishing is here to stay and is trend that will explode over the next couple of years. And I think as the social networks grow, essentially they are mechanisms for self-publishing. And so that’s what social networking is becoming.A blogger is now a professional journalist almost who writes prose for a specific purpose. But when you’re self-publishing in general, there are two silos of it. One is I have to say something, whether it’s prose or I want to announce something or talk about something or give my opinion about something.The value from just raw text is horrible. To extract value from it, you have to do concept matching and extraction and all sorts of horrible algorithmic things to try and guess it. If somebody said, “I really like Harry Potter,” did they mean the book, the movie, how much did they like it, who was the person anyway, etc.? So you try to desperately try to make sense of it all. |
| 21:15 | If you can get somebody to publish a book review or a movie review, give them three stars and syndicate in XML to everyone out there. It becomes much, much more valuable to everybody concerned. And so what Facebook, Twitter, Myspace, a lot of the Yahoo properties is really creating a light structure around the ability to publish.And I think when you provide that light structure, which has a UI challenge to it, you enable the individuals to self-publish in a variety of ways. So we’re used to writing blogs in text prose and it’s syndicated. Now bloggers really write considered posts. If they wanted to just throw something out, they update their Facebook profile or publish something on Twitter and say, “I’m going shopping today.” |
| 22:00 | You don’t see serious bloggers putting too much a rule on going shopping today. So the tools, you have a lot more tools servicing different reasons and different styles of publishing. And that will increase. So I think in general, this concept of self-publishing is exploding. There’s a million different ways of fragmenting that and there’s tools trying to take advantage of that, leverage that, enable all that, all towards a goal of having more structure.And I think what we will see happen over the next few years is as more structure gets implemented into the publishing, it increases the value through the value chain. Now you get real semantics. The semantic web comes into being but you have to start at the source. The reason RDF is trying to implement structure, or extract structure from existing elements which is very hard. The effort really should be in putting more structure and getting more meaning right upfront at the publishing source and then that can be leveraged down the chain. |
| 23:06 | Sean Ammirati: Two things I wanted to ask you about. First is have you looked at the lifestreaming stuff and do you think that as we move in this direction, do you think there is a need for everybody to have one destination where they aggregate all that stuff? Or do you think it should intentionally be atomized and just spread across the web, only to be delivered in context?Salim Ismail: By lifestreaming, you mean the Justin.TV’s, the uStreams, is that what you are going at?Sean Ammirati: No, I mean more like steverubel.com where he is literally posting everything - his Twitter, his blog posts, everything about him – all in one place. I’m talking about that kind of thing.Salim Ismail: Oh. You’re question is do I see that increasing?Sean Ammirati: Yes.Salim Ismail: I think you’ll definitely see that increasing. If Justin.TV is an edge case, and a very leading edge case of that concept where you can see what Justin or Justine or whoever it is, is exploring people what they are doing at any particular time. And again, it’s the same concept of somebody publishing an HTML page, a Geocities homepage 10 years ago, on what they were doing in any given time. |
| 24:18 | The question is twofold. First of all if you get content, as you get more and more content about that person, back at the earlier conversation a point comes into play. How do you now make sense of all of that? Do you really care what Steve Rubel was doing at every minute of the day. There are a large number of people that do; there are a large number of people that don’t.There’s a very vast percent of the population that doesn’t give crap whether I wore brown shoes or black shoes today. And thank God for that. And most people won’t want to know in the same way that most of the HTML or text is published, most people don’t care about it. I think what’s interesting and extraordinarily important which was led by the blogging movement is whereas in the old days you’re voice could not be heard. |
| 25:09 | What blogs introduced was the concept that you have the option to be heard. Somebody who has important view on abortion, if they live in the middle of Kansas and they’re a waitress in a restaurant, there’s no way their viewpoint could actually get out there.Today, their viewpoint can imminently get out there and if it is a good viewpoint, it will be heard and people will notice it. And the people that have a lot of content and have a lot of things to say naturally rise to the top: the Scobles, the Rubels, the Dave Winers, come in line there.And I think that ability is incredibly important for the general human condition. And I think it democratizes information. I think it democratizes value. I think that is probably the most important thing happening in the Internet today. |
