Selling intangible goods

Most traditional business models are based on the principle that the goods which they sell are physical objects. These objects can then be sold as individual items in return for a certain amount of money. The merchant exploits the fact that it is difficult to make a copy of the objects. However, all of this changes when intangible goods are being sold.

Since this is a very young field, there are not many reference works available. However, many patents have been granted in this field. When appropriate, a reference to a relevant patent has been included. Please note that this does not mean that that patent covers the technique in question, or even that the patent is a valid one. All it means is that the patent in some way relates to the technique in question.

Why information can't be sold like physical goods

Information is a very broad term. Essentially anything that can be digitized, in other words, encoded as a stream of bits, is information. Information like books, databases, magazines, movies, music, stock quotes and web pages can be sold. Such information is sometimes referred to as information goods.

Information goods are costly to produce but cheap to reproduce. Books that cost thousand of dollars to produce can be printed and bound for a dollar or two and million dollar movies can be copied on DVD for a few cents. In economical terms, production of an information good involves high fixed costs but low marginal costs. The cost of producing the first copy of an information good may be substantial but the cost of producing (or reproducing) additional copies is not usable.

This sort of cost structure has many important implications. For example, cost based pricing just doesn't work. A 10 or 20% profit margin on unit costs makes no sense when unit cost is zero. Therefore, information goods must be priced according to consumer value, and not according to production costs.

Exploiting lock-in effects

When the costs of switching from one brand of technology to another are substantial, users face lock-in. Having locked in customers is a very favorable position for the vendor, since he is assured of repeat sales and the sale of upgrades and add-ons.

Shapiro and Varian (1999) give Bell Atlantic as an example of a locked-in company. Bell Atlantic selected AT&T as supplier of its telephone system. The switches supplied by AT&T employed a proprietary operating system, which meant that Bell Atlantic was entirely dependent on AT&T when it wanted to add a new capability or connect the switches to a new piece of hardware. Since the replacement costs of the switches were extremely high, Bell Atlantic was locked into the switches. AT&T was thus able to charge high prices (for example, 8 million dollars for the software that recognized toll free numbers, or 10 million dollars for voice recognition capability for the telephone directory system).

Lock-in and provider switching

The phenomenon of lock-in also often occurs in subscription-based cable TV systems where users buy a decoder which can only be used to access one particular TV provider. Switching to another provider means having to buy a new decoding apparatus and throwing away the old one which is very expensive. Effectively, the user is locked in to that provider. A competitor can only overcome this problem by lowering the cost of entry, that is, by renting out its decoder or offering a premium for every old decoder that is exchanged for a new one.

A different approach to combat lock-in is to operate as gateway between multiple providers and individual users. A user needs only one box which connects to the gateway. The gateway operator facilitates the subscription to one chosen provider and facilitates a later switch to another provider without the user being required to buy a new box. The gateway operator of course gets a fee every time a subscriber moves to a particular content provider (unpublished WO application EP01/07966). An advantage of this business model is that the user are now effectively locked in to the gateway operator, which gives him a powerful position to negotiate with individual content providers who want to access his user base.

The lesson from this is that lock-in can be an advantage if present in an invention and at the same time can be a disadvantage when present in the prior art and overcome in the invention.

Lock-in through proprietary data formats

For data processing a popular way to lock in users is to keep the data format secret. Users can't move to another program without buying a conversion tool or writing it themselves. However, a risk of this approach is that if the company goes out of business, its customers face inaccessible data or having to keep legacy programs around in order to access it. This generally makes companies very reluctant to buy programs that work with proprietary data formats.

A preferred solution therefore is to publish a specification of the data format and to patent the implementation. This way the company can still gain market exclusivity and at the same time offer the security to its user base that they can always access to the data format because they know the specification.

Lock-in through incentive programs

A simple way to lock customers to a particular provider is to offer a bonus for every purchase which can be spent only at that particular store (US 5,774,870). Many brick-and-mortar stores give away points (such as Douwe Egberts or Esso) or Air Miles or frequent flyer miles. When the user has collected a substantial amount of points, it seems a waste to him not to spend those, so he keeps buying at the store which offers them. Because of the points, he is effectively locked in to that store.

The trick to the store is to put the number of points required for a particular item at a high but reachable level. Further, it is advisable that the points have a limited lifetime so that the user feels the need to spend them or to lose them all.

Customizing the product

When selling a product such as a computer program to someone, there is always the risk that the buyer will redistribute the program to someone else, which means a lost sale. While it is possible to prevent this through copyright law, this is hardly practical when individual home users are involved. The obvious solution is to dilute the value of the product to anyone but the original buyer, while enhancing or maintaining the value to that original buyer.

Making the product buyer-specific

A simple way to make the product of interest only to the buyer is to add the name of the buyer prominently for example on the title bar or on the start-up screen. The buyer himself won't mind - in fact, he might even like it - but other people will. Further, the original buyer may fear that because his name is included in the program the authorities will be able to track the pirated version back to him and make him pay a fine. However, with a little programming skill such a name can easily be wiped from the program code.

A better way is to customize the program itself. Before buying, the user selects the components he needs and pays only for those components. This enhances the value to him because the program does exactly what needs and he does not pay for useless extras. It also dilutes the value to others who don't want that exact same customized version. However, it is still difficult to automate generation of those customized programs.

Customization through watermarking

Audio-visual content can be personalized by the use of watermarks. Using a watermark, the name of the owner can be embedded into a picture, into music, movies etc. Watermarks are difficult or impossible to remove so tracking as outlined above becomes possible. This dilutes the value of redistribution it.

