Pricing Considerations for Electronic Products in a Network and Requirements for a Billing System - from a Publisher´s Viewpoint

H. D. Köhler, VCH Publishing Group, Weinheim/Germany

1. What you buy and what you pay for

When you buy an apple - and if you are not Adam who paid for it with the loss of paradise - you pay for it in most cases according to its weight. You can eat, or in a more general sense: consume, the apple that's it.

When you buy a dog or another pet, it's yours like the apple, but not quite: you have acquired also a number of responsibilities enforced by laws or social or ethical norms.

When you buy a company, it's again a little bit different: you have acquired some objects, some rights, some responsibilities, some duties, and some obligations.

When you buy a book or a copy of a journal, I think this is something different again. You have acquired the right to read the copy, to resell it, to destroy it. But have you acquired the right to reproduce it - in parts or as a whole ? As you are allowed to breed your dog?
Do you think you paid for the printed paper? Certainly, you did. But equally certainly, you were not as much interested in the paper as in the contents of the book, i. e. what we nowadays call "information".

Finally, in a network you are able, for the first time, to acquire this information without any physical object carrying the information. That is, you now have to pay only for the information. Don't you? You don't pay for anything you did not really want, like a kilo of paper or a plastic tape or disc. Do you? You have access, you select the piece of information which we might still call a "document" and you pay for exactly this. That's fundamentally different from ownership and avoids paying for redundant material in which you are not interested.

Is this reality ? Or a dream? Or a nightmare?

2. Basic Economic Considerations

As I have tried to exemplify by comparing it with apples, dogs and companies, information is a very special merchandise:

• information is immaterial, it can't be consumed;
  
• information can - at almost no cost - be stored, fixed to a carrier material, or transmitted electronically. It thereby can easily be accessed and reproduced;
  
• the utility of information varies according to the kind of use and the time of communication;
  
• information is a product, but it is also more and more acknowledged to be a factor of production - like land, labour and capital.
  
• last, but not least: free flow of information is a basic human right.

With other products prices for information are fixed - at least in a market economy - by individual suppliers who act in a marketplace where there is a demand for their products.

If access to information were unlimited and free, a marketplace with prices paid for information could not exist. Only non-profit organizations and unselfish individuals could invest in information and supply it free of charge.

Copyright law has been developed to protect information from unauthorised dissemination and reproduction, thereby encouraging originators to invest in information and to supply it to the public in the marketplace. Therefore, the purchase of information resembles a license contract. It gives certain rights and imposes certain restrictions. The bond between the originator of information and the information itself is not cut by the sale.

Prices are not determined by cost but cost plays an important role for suppliers: in the long run. They can only survive if the revenues recover their full costs. However, in the short run a supplier might sell at marginal cost or even below.

Production costs for information in a network are - almost by definition - 100 % fix. Incremental costs are, hence, zero. An additional customer, additional access to or download of information cause no extra costs besides telecommunication costs.

This situation gives the "content provider" the option between two principally different pricing strategies1:

• The "skimming price strategy": a high price strategy which provides a healthy margin but risks a depressed sales volume. The high price also makes piracy attractive. Costs for protection against piracy can eat up margins.
  
• The "market penetration strategy": a low price increases the likelihood of full exploitation of the sales potential and discourages potential competitors. Piracy is less attractive, and protection against it might be superfluous. At the extreme, free access to the product is encouraged.

As we all know, the latter strategy has - in the absence of billing systems - been the only option for content providers in the Internet until recently. Obviously, revenues have to be gained later and/or from related products which are sold at profitable price levels.

Theoretical research as well as some prominent empirical examples indicate that this strategy can be extremely successful. However, a short life time of the product and price acceptability problems with users who have become accustomed to having free access, make it very risky. It is a typical strategy for big companies, especially the so-called "new players".

For publishers who want to supply the same information both printed and in a network, the situation is different. Although the incremental costs of supplying an electronic version are minimal, the publisher's main interest often is to protect the income from the printed product. Then a low price strategy for the electronic product is not viable.

In that case, the

• "two-tier pricing strategy"

is usually followed. The price for the combined purchase of printed and electronic versions is considerably reduced compared to printed and electronic versions, each purchased separately.