| 25:55 | If you look at people on the East Coast, the Boston based guys; they’re enabling blogger in repressed countries to blog. They’re helping bloggers stay anonymous and publish what’s happening in repressed countries and repressed regimes. It’s unbelievably important.And that trend continues through out the world. And again it will take hold at different places at different times, but I don’t think you can stop the trend.Sean Ammirati: Thanks. So the other question I wanted to cover from your statement is that you said, “People do this on Facebook, they do that on Twitter and they do it on a number of Yahoo properties.” I was curious, what are the places on Yahoo properties that you see this doing this most.Salim Ismail: On Yahoo properties?Sean Ammirati: Yup. |
| 26:54 | Salim Ismail: I can’t really answer that question because I’m not familiar enough with all the Yahoo properties. When I first arrived, I was chatting with a senior person at Yahoo and he was saying, “So, what’s you’re plan of attack for Brickhouse?” And I said, “I think we’re going to spend a few days just getting my head around all the different things that Yahoo is involved in.” And he said, “No, stop right there! I absolutely forbid you to do that. Please do not do that, you’ll be gone for a year. Yahoo does everything.”And in most cases, they are number one and number two in that vertical, whether it’s finance or personals, etc. But there is an unbelievable amount of stuff out there, and you’ll spend forever doing it. Don’t go focusing on that.I am aware there are places like Personals which is doing some amazing stuff around lot people who self-publish, on user profiles. Yahoo 360 did an extraordinary job. They didn’t quite cash the marketplace like other things have but it’s such a crap shoot.I would like to say one thing, comment about my view on why things succeed or fail. You have to do everything right, and then you have to be very lucky. There we’re 20 other sites like Flickr and they we’re doing similarly cool things and Caterina and Stewart did a couple of right things and hit the marketplace at the right time and that kind of exploded. |
| 28:18 | If you talk to Zuckerberg or the Yahoo founders or the Google founders on authentic moment they would say that what actually made them successful was really luck. Something happened at the right time, at the right place. What you have to do is prepare yourself so that when that lucky moment happens, you can take advantage of it.And that’s really what the startup world is about. Because there are 200 startups doing video, how does YouTube be the one? And if you read my blog, there was this post I did, I wrote this analogy about turtle eggs. There’s 200 turtle eggs that could hatch. 50 of them don’t hatch because they get late. 50 of them don’t hatch because the shell was too thick.150 start running towards the water and they’re falling into cracks and getting stuck under the stones and the animals are eating them. And then they get to the water surface. And then maybe 50 get out into the water, and the birds are trying to eat them, the fish are trying to eat them. And maybe 5 will get down at the bottom. |
| 29:13 | In any particular space, whether it’s video, whether it’s personal or whatever, there’s 200 startups in each of those trying to do their thing. 5 will succeed; you just don’t know which 5. And the VC game is bet on 10 and one or two will succeed and then you put all you’re weight on that one.That’s really the process that needs to go through for innovation anywhere. So it’s a very evolutionary process. Nobody has ever come up with a better one. People desperately trying to design well, it’s really an evolutionary process where you have 200 startups trying to do this and how do you succeed.What’s interesting about my role in Yahoo is I am able to leverage Yahoo’s unfair advantage. We could do that in an interesting way. Pipes for instance, there’s 10 other companies doing things like Pipes. It succeeds partly because of the coolness, the cool factor and the extraordinary technology innovation but more so because it has got the Yahoo name on it. |
| 30:09 | Google Maps, or AOL or Microsoft all has the same thing. Once you’ve got an established brand, people are going to listen and watch very carefully what we did. And therefore, you were able to succeed.So leveraging that and combining that with the amazing amount of ideas running around Yahoo inside and outside is what is cool about my job.Sean Ammirati: I think that’s a good note to end on. So Salim, I just want to thank you for taking time today to sit down with us and we’ll talk to you soon.Salim Ismail: Thanks very much, I enjoyed this. |
Note: Picture of Y! Brickhouse from Flickr - itsdhiraj & Salim is also from Flickr - Sifry
September 24th, 2007 at 3:16 am
[…] à l’écoute d’un des derniers podcasts de Read/write Web : Interview de Salim Ismail, le boss de Yahoo Brickhouse, je m’interroge sur la notion de valeur. Surtout qu’avec […]