Sunday, June 08, 2003
The following are my notes from the May, 28, 2003 interviews with Terry Semel, conducted by Kara Swisher, and Sergey Brin and Larry Page, conducted by Walt Mossberg, at D: All Things Digital. As with all my conference notes (wherever, whenever), please don't mistake these jottings for a verbatim transcript or a complete portrayal. They are necessarily paraphrased and incomplete, and the product of my highly selective and imperfect notice and attention.
Swisher: There have been some remarkable changes at Yahoo since you came on board. (Mentions articles that initially expressed skepticism about Semel's suitability for the role.)
Semel: For me, it was time for a life change. I examined things, invested in some Internet related things. I wanted to be involved with a business that had great products. I was attracted to Yahoo because it was the best brand on the Internet, and it had a huge audience. There are now some 112 million active registered users.
Swisher: What is Yahoo now?
Semel: That's a very simplistic question. Not sure if there is an answer. Some people would refer to Yahoo as a portal, or a platform, or a network. Probably it's the most relevant place to come for anything you want. The premise is if services and products are great, you'll spend more and more time there. It's not just search, travel, sports, or finance, but getting more personalized, so you the user can design what you want when you want it.
Swisher: Is it a media company?
Semel: I'm surprised you'd use such a traditional word, but it fits if you like.
Swisher: (Asks about Yahoo's success in maintaining advertising revenue.)
Semel: Yahoo's advertising sales now exceed AOL's. Traditional advertising required a lot of handholding. You're starting to see more creative advertising, more use of rich media. Some similarities to TV advertising, but there will be great differences. Broadcast television remains the most effective way to reach the masses. The Internet is the second most effective. In the last year we've seen an enormous change of attitude on both sides. We have a much deeper list of clients starting to believe they're getting their money's worth. We have seen growth for five quarters in a row in our traditional advertising business.
Swisher: (Asks about sponsored search.)
Semel: Sponsored search is auction oriented, very effective for small and medium sized businesses. Works well for all three sides: the advertiser, Yahoo, and the user. Yahoo still is committed to pure search, but users find it a little more organized and relevant when they're getting a "recommendation." It's like the yellow pages. There's a stronger temptation to go to a company you've heard of with a big ad.
Swisher: Let's talk about premium services and extras. There's no proof people want to buy these things. What do you think?
Semel: Look at the track records in other fields, like cable television. People are accustomed to spending money for certain things. Conservatively, 50% of AOL users are regular Yahoo users. Yahoo is essentially "free" to them, so maybe they're willing to pay for certain basic extra services (extra email storage, centralized data access). Yahoo is co-branding with SBC to provide access and services. There's not a single doubt in my mind that people are getting accustomed to paying for certain specific services or products, while still getting certain quality services for free. If we continue to improve the quality of our free services, they will support the pay services, and vice versa. We thought it was important to offer access to listings for personals, jobs and real estate. We decided we needed to buy the jobs segment and personalize it. In the case of personals, we decided we can build it in house and do it really well, but I gave it an advertising budget of zero. The goal was to become #2 in that market with no advertising beyond our own network.
Swisher: What about entertainment? (Mentions AOL, People Magazine online.) As a content maker, what do you think of that happening? What about music?
Semel: 10-11 million people use Yahoo's Music Launch service. There have been 125 million music videos downloaded. Music can and will be successful in its own right. So many people use the 'Net to legally listen to the music they want. There are music clubs. These are perfect examples of the kind of extra services people will pay for.
Swisher: (Asks about whether the music industry will cooperate with online services to give users what they want.)
Semel: As former chairman of one of the five major labels, I see a shift in attitude. The tone used to be, "let's sue them." The industry has done a good job moving away from this. What Steve Jobs has accomplished in obtaining cross-licensing is great. As times change, repackaging and repricing make the most sense. [I missed a bit here...] They're going too slowly though. Now is the time to bring the Internet in, before it becomes too late. The industry should get involved sooner rather than later.
Audience Questions for Semel
Audience member: Jack Warner once cautioned the movie industry not to sell to TV. The result was to encourage innovation in the new medium. Will this happen on the Internet?
Semel: I'm a total believer in the changed medium. Things start out looking like what you're used to, you begin by slicing and dicing what's there. HBO started by broadcasting movies, then moved on to do its own series and other things. The Internet too will begin with slicing and dicing what we've seen before, but that won't be the big killer app. That's going to be original hits that take advantage of the unique aspects of the Internet, and they'll be very different from what we usually see. Music, games. Companies like Yahoo have enormous communities, posting and talking to each other all the time. There's no doubt the Internet will become a major vehicle for games, all over the world. If all you do is copy what came before, you're going to look like an old newspaper. These new ideas could come from Hollywood, but they're just as likely to come from users.
Audience member: Is there a role for Yahoo in convergence, everything coming into the home through one box?
Semel: Yes, it's important to deliver what you want, where and when you want it.
Audience member: Yahoo's cash position is huge. Are you thinking about media properties, your own programming? Semel: Yahoo has north of 2 billion cash right now. Three years out, broadband will be in 50% of households, and more and better content and programming will do well. We don't need a studio now, but we'll be involved in helping people who are doing things we feel are good for the Internet. HBO started out with 100% licensed content, now probably 50% of what they offer is original. And they're thriving.
David Kirkpatrick (from Fortune): Will traditional media throttle Internet content? (Referencing things like Diller's HSN.)
Semel: On the Internet, anyone with a good idea can do it tomorrow. Individuals can and do create stuff. It's much more open to the creation of thriving businesses. The Internet world offers another opportunity to reach masses. As a citizen I worry about countries where one person controls the whole thing. The Internet helps, generally speaking doesn't take positions on issues.
