With so much talk about business models nowadays, it may be useful to say what we mean by such a model.  Granted, that most people who talk about new business models do not themselves have a business, but the question is not idle because even long-established publishers review their models continually.

• If the cost of paper rises, how will that flow through the income statement?
• How do we map that cost against a series of price increases for our publications and overall inflation in the economy?
• If electronic subscriptions rise by 10% per year, and print subscriptions decline by 5% per year, how will this affect overall free cash flow?

The tweaking of the model — the set of dynamically arranged components whose values vary with one another — is endless and, if it is done properly, can provide enormous management insight for setting priorities and identifying trends. For example, a model that shows disproportionate overhead allocated to a print business in the years ahead may drive a publisher to develop marketing plans to migrate more aggressively to an all-digital business; and a model that shows strong advertising revenue for a print business but meager advertising for an electronic one (a common phenomenon) will motivate a publisher to hold onto the print format for many years.

Business models have lots of moving parts — the more complex the business, and the greater the amount of capital required to run it, the more parts that have to be tracked.

What I would like to do here is to focus on just one of those parts, the revenue streams, and to categorize the different sources of revenue available to publishers.  The point to be made is that  revenue streams are not all that exotic.  They fall into a small number of categories, all of which have been with us for decades.  For some oranizations, hybrid models combine two or more categories (some examples below).  Indeed, any innovative publisher will always be looking for ways to open up as many revenue streams as possible.

Traditional Publishing. This is a terrible name for a venerable and excellent means of managing a publishing business.  In an era of Open Access, traditional publishing is sometimes called Toll Access, a term whose bias shows.  Somewhat inelegantly,  I prefer to call it “user-pays” publishing, where the revenue stream derives from the person who actually reads a document or that individual’s proxy, e.g., a library.  In publishing today, this is the most prevalent model.  A bookstore is part of the user-pays ecosystem, whether the store is on Main Street or in the Internet Cloud.  The virtue of traditional publishing is that it aligns payment with demand.  It is also subject to the influence of dominant market players; nor does it have much of a story to tell for publications where there is no or little demand or when that demand can  not be met profitably (e.g., some areas of academic monograph publishing).

Author-pays Publishing. Pioneered in scholarly communications by BioMed Central and then  copied by PLoS, Hindawi, and now many others including Sage, Wiley, BMJ, and AIP, the author-pays model turns traditional publishing on its head and develops a source of revenue upstream, where the intellectual property is created. From an economic point of view, vanity publishing is a form of author-pays publishing, but the term “vanity publishing” is not apt when talking about rigorous editorial policies for some author-pays services (the flagship journals of PLoS are a case in point).   Author-pays publishing does not address a market demand, at least not the demands of the same market as traditional publishing.  When the author becomes the customer, organizations shape themselves to satisfy that customer. We are still watching this drama unfold.  The launch of PLoS One, for example, represents an adjustment of the methods of peer review to accommodate the needs of a different class of authors.  We will see more such accommodations in the future.  The “post-publication peer review” of The Faculty of 1000 is perhaps a harbinger.

The author-pays model is not the same thing as Open Access.  OA is not a business model; it is a property of a publishing service.  While author-pays publishing may in fact embrace the property of OA, a service can also be OA without requiring that authors pay for publication.  The most common form of OA content is supported by media advertising; the SuperBowl is the world’s largest OA event.

Institutional Sponsorship. Often the revenue for a publishing service simply derives from the sponsorship of an institution.  The many repositories at colleges and universities  are examples of this.  So, for example, MIT’s DSpace repository is paid for out of MIT funds, and the California Digital Library’s eScholarship service is a budget line for the University of California.  Any institution (e.g., a governmental agency) may decide to sponsor a repository or publication program.  Many such repositories are OA, but they don’t need to be:  a university, for example, may choose to create a repository that is only accessible to members of that university’s community.  My impression is that  the institutional model has not proven to be as successful as many of its advocates had hoped, but there are no signs of it disappearing.  It may be that institutional repositories will evolve to complement the many OA repositories now being established by organizations without a direct institutional affiliation, some of which are commercial in nature (e.g., Sage, Wiley).  With the current budget strains at universities today, institutional repositories may have to struggle to win administrative support.

