This article is adapted from a presentation at the 1999 National Electronic Publishing Seminar sponsored by the University of Virginia and the Library of Congress.

The ideas and opinions expressed in this article are my own and do not necessarily reflect the views of ACM. —B.R.

Online success is context bound. Publishers put many different types of content on line to appeal to reading communities with varying characteristics. Certainly a newspaper and a journal of philosophy do not have the same time criticality or shelf-life. Practicing pediatricians and research physicists have different ways of keeping up with their fields and are likely to differ in their online behavior.

My observations about what contributes to online success are therefore conditioned in part by my experience at ACM. ACM is a not-for-profit scientific and educational society with 80,000 members in the field of information technology. We publish about 50,000 pages each year in twenty journals, thirty-five newsletters, and seventy conference proceedings. Most of the publications are archival outlets for scientific research. Publications are priced modestly: A typical print journal costs about $35 for a member and perhaps $145 for a non-member library. Institutional members, both academic and corporate, pay dues for which they get complete packages of printed ACM publications, heavily discounted from the aggregated individual nonmember prices.

In the past, ACM publication revenue has split fairly evenly between members and institutions. In this respect ACM has differed from most science, technology, and medicine publishers whose main customer is the institutional library.

Criteria for Success

About eight years ago, after a decade of small-scale electronic-publishing experiments, ACM decided to go on line. An ambitious plan was developed. It involved the transformation of the publications production system and the establishment of what we called the Digital Library, an organized body of well-indexed literature, with a variety of views into it, supporting the scholar's need to browse, search, bookmark, retrieve and store articles of interest, from his own desktop, at any time.

A great deal of that plan has been accomplished; there is even more still to do; the work mushrooms as people realize what can be done in this environment. So far, what we have done on line has been successful. One of the chief reasons for that success, I believe, is the significant amount of time spent on planning. Plunging headlong into electronic-publishing experiments is often a good way to gain valuable experience, but it is rarely a recipe for success — at least not success measured in financial terms.

Thorough planning is important when exposure is high, and especially so in an arena marked by great uncertainty, rapid change, and low predictability. The financial hit can be great between up-front capital costs and risks to current revenue streams. In ACM's case, about a third of the society's total revenue is generated by its publications — and all of it was tied to print. The decision to transform our print-based operation to an online-publishing business was therefore not a trivial one. A great deal was at stake.

A second factor that can be critical to success is the often-pedestrian political process of attaining total organizational buy-in. Getting such a common focus slows down implementation. It can be a highly frustrating process. Going on line with publications, however, is part of a larger change. Existing organizational structure and power relationships will be shaken up. This is sure to happen even with planning, but at least all parties are informed. And organizational buy-in can ensure at least an initial level of commitment.

"Servicing users of online information is a major new requirement whose expense is often underestimated"

Moving an organization in a new direction is hard. Few realize initially just how major a change is entailed. A degree of insecurity is natural whenever jobs change and retraining or new hiring is required. The forces of change are buffeted on one side by the trepidation of staff, and on the other side by the ineluctable inertia of large organizations. Planning, and communicating the plans, can minimize internal opposition and resistance.

Typically, new sets of publishing skills, many of them technological, are required to take a publication online. Similarly, new competence in and understanding of publishing applications, their development and support, are required of an organization's information-systems professionals, who now become more involved in the publishing side of the business. It is common for tensions to arise between the publishing and IS divisions: lines of responsibility must often be redrawn, and struggles for turf and control are common. Such tension is not simply the inevitable accompaniment of change. It actually reflects the fact that online information is inextricably intertwined with the technology that stores, retrieves, organizes and displays it. This is not the case in the print world, where the means of content preparation are more readily severed from the product.

Why Go Online?

There are many reasons for the decision to go on line. For a successful outcome, they need to be examined up front and evaluated as part of the planning process.

  • Membership Demands: For ACM, as might be expected, there was a strong push from the membership. Although the members of the Society are computing experts, they are, as a group, decidedly uninterested in association finances. It is important to understand the reasons behind this pressure and to evaluate it in terms of financial impact.

  • Fear: Many publications are also pushed by fear about the future if they do not go on line. Even if they cannot foresee a revenue stream, much less a profitable one, there is a pervasive unease about the possibility of losing their market altogether.

