Online Extra: Q&A with UC Berkeley's Hal Varian

The Net has been hugely successful in communications

Hal Varian, dean of the School of Information Management & Systems at the University of California at Berkeley, is an expert on the economics of information technology. His writings include a book on business strategy, Information Rules: A Strategic Guide to the Network Economy, as well as columns for The Industry Standard and The New York Times. He spoke with BusinessWeek's Heather Green about the future of consumer business models on the Net. Here are edited excerpts from their conversation:

Q: What do you think the value of the Internet is now?


I guess I come from a different background. In education, we always saw the Net as a way to facilitate communications among interested parties. The business model wasn't the topic that was at the front of our minds. The Net has been hugely successful in communications. So with business, if you look at this in a similar context -- how they will use it for internal communications, to connect with suppliers, partners, and customers -- I see a bright future. If we look at how it is changing the nature of the market, it's not that it can't do it, but we are struggling to find the business models that work.

Take Napster. Everyone knows, including the record companies, that you want to distribute digitally -- it's cheaper for one thing. But no one has an answer to how you do it and make money. What it means is, it's a time for experimentation, you have to try different things. The movie industry, which will run into this problem as well, can get a free ride on the music industry's experiments, in the same way TV piggybacked on radio's struggle [to come up with a business model].

Q: What have we learned about certain business models, and what does that teach us about future models?


If you think about payments, one thing that has made people interested in wireless was that it has a built-in payment system. The Net would be a very different place if we had a payment structure coupled into it. If it were like France's Minitel, it would have worked. This is the issue that makes wireless attractive. DoCoMo is making a bundle because they are doing micropayments. We have tried different methods to achieve this [with the Web]. PayPal is interesting. But it's the multiple-standards question, the chicken-and-egg dilemma with the Web and micropayments.

Q: Is online advertising a big bust?


People say it doesn't work. I think they are wrong. One ad model that works extraordinarily well is classified ads. Think about it -- eBay really isn't much more than classified ads. On one hand they are doing the advertising, and then they are also taking a cut of the transaction. The auction stuff is a side game, from my point of view. From my perspective, what's interesting is the matching process. Think of the Net and a classified matching model with lots of buyers and sellers, not a [model that features] broadcast, branding, emotional appeal.

Q: Any other interesting models that come to mind?


Well, take what BusinessWeek is doing. A couple of years ago, we were thinking about the online version as a substitute. Now, we think about how it can complement the offline version. I read BusinessWeek every week, but I will often clip stories. For instance, I write columns for The New York Times. Last week's column was the 12th most e-mailed column for them. What are they doing? They are learning about what people are reading. That helps them make the content better.

People say they aren't making money online. Well, not so clearly. But think about the example of Filene's Basement. They did a study of the revenue producers and found they weren't making money off the rest of the store. So they said, hey, we should get rid of the rest of the store. But what is Filene's Basement the basement of? What we should think about is what complements the existing content. What I will look for in all these online instances is the move to exploit these complementary aspects.

    Before it's here, it's on the Bloomberg Terminal.