Marketing Services. Marketing services that support content cover a lot of territory.  The best-known marketing service of course is advertising, whose revenues can in some instances dwarf what is forthcoming from paid content.  But marketing services go beyond advertising.  For example, some content-based services are used as promotional material to drive traffic to specific Web sites; others capture fields of data that can be converted into sales leads.   An increasingly common marketing service is the use of free versions of content on the open Web to drive sales for print on demand. The strength of marketing services is that when they work, they can generate huge amounts of money (the SuperBowl example again) and such services seem relatively painless, as neither the producer nor the end-user has to pay for them–although some would say that the demands marketing services make on our attention is a high price to pay.   For scholarly publishers, there are inherent problems with the marketing services revenue model.  For one, marketing services can be corruptive, as the solicitation of such revenue could pressure a publisher to alter the content in some way.  Another problem is that publishers of scholarly material often work with audiences that are too small or too highly specialized to attract much interest from marketers.

In addition to these four categories of revenue streams, two others are often cited, though I believe they actually fit into the categories above:

“Freemium” Publishing. This ugly word is a conflation of “free” and “premium.”  The notion is that some portion of a service can be made available for free, but other parts will require payment.  The freemium model is in use just about everywhere.  Sign up for a free account at LinkedIn, but if you want additional features including broader access to other LinkedIn members, you must subscribe to the professional (meaning paid) version.  You can store your files and share them with DropBox, but once the files stored exceed a limit, you will be charged by the amount of storage you require.  You may read abstracts of scientific articles as much as you want without cost, but if you want to read the entire article, you either have to purchase the article or subcribe to the journal.

Freemium services are really a hybrid of two or more of the categories listed above.  A service that provides free content as a mean to “upsell” a user to paid content is, first, a marketing service (free content as promotional tool) and, second, a traditional publisher (pay for the really good stuff).   With Web publishing, just about all services have a freemium dimension:  use free content (e.g., abstracts) to bring users to a site and then attempt to convert those users to paying customers.

Licensing. By “licensing” I am not referring to the licenses that librarians execute to get access for their patrons to digital materials but the business arrangement that one publisher makes with another.  Thus an American university press may work exclusively in English, but may license rights to selected titles to publishers working in German or Japanese.  In some instances a licensee may use the content in a way that is invisible to readers.  For example, search engine companies have licensed a great number of reference works, dictionaries in particular, to enhance search functions.  Thus the word “slow,” licensed by a dictionary company to a search engine, may be used for stemming, the creation of legitimate variant endings (e.g., “slowly,” “slowest”).

As a revenue model, licensing is a special subset of traditional publishing, where the recipient publisher pays for the content from the first publisher and then resells the content into the second publisher’s own market.

Much content is available in hybrid models, though what the hybrids “hybridize” all derive from the four primary categories above.  The hybrid textbook publisher FlatWorld Knowledge hybridizes marketing services (the availability of the full text of a work online for free) with traditional publishing (the sale of PDFs and print-on-demand editions).  A movie studio can travel across a couple categories over time.  Thus the initial release of a film is almost always in the traditional model, with viewers paying for content at theaters.  Later the same film may show up on television, where it is used as a marketing service to attract viewers for the benefit of advertisers.   I expect that we will be seeing author-pays publishing develop a modest revenue stream from traditional publishing, with articles originally appearing in OA (author-pays) repositories being collected and edited for publication in books (traditional model).

So, four categories of revenue streams, which can be mashed together at times or can, for certain properties, stand alone.  Not all of these models work equally well with all kinds of content, however.  To offer an extreme example, you would never attempt to make a feature film on the author-pays model, nor would you use the traditional model of user-pays for highly specialized scholarly material where there is literally no market.

Matching the revenue model with the type of content is important.   The author-pays model, for example, works well for narrowly defined communities, where authors and readers work in close collaboration. The publisher of content that attracts large audiences (this is true for no aspect of scholarly communications) can analyze and package those audiences for advertisers, but some specialized publications (for example, in the medical area) can also derive revenue from advertising provided that the audience is precisely the right one for the products advertised.  There are many kinds of publications for scholars (e.g., highly specialized reference works) that can only be supported through institutional sponsorships, as the costs are too great for the author-pays model and the market too small to earn enough revenue through the traditional model.

When we talk about business models, we have to be clear that there are only a few ways to derive revenue and that digital technology and appealing to organizational mission do not change that.  Whether a publication is the product of a commercial publisher or a not-for-profit educational institution, the publisher must determine how it will generate revenue and align its cost structure appropriately.