This is a digital reprint of an interview I did about ten years ago with UC Berkeley’s Hal Varian. At the time Varian was co-author of a bestseller: Information Rules: A Strategic Guide to the Network Economy
; it’s still worth reading today. Today he’s the Chief Economist at Google. There are still a number of good things in this interview that the media companies could learn from… (I’m a bit embarrassed by the silliness of my questions, but hey.)
I suppose we should begin by asking you for your definition of “information” and what you call “information goods”.
When we talk about information goods, we mean anything that can be digitized. Text, pictures, moving images, sound, all the media that can be delivered over a digital connection. Some people call them digital goods.
Information goods have some interesting properties. On the supply side there’s normally a big fixed cost to create the first copy, of say a movie, and then a negligible cost to create additional copies. On the demand side, the interesting feature is that you don’t really know what information is until after you’ve consumed it. So you have to experience it to know what it is.
When you’re selling information, you’re dealing with how do you give free samples, how do you give part of it away, how do you establish a reputation so people will purchase the information you’re providing, etc. etc.
I read about a travel publishing company that put its contents on-line, and their book sales went up, because people wanted the books with them when they traveled…
Yes. Another example is the National Academy of Sciences. They found when they put all their content on line and people could actually look at what the content was, they were more likely to buy.
What are some of the techniques you find companies use to create and sell information products? How do you sell an information product to different customers at different prices? How do you find out what the different customers will pay? Can you do this on a website?
The trick is to “version” your information product: construct a product line of your information goods that will appeal to different market segments. A common way to do this is to use delay: issue a book first in hardback, then, a few months later issue a cheaper edition in paperback. The people who are really interested will get the hardback, whereas people who are only casually interested will wait.
We see financial sites on the Web that sell real-time stock quotes, but give away quotes that are 20-minutes delayed. A movie first comes out first in the theater then six months later in video.
Then there are other things, user-interface, for example. If you look at Dialog, which is a search company, they have two types of search engines- one is a professional search engine, with Boolean searches and all sorts of options, and then they have an “ordinary-person” search engine, with a stripped down interface. It’s nice because the ordinary person wants to use the simpler interface, while the paying professional uses the professional interface. So there isn’t any cross-market cannibalization.
Other dimensions on which to version your product are user convenience, image resolution, capability, features, tech support, etc.
You mention Gresham’s Law of Information in your book. What is it?
Gresham’s law said “bad money crowds out good”. We coined “Gresham’s Law of Information” which says “bad information crowds out good”. Low-quality, cheap information can displace high-quality, authoritative information: look what happened with Encarta and Britannica. However, Britannica is now fighting back and has come out with products that are much better suited to computer use. Smart consumers will look for quality information.
Your example of the struggle between Encarta and Britannica, how Britannica lost out to the upstart $49 Encarta…
Right, although they’re coming back. They’re doing some clever things now. What happens there is the incumbent in the industry has a very low marginal cost, so they should be able to beat the entrant but they can’t quite change their business model. It’s hard. Telephone companies are having this problem, the print/publishing media is having this problem, TV networks have this problem vis-a-vis cable.
(This was before Wikipedia!)
Since there’s a high cost of innovation and a low cost of imitation on the web, isn’t it harder to keep “first-mover” advantages?
You’re right, we talk about this — the competition is only a click away. But the clever company, which has that first-mover advantage, will try its best to create “lock-in” for their customer base. For example, look at what Amazon has done- one click ordering, keeping information on what you purchase so they can recommend books to you. If Amazon is recommending good books to me and I want to switch to say Barnes & Noble, I have to start all over.
Another good example of that is e-toys. You put in the birthday of your nephew, your neice, and your cousins, whatever, and they send you a reminder that your nephew’s birthday is coming up and here’s a nice stuffed rabbit that’s very popular with children in his age group.
Can you tell us a little more about your lock-in strategies?
Since the competition is just a click away on the Web, it pays companies to invest in building customer loyalty. The best way to do this is to produce a product that is so much better than the competition that they don’t want to switch! But there are other ways too, such as loyalty programs, like frequent flyer programs that reward frequent purchasers.
What about lock-in strategies for suppliers and partners?
What we were thinking about there was that if you have a group of loyal customers that are purchasing your products, and there may be other complementary products that they would also purchase, but you may not be the best firm to supply that. So then what you do is sell access to your customers.
The portal companies are doing this. For example, I go to Yahoo, and Yahoo charges other companies to have access to me. Let’s say e-toys wants to move into baby or children’s clothes. They might not do that themselves, but they could partner with other companies that do that.
So once you have a loyal customer base, then you can sell access to that customer base for other products that complement what you are selling.
What about the dangers in this, with privacy issues?
It’s certainly convenient for me to be reminded when my anniversary is or my nephew’s birthday or something. That’s a service, a good thing. Of course they can use the information about me in ways that could be detrimental- they could sell it to mailing lists and I get deluged by email. So the trick is to make sure that consumers give their consent; you want to know exactly how the information is going to be used by the company in question. There are companies like e-trust which meet a very important need.
I was looking at ANX, the auto-industry supplier network, and I found out that Chrysler, despite its enthusiasm during the pilot, isn’t part of the production version of ANX. And if you go to the Chrysler supplier website, you find they’ve created tons of business applications. So when does it make sense to join a standards organization and when does it make sense to go it alone?
There’s this fundamental equation that says that the value to you is your share of the market value times the size of the total market. So some of your actions, like standardization, can increase the total size of the market, but it can decrease your market share because it creates more competition. So you have to trade-off these two effects.
So you’re saying if the total size of the market gets bigger, and you make a bigger profit despite a lower market share, then you are on to something… How do you protect intellectual property on the web? Will the current move of providing patent protection to internet business models help or hurt the future of e-commerce?
The point is to maximize the value of your intellectual property, not maximize its protection. You can charge a lot lower price for content on the Web because you can reach a much larger audience.
I’m quite unenthusiastic about patent protection for Internet business models and feel that it will retard progress in this area.
(Like I said, my questions are quite stupid, but the versioning of information goods - that’s still something the media companies can learn about! This cartoon was also done about the same time…)
Finally, to get you up to speed, here’s a decent
interview with Prof. Varian with the
global-warming deniers at
Superfreakonomics >>
![\varphi = [1; 1, 1, 1, \dots] = 1 + \cfrac{1}{1 + \cfrac{1}{1 + \cfrac{1}{1 + \ddots}}}](http://upload.wikimedia.org/math/8/b/6/8b60ba0565a178cde9cc400bd6c253ab.png)
argues that “the model of American capitalism that worked so well to raise the fortunes of millions of people last century appears to have hit a wall. What’s good for General Motors may no longer be good for the country. In its place must arise a new model of the company, one that serves society as well as rewarding shareholders and employees.” 
- which to me is still the best book in this space.