Larry Page and Sergey Brin
Mossberg: You've created great technology that works, but you're sort of the oddball in the search industry. Take sponsored search results. You don't do that.
Page: We've taken a hard line on that. Our search results are the best things we can produce. You can tell our advertising is advertising. It's even more evil when you can't tell if a result comes up because of paid inclusion.
Mossberg: I think this is one of the best things about Google. But from a business point of view, you're not a public company. Why would someone buy a Google ad?
Brin: It's important to us and the users that ads be identified as ads. We think it makes them more effective. (Gives example of purchasing a green laser pointer from one of Google's advertisers.) The ads themselves work well. They're a good revenue stream for Google and for the advertisers. In fact, the ads work so well they're being run on non-search properties, like the Knight Ridder sites. Instead of having a flashy banner you'll see a set of text links, related to the news story you're reading. It's not a perfect product; sometimes you have unfortunate subject-ad pairings—headline, "Boy drowns in washing machine," next to a washing machine ad.
Mossberg: How do you guys make any money?
Page: We license our search technology to other sites, and to the enterprise market. Advertising is a very large source of revenue as well.
Mossberg: What about the simplicity of the ads, why is that important?
Page: We felt like banner ads slowed down our sites and weren't relevant. The click-through rates on the search-relevant text ads are much higher.
Mossberg: There are companies that are pissed off at you guys because you've become the gateway to the Internet. There are people who think Google disadvantages them and their business.
Brin: We were sued in one case I think and I believe it was dismissed. People tend to get really upset when they used to have a big flow of traffic from Google, then they don't. There's another set of people who are getting that traffic. There's not a great deal of stability in our search results. A site might be down when we crawl it, for example. The more important issue is that we continue to develop our algorithms and have a rapid development cycle. People can't necessarily rely on search results remaining static.
Page: We've earned people's trust this way through the quality of the search.
Mossberg: Do you think your average users understand why some results are ranked higher than others?
Brin: The whole system is very complex. I couldn't tell you why in a given set of results one thing is higher than another. I would not recommend following those spams that promise to increase your search result standings.
Mossberg: Have you got this figured out for the next few years?
Page: There's a lot we can do. I still think using Google's terrible. There's still a huge number of things we can't answer. There's probably something out there that explains every complex query, and Google can't give you the exact right answer instantly.
Mossberg: (Asks about specialized searches and features, like Froogle.)
Brin: Most of the products we develop are suggestions by enthusiastic researchers. True of Google News, true of Froogle. We encourage this, and ultimately it's highly motivated people on our teams who conceive these things. You look at our track record on things we planned to launch that were successful, and the correlation is pretty weak.
Page: Innovation in general is like this. You try a lot of things, some of them work out really well.
Brin: We have a list of that we call the Top 100, research projects in development. It's really more like 200. Some people work on them whether we want them to or not. (Audience laughs.)
Mossberg: It sounds like summer camp.
Brin: People create much better things when they're inspired and feel they have ownership, it's what they want to do.
Mossberg: Let's talk about browsing. This doesn't just mean the Web, but the way you interact with the PC, search as more of a metaphor than it has been. Is the kind of thing you do at Google something that could unify all this?
Page: No, Google works best on a large universe of data. With smaller universes, something like the Apple Music Store, you may not even need search. It will be nice to have better search functionality for your own information, but software can provide other alternatives.
Mossberg: What about searching other documents: PDFs, images. AlltheWeb includes music and videos. Will there be a way to use Google to search things that are not on the public Web?
Brin: You can now throw away all your junk mail catalogs because you can search for and browse them on Google. Regarding music and video, there are legal issues, issues with the results really working, issues with player compatibility; generally, usability issues beyond search that make these problematic.
Page: We do have a fair amount of influence now. We try to consider what it means to make all this available.
Audience Questions for Brin and Page
Audience member: Do you help users with queries on things like health care issues to narrow down their results?
Brin: There are great sites about diabetes, and we don't have the ability to give you better information than they can. In the future, who knows?
Audience member: Why search? Why does it work? Page: It was obvious for us because we didn't really want to form a company. There were 10,000 searches a day at Stanford on Google, it was working and growing. We didn't fully understand what would happen, but had strong indications.
Audience member: Google vs. bookmarks. Why don't you let people put their bookmarks on your home page? You'd do serious damage to Yahoo and AOL.
Page: Would you like a job? (Good laugh from audience.)
Same audience member: As one of the two unemployed people here, yes! (More laughs)
Brin: This is the kind of thing people work on on the Top 100 list. It sounds like a good idea but would need to be tested.
Esther Dyson: Asks about Google's purchase of Pyra: what have you discovered, been surprised by, found out?
Brin: We've let those guys go at it, continue to develop their product. They have so many ideas on their own, and there's a whole industry of third party applications to tack onto Blogger. I'm just trying to make sure we don't add too much "value!" (Big laugh.)
Audience member: What are the implications of being able to find just about anything on Google?
Brin: Larry told me this some time ago: people's interests are, and have become, esoteric and diverse. The wealth of information has enabled people to specialize in much narrower interests. It makes it easier for someone to specialize in a localized sphere of knowledge.
Page: I've been waiting for them to start teaching searching, alongside spelling, in school.
Unless otherwise expressly stated, all original material of whatever nature created by Denise M. Howell and included in the Bag and Baggage weblog and any related pages, including the weblog's archives, is licensed under a Creative Commons License.