Because watermarks can be detected automatically it becomes even practical to search the whole web for any watermarked unauthorized copies. For instance, US company Digimarc offers an automated tracking service for images (US 5,862,260). A special program, called the MarcSpider, download images from the web and checks for the watermark. If the watermark is found the owner of the image is notified. Should the image have been copied illegally, the owner can take action. Of course, Digimarc asks a fee for this tracking service.

Customizing the price

In addition to customizing the product, the price can be customized as well. This is known as differential pricing or one-to-one marketing. In traditional shops, a particular product can only be offered at one price, since consumers will immediately notice two identical products with different price tags (trust me, I've worked in a shop for years, they do). However, in an electronic market, it is harder for consumers to find out what someone else is paying, so it becomes possible to offer different prices to different buyers, either on an individual or a group basis (e.g., student discounts).

Personalized pricing is already widely used by airline companies. The price of a ticket depends on when the booking is made, whether the trip is in a weekend or not and even on the traveler's history. Similarly, database providers typically charge different prices to companies, government agencies and universities for access to the same database.

One way to achieve differential pricing is to localize the product (for instance, by translating it into a local language) and to price different localized products differently. While the Indian version of a book may be cheaper than the English version, few non-Indians will buy this version.

It is also possible to differentiate the price based on the customer. For example, if a user profile reveals that a user has a large spending potential, the price can be set at a higher level. For instance, students are often given discounts on products. The "rule" on which the price is differentiated must be clearly identified and seem objective, otherwise customers will react negatively (why does my neighbor get a 10% discount and I don't?). The proliferation of price comparison services has made it much more difficult for merchants to differentiate the price based on the customer, since it is difficult to identify the customer when the price comparison service is the entity that retrieves the price from the merchant.

Providing complementary products

One way to sell intangible products is to create two products, one of which is given away for free and the other, the complement of the first, is only available for sale. A complementary product is a product which needs to be used in combination with another product in order to be useful. A famous example is of course the razor blade which is used in combination with a razor. Gillette became the number one company selling razors by giving away the razor with one or two sets of blades, and charging a relatively large price for replacement blades.

Complementary products also exist in E-commerce. A well known example is the Adobe Acrobat Reader. This is a product which can browse and print PDF files and which is given away freely. Its complement, the Adobe Acrobat products which allows the creation and editing of PDF files, can only be bought from Adobe. By giving away the reader for free, Adobe established a large user base which can read PDF files, making it attractive to publish information in this format. This in turn generates sales of the full Adobe Acrobat product.

Using patents to create complementary products

Patents can be used to create complementary products. One technology is licensed at attractive terms (preferably a free license), while the other is more costly. The availability of the first technology (which typically relates to decoding, reading, viewing, browser or other end-user technology) encourages developers to add support for it in their programs. This establishes a large user base of end users who can use the first technology, which creates a demand for supporting the format in creation software. Suppliers of editors, encoders, or writers then need to pay for a license to incorporate the second technology.

An example of how patents can be used to create complementary products is the Graphics Interchange Format (GIF), the most widely used image format on the World-Wide Web. The company Unisys holds a patent (US 4,558,302) on creating images in the GIF format and on decoding them to a format suitable for display. The licensing conditions are simple: a decoder can be built for free, but an encoder requires a paid license. This encourages people to include support for displaying GIF images into graphic viewers but requires the creators of graphics editing programs to pay a license. The main reason that the GIF format has become so popular however, was the use on CompuServe's image library, especially its adult section.

Also important was the fact that Unisys never asserted its patent until GIF had become popular on the World-Wide Web as well. This created the impression to many people that the algorithm was actually free for all. The negative PR from Unisys' aggressive assertion policy has led to the development of the "patent free" format PNG (Portable Network Graphic).

Versioning the information

The principle of versioning information is based on the fact that some customers are willing to pay higher prices in return for some additional advantage. Book publishers first sell a hardback book and then issue a paperback several months later. The impatient consumers buy the high-priced hardback, while the patient ones buy the low-priced paperback. Sites that provide stock quotes typically charge a higher price for real-time quotes than for 20-minute delayed quotes, because real-time quotes are more valuable to the persons who buy them (typically professional brokers).

Versioning is also useful for software. Many programs these days are available in a Standard, Professional and Gold version. These different versions offer different capabilities and come with different price tags. Often, the Standard version is given away for free. This encourages potential customers to try out the program. Once they are used to it, they will begin to notice its limitations. At that moment, they may decide to buy the Professional or Gold version.

Digital rights management

Copyright law has traditionally been used to enforce a merchant business model for sellers of books, music and other information goods. The creation, distribution and publication of the work can be restricted. However, in electronic networked environments, it is very easy to copy and distribute an information good, so for this business model to succeed, technical measures are necessary to enforce these restrictions. This is where digital rights management (DRM) comes in.

In a DRM system, the owner of a content item can add access and usage rules to the content item. A device which receives the content item, such as a television, set-top box or personal computer, must enforce these rules. For instance, if the rules permit the playback of the content item but not its recording onto a storage medium, the player must not allow the user to make a copy.

The owner can now require users to pay for specific rights. Of course, the more rights a user wants, the more he has to pay. The rights are typically bought in the form of a license, which the receiving device needs to access the content. The content itself is encrypted to prevent devices that do not adhere to the DRM requirements from accessing it. The license may contain the necessary decryption key. Alternatively, the decryption key can be embedded in the receiving device, and the device then decrypts the content only if a valid license is present.

The license can be embedded into content, especially audiovisual content, by means of a watermark. A watermark is a visually or audibly imperceptible modification to the content, which can only be detected if a certain piece of information is known. This allows a content owner to embed, for instance, a copyright notice in the content. Should a dispute over the authenticity arise, the owner can detect and retrieve the copyright notice and thereby prove he is the owner of that content. The watermark should be next to impossible to remove. However, this is a difficult problem.