  • Competitive Advantage: There are advantages in the online medium itself. It allows for greater interactivity than print, combining communication with publication. (Those are not the same thing, in my opinion, despite being used interchangeably today.) The digital environment provides multimedia opportunities for visualization of scientific data. (Although this is often cited as a motive for going on line, I would urge some caution: Multimedia presentations can be expensive; it may also take a long time before we evolve a deep understanding of how to effectively present ideas in the new medium, and to develop the appropriate authoring and editing tools. You don't make someone a sculptor by handing him a block of stone and a chisel. It is likely that effective use of this very new medium should be informed by a different pedagogic theory than the one implicit in printed presentations.)

  • Cost Savings: Some are motivated to go on line by the hope of cutting costs. We are probably somewhat less naive about that today than we were a few years ago, yet there is still a tendency to think that once the initial capital investment in the new technology infrastructure has been made, then publishing on line will be cheaper than print. In this regard, it is vital to understand that the information-technology industry makes its money from revisions and obsolescence. The plain fact is that the technology infrastructure for online publishing is inherently unstable. In planning a transition from print to online, it is therefore prudent to take into account the need to continually update the infrastructure when calculating your ongoing operational costs.

    In ACM's case, it has been necessary to beef up the information-systems staff. Other personnel have been retrained and shifted their focus, particularly for the service-support function for digital-library users. Servicing users of online information is a major new requirement whose expense is often underestimated. The publishing division has also required new staff, but that expense has been offset by savings in production.

One additional motive for going on line (and for ACM perhaps the primary consideration) is a perception that something is fundamentally wrong with the traditional scientific-publishing enterprise as a whole: lack of timeliness, e-print servers, more sub-disciplines and specialty publications, rising prices and stressed library budgets, information overload. (See Denning and Rous, The ACM Electronic Publishing Plan 1995, for fuller treatment.)

The system of recognition, promotion, and research funding is the peculiar engine for scholarly publishing. Submissions to journals increase while readership declines, tending dangerously toward the write-only journal. This tendency is the most extreme evidence of the breakdown in traditional scientific publishing; write-only journals emerge as submissions explode and readership dwindles. The human-processing bandwidth is fundamentally out of synch with the publishing system.

For all those reasons, ACM felt that a fundamental change was needed. Going on line appeared to offer a way to address some of the problems with traditional print publishing as we understood them.

Critical Non-Technological Factors

Three nontechnological factors are vital in planning the transition to online publications. In some ways, they are even more important to success than the technological issues.

  • Legal factors: How will copyright and permissions policies function in a networked environment? What, if anything, will be done to ensure the effectiveness of the policies?

    ACM spent more than a year hashing through new usage policies. The ACM Interim Copyright and Permissions Policy was then first published in December of 1994 as a landmark model of how to work with copyright on line. ACM believed that the overhead of building in technological enforcement was greater than the benefit and was likely to make some legitimate uses overly cumbersome. We adopted the view that most people will honor a copyright and permissions policy as long as it does not impose major inconveniences, and that any systematic abuses could be tracked down and dealt with after the fact.

  • Sociocultural factors: How are the patterns of learning behavior and knowledge acquisition changing in general in society, and in your particular market niche? What are the expectations of a community of users in the online networked environment?

    Detailed market intelligence showed an extremely high level and type of Internet connectivity for ACM members and a large amount of time spent online in professional information gathering. (There were contradictory indications about willingness to pay for online publications; the sentiment in favor of free online information was strongly expressed.)

  • Economic factors: What is the online business model and how does it relate to existing print business?

    ACM took a long time to develop its business model. The development cycle included review by various ACM constituencies and subsequent modification. A librarian advisory group and a scholars' advisory group played important parts in this process, as did the publications board and ACM society governance.

Those three factors — the business model, online usage policy, and understanding what your users want and need must all fit together cogently and dovetail with your basic purpose in publishing.

Developing the ACM Online Business Model

The fundamental goal of our business plan was to remain at least net-revenue neutral in the transition to online publication. We recognized that we would have to sustain bimodal publications for an indefinite period. ACM successfully covered its up-front capital investment through a fund-raising campaign to which individual members responded. Not only did that campaign raise the necessary capital, but it also raised the awareness (and expectations!) of the entire membership about ACM's electronic-publishing plans. And it provided another positive indication of membership desires.