I would like to emphasize here, that the idea of increased market penetration does work for consumer products but has little value for highly specialised STM-material. The price elasticity of demand is minimal when you have already reached - at whatever price - the small community of those who can understand and utilize your information. Insofar, there is no difference between printed and electronic products.

An upper price limit, of course, is given by the value which the information adds to the product of the user, or by the opportunity cost of the user to get the same information from another source.

If the product has an appeal to a larger group of users at a low price, but also has a high value for a small group of users, then obviously price differentiation is the optimal strategy - again for printed and electronic products alike. With regard to protection costs (in this case also to separate the market segments), this strategy resembles the "skimming price strategy".

Up to now, only prices for the product have been considered. For the user these product prices don't represent the full cost of information. Cost of search for the right information, cost for the individual computer equipment, and last but not least, for telecommunications can easily exceed the price for the product. That's all too often forgotten.

3. Practical Pricing Considerations

For many years prices for online information were based on connect time. While connect time seems to be perfectly appropriate to pay for telecommunications or use of computer equipment, it certainly is not a good indicator of the quantity or the value of the information which is retrieved or transmitted.

Slightly more sophisticated is baud rate pricing which more closely measures the quantity of retrieved information. Nevertheless, both pricing methods are equivalent to selling books according to weight. They undoubtedly have the advantage of being simple and yielding in predictable retrieval cost, but should only be used in combination with a pricing method which relates to the value of the information retrieved.

Such an approach to pricing has to refer to units of information which can be identified as the information to be paid for.

Probably the single most important decision in pricing for electronic products in a network is to choose the size of the unit for pricing2. If it is a whole database, then you have chosen a kind of subscription system. If it is a smaller unit like a "document", a "page", a "screen" or a "data-field" , then you have chosen what often is called a transactional system.

The transactional system seems to be the "dream" for those who seek "just" prices.

The smaller the unit is, the more precisely the user pays only the price of information in which he is really interested. That also allows the content provider to identify exactly the selling and non-selling items in the database.

On the other hand, there are a number of severe disadvantages to small pricing units:

• browsing and searching in order to identify the "target"-units can become complicated and time consuming; tables of contents and catalogs can become very long and less helpful;
  
• administration and billing is more costly the smaller the units are;
  
• in order to recover the costs for the whole database, some parts of which are likely not to sell well, the price for each unit has to be high, potentially provoking resistance from users;
  
• costs for the user as well as revenues for the content provider are unpredictable.

In a databank with well defined factual data these disadvantages do not really count. However, the less well defined the content of a database is and the less predictable the kind and quantity of its use is, the more a larger unit and a subscription system seem to be preferable.

As you could expect from a publisher:
We are back to the subscription system, to the good old pricing system which has been used by publishers for printed products for a long time. It works perfectly for electronic products in a network as well.

Obviously, all the refinements of journal subscription pricing can also be applied to electronic products - and combinations of printed and electronic products (a good example for a mutually agreeable pricing system is the site licensing contract set up by a group of German publishers - assisted by the Börsenverein - and the University Library Bielefeld).

Also obviously, the subscription system has to be complemented - for the "occasional" user - by a transactional system which fulfills the function of a doc.del.service, which ideally would give global access to the products of all publishers.

4. Requirements for Billing Systems in Networks

Not long ago the lack of appropriate billing systems seemed to be the biggest problem for the commercial use of networks, esp. the Internet.

In the meantime, software based systems (like ecash by DigiCash or IBM´s Internet Keyed Payment Protokols (iKP)) have been developed and seem to fulfill most requirements from users and content providers to a satisfactory level - with one exception: their market penetration is still minimal; they are not all yet globally accepted.

A billing system should provide

• easy access for users (by pass-words, credit cards, smart cards etc.);
  
• security for the payment;
  
• anonymity for a user and/or untraceability of a transaction, if so required by the user;
  
• reasonable costs relative to transaction/subscription costs.

A publisher, of course, would like to know for marketing reasons who his customers are, and therefore is not keen on anonymity and untraceability - but as homo politicus has to understand and accept that requirement.

Let me close with a - slightly changed - quotation from A. A. Milne:

"Almost anyone can be "a content provider in a network"3, the business is to collect money (and fame) from this state of being."