One of the most problematic aspects of the business model was the handling of institutional access. Our greatest anticipated risk was loss of revenue from both member subscriptions and member dues by providing institutional access. Most ACM members work in corporate or academic institutions, and we had to estimate what percentage of them might drop their subscriptions if they could gain free access through their institution's subscription. We assumed that the multiple-use factor, (which we used in pricing institutional subscriptions for print publications), would go up. The added convenience of service 24 hours a day seven days a week, and access from the user's desktop, might induce some subscribers (who had maintained personal subscriptions rather than cross campus to the library during specific hours of operation) to drop them. We then had to assume that some percentage of members join ACM solely because of the low cost of publications for members. Those members might be inclined to drop their memberships, as well as their subscriptions, for the same reason.

"The full financial benefit will be reached only when the presses stop running"

In order to offset those risks, ACM had to increase institutional fees for Digital Library access compared to print, and build into the business model a service differentiation; patrons of institutions that licensed the Digital Library would receive basic access while personalized services, such as technical profiles, alert services, and access information for professionals with shared interests, would be reserved for individual members.

Of course, the biggest gamble was the belief that in building a Digital Library, ACM was going to fill a need people would be willing to pay for.

A corollary to that belief was our assumption that an individual electronic journal, in itself, did not provide enough value compared to a printed journal, which is more portable, easier to read, delivered regularly rather than requiring active access, and easier to scan through.

We assumed that the greatest value could be found in a thoroughly indexed Digital Library — a collection of all ACM publications — which gave the ability to search the entire corpus of everything ACM published. The business model and marketing campaign therefore deemphasized subscriptions to individual journals in favor of a single, annual access fee for unlimited usage of our Digital Library.

ACM made one other major assumption in its business model, one that I thought was somewhat shaky. ACM assumed that savings on the print side would at least offset increased costs for the online publication. That assumption has been correct, however, for editorial-production costs; the new in-house costs of preparing the content for online publication are balanced by savings of two kinds: external typesetting costs, and the expenditure on stock and postage. (Note that the savings associated with print manufacture and distribution are only incremental while in bimodal operation. The full financial benefit in terms of savings will be reached only when the presses stop running.)

It seems, however, that it might be more expensive than originally planned to provide value-added functions to the Digital Library service. Once publications have been placed on line, there are increased (and growing) costs in providing additional services to the user, such as sophisticated search, technical-interest profiles and alerting, personalized binders, reference linking, virtual author home pages, interactive discussions, and streaming video. The list of value-adding services and expense continue to grow as an online information resource evolves. Fortunately, the publisher does have a measure of control on these expenses by deciding just how many services to add to content.

The ACM Digital Library went online in 1997 for a period of free access for anyone in the world who could connect via the Web. The collection currently consists of about 22,000 articles from 1991 to the present, totaling about 200,000 text pages. Searchable tables of contents extend back to 1985. ACM does plan to expand its full-text coverage back to earlier years, but we felt that an initial five-year collection constituted a critical mass.

The business model incorporated a decision to keep the search and browse functions of citations and abstracts free and open. In January 1998 we restricted access to full text to paying customers only.

Individual members' Digital Library access fees were treated as an optional add-on to membership dues. (An individual can still subscribe to a single journal, either in print, or online, or both. An online-only subscription is 80 percent of the print-journal price, while getting both versions costs 120 percent of the print price. But for $87 in addition to dues of $88, the cost of the entire Digital Library, including current year and archive, is very attractive to an individual member. It is less than the cost of three journals. For student members, Digital Library access costs only $10 on top of their $25 dues.) Nonmembers (whether individuals or libraries) are not allowed full-text access to the Digital Library except on a pay-per-article basis.

Institutional Member access fees (usually paid by the corporate or academic library) almost doubled over print-package prices. The increased Institutional Member prices reflected ACM assumptions, discussed above, about lost revenue from individual members who were themselves patrons of those institutions. It also reflected an assumption that an institution that required multiple print copies of ACM publications for its various libraries would require only one Digital Library access license. Although the increase was steep, the new total was not perceived to be that high, owing to ACM's historically low library-package pricing. Only three basic packages are offered to institutions, and they cover everything ACM publishes (except books).


In the first fifteen months of online service, ACM gained 30,000 paying subscribers to the Digital Library — 18,000 voting members and 12,000 student members. This compares with an individual subscription base to both journals and proceedings of about 100,000. Financially, ACM is ahead of expectations, both because the business model was conservative and because the Digital Library has proven to be very popular. Net publications revenue has risen despite a significant drop off in printed subscriptions (more than 25 percent, which was mostly anticipated). What comes as a surprise, though, were the number of people who did not drop their print subscription when they added Digital Library access. Presumably those people see an advantage in getting a printed copy of their specialty journal delivered to them, while they use the Digital Library to search and retrieve additional material for which they could never afford subscriptions.

"It appears that consortium buying is beneficial to both sides"

The predicted loss in membership has not occurred, at least not yet. It may be a little too soon for members to realize that their institutions are providing access for them. In fact, during the first year, the reverse happened. There has been a significant increase in members. Most significant is the rise in student memberships and in international memberships.

ACM had more institutions buying access than predicted for the first year. The ACM Librarian Advisory Group advised us that few libraries would be able to increase their budgets much for the ACM material in a single budgeting cycle, even though the total ACM price was still quite moderate. Somehow, a significant number of the institutions managed. Nonetheless, the number of Institutional Members opting for Digital Library access has grown much more slowly than the number of individual members.

ACM did not anticipate at all the level of consortium buying that has occurred, nor the way new markets have opened up as a result. It appears that consortium buying is beneficial to both sides. ACM discounts consortia prices based on the size of the consortium and the United Nations' three-tier classification of countries by wealth. The pricing model for consortia allows each library to save money; it increases the global distribution of ACM scholarship; and it allows some smaller institutions, or ones with only occasional need for the material, to buy access. Discounts to consortia are increasing the volume of buyers and the discounts are coming at the time when prices were significantly raised.

It is a perhaps fortuitous that consortium buying has developed just now as the libraries try to find ways to cope with escalating prices. Consortium buying could apply equally to print. But it is most effective for online publications, since the cost of providing electronic access does not increase linearly with the volume as does the cost of manufacturing and distributing print copies. (Note that there is a step-function at work on line — costs jump suddenly when a given service plateau has been saturated.)

It was also a surprise to see how few articles were bought on a transaction basis. The per-article fee is not high: $10 for a non-member, $5 for a member, and $3 for a student. Possibly, the immediate market of Institutional and Individual Members see the advantage of a single Digital Library access fee, while the larger market of occasional users has not yet been reached.

ACM's business plan underestimated the service and support requirement for users of the Digital Library. There are many new problems: the behavior of particular versions of browsers with particular files; local-printing problems; download performance; broken links; authentication failures; file-access permissions; and the proper structuring of a search query. In addition, a typical online user has an expectation of much faster support response and service at all hours. These are just some of the problems a service staff must now contend with. Some publishers have opened entirely new departments to service their online users. ACM shifted some people into a support role, and retrained current service personnel. Even so, new staff was needed.


ACM has achieved an initial success with its Digital Library offering through organizational planning. Today the key questions facing the organization are: Is this unexpected level of success due to novelty or will it be sustained? When will online growth peak? If aggregators succeed in creating broad, horizontal digital libraries, how will the vertical niche occupied by ACM's Digital Library be affected?

Those important questions need to be answered as we consider how much more to invest in expanding the content and functionality of the digital library.

Bernard Rous is currently deputy director of publications and the electronic-publishing program director at ACM, where he has worked in publishing since 1980. He has been responsible for development and management of database-publishing systems for reference publications and development of early CD-ROM and hypertext products. He was project manager for SGML publishing production systems, and he drafted the copyright and permissions policies for a networked environment. He oversaw the evolution of the business model for ACM's Digital Library. Bernie received his education at Brandeis University and the New School for Social Research. You may contact him by e-mail at

Online publications cited:

The ACM Interim Copyright and Permissions Policy. Most recent version dated December 18, 1998.

Denning, P., and B. Rous. The ACM Electronic Publishing Plan. 1995. Also appeared in Communications of the ACM 38, 4 (April 1995), pp. 97-109.