Chapter III




This dissertation involved five major steps.

                                1.     Provide an historical overview, with a summary of previous works and a review of current                                           literature.

                                2.     Develop and implement a survey.

                                3.     Present an analysis of the results from the survey.

                                4.     Analyze and discuss current Web offerings of various newspapers, comparing these to other                                          Web competition such as television, radio, magazines and cable television.

                                5.     Draw conclusions, make recommendations and present implications.


1.) In step one of this dissertation, the researcher reviewed and summarized the "birth of electronic publishing" (Neustadt 1982). It is important to review past events, and to understand how newspapers have evolved into electronic publishers. Criner (1996) explained that Videotex experiments during the early 1980s made the Los Angeles Times and Washington Post early adopters of electronic information dissemination. She believed that these newspapers (and others) would not be providing the quality on-line services they are today, without those early investments. Much has been written about the Internet, the World Wide Web, cable television, and their relationships to a host of other interactive services. Reviewing this literature provided an insight into current trends in electronic publishing, and yielded additional questions for the survey.

2.) A survey of electronic publishers constituted the second step of this dissertation. Because few scholarly pieces have been written about the newspaper industry and the problems they currently face, a survey of electronic publishers including newspapers, magazines, radio and television stations that are operating Web sites was designed and implemented.


In an effort to create a valid and worthwhile survey, four associates of the researcher, including the advisor of this dissertation and three people with on-line newspaper backgrounds, participated in the initial validation of the instrument. Through a number of iterations, several questions were found to be irrelevant, and were eliminated. Additionally, several questions were added. Next, thirteen experts (webmasters and/or executives) in the on-line publishing field were asked to participate in the survey and provide feedback about the survey questions and their validity. Of the thirteen participants, ten responded, which included representatives from six newspapers, one magazine, one radio, and two from the television industry. Their feedback brought several additional questions, the elimination of several others, and the re-wording (for better clarification) of others. Overall, the participants felt that the survey was "excellent" and should yield very good results. It should also be noted that the survey was offered on-line, a first in the electronic publishing industry.

A Web site was established during the month of July 1997, making the survey available in electronic form. A database was created to collect the survey response data. The Web survey instrument had over 3,000 lines of code to ensure proper validation of entered data. For example, there were checks for numbers in numeric fields, and required fields had to be filled in. The survey instrument can be found in Appendix A.

In an effort to maximize responses, a collaborative effort was made between this researcher and the research department of Editor & Publisher Magazine. In this way, survey recipients were more likely to recognize the well known industry name of Editor & Publisher, and be more inclined to complete the survey. Another advantage in working with Editor & Publisher is through the knowledge and experience they have gained in implementing their past eight surveys in the area of electronic publishing. The letter that was sent as e-mail to the potential respondents, as well as the survey instrument can be found in Appendix A. The survey instrument was submitted to the Nova Southeastern University institutional review board in July 1997, and approved.

Survey implementation

A list of potential survey participants was obtained from the research department of   Editor & Publisher magazine. This list was the result of several years of effort by the research department to determine key people within the four (newspaper, magazine, television, radio) media groups.  Survey participants (2,965 Web publishers) were solicited via e-mail and asked to take part in the survey. Though other forms of electronic publishing exist, such as bulletin board systems, Lexis-Nexis, Dialcom and CD-ROM publishers, they are not perceived by the newspaper industry to be any real threat to newspaper publishers. Accordingly, no attempt was made to include other forms of electronic publishing outside of the Web. Note, at this time, cable television is only a potential threat and is not considered in this study.

The objectives of the survey were:

a.    to gain a better understanding of the driving force behind Web publishing activities.

b.    to measure the level of interest in the Internet among the four media groups, newspapers, magazines,

       radio, and television.

c.    to understand who survey participants consider to be their competitors

d.    to report how survey participants measure consumer activity within their Web site.

e.    to understand if Web prducts are helping or competing against their core products.

f.     to report the staffing sizes of various media, and to identify any trends with regard to staffing size.

g.    to ascertain if Web publishing is profitable.

h.    to provide a comprehensive analysis of the survey data.


The methodology behind this survey is a large scale sampling of a finite group, consisting of participants from the newspaper, magazine, radio, and television industries who currently operate a Web site. Each element had an equal chance of being chosen to become part of the sample. It is important to note that self-selection occurs in nearly all surveys of people (GVU survey, 1997). In the case of this survey, self-selection occurred because entities in the sample were given a choice to participate.

The survey consisted of 61 questions, many with multiple fields, resulting in a total of 197 database fields, most of which were measurable in some form. In taking a close look at these variables, questions were answered, such as: How many full time employees work on the Web edition for those media with less than 50,000 page views per week, 50,000 - 250,000 page views per week, 250,000 – 1 million page views per week, and greater than 1 million page views per week. Are there consistencies (i.e. does the Web staff size go up in size with an increase in page views?). Other questions such as allowing their core product to be "scooped" by their Web product, providing archival information, offering classified advertising, promotional budgets, as well as sales and profitability were addressed by the survey.

3.) Step three of this dissertation consisted of an analysis of the survey data, which was performed by implementing the following procedures:

        a.     Ran all descriptive statistics (Frequency distributions and percentage tables), and cross tabulations.

        b.     Ran covariances on all variables to one another to determine which ones have significant statistical correlations. Determined which ones tend to move together and which ones are causal to others. (See Assumptions, Limitations, Delimitations for more details.)

        c.     Ran scatterplots on those high variance variables to determine if regressions can be developed.

        d.     Developed specifications for models to see if linear relationships exist. For example, by  specifying a pricing model within all four industries, they can be approached separately with an econometric model. If price of banner (TV) equals function of (separate budget, promotion dollars, niche product of community affairs), and, if these three variables all are significant, it can be said with some certainty that there is a relative impact on the price of banner ads.

         e.     Looked at variables and determine what, if any, inferences can be drawn based on the relationships that exist. As mentioned above, there are many variables that can be compared across the four industries.

4.) In step four of this dissertation, the researcher compared the various practices of electronic information competitors. Based on survey results, as well as findings from the research, newspapers, magazine, radio and television Web sites were examined by looking at ways in which they disseminate the news, advertising and other forms of information through their Web site. For example, how many on-line articles are available on a television station Web site as compared to a 250,000+ circulation newspaper Web site?

5.) From the results of the first four steps above, conclusions were drawn, recommendations made and implications (what may or may not happen to the Web publishing industry) presented.



The first assumption of this research is that a study of the World Wide Web and its

relationship to electronic publishing of newspapers is a worthwhile subject of scholarly attention. With the Web in its infancy, just a few years young, few scholarly pieces have been written, and, with the billions of dollars being spent each year by Web publishers, there was clearly a need for a study of this caliber.

The second assumption is that the Web is a fast growing medium, and may have a long term impact, not only on the newspaper industry, but also the magazine, radio and television industries. The relationship of the Web among these four industries is unknown, however, it was intended that the survey, which was implemented as a part of this research, will clear up many unanswered questions and assumptions that were found in doing the literature review.


This survey was limited to a select group of Web publishers, those in the newspaper, magazine, radio and television industry. The survey instrument was created in electronic format and published on a server accessible through the Web. Potential participants (2,965 from a list supplied by the Editor & Publisher research department) were solicited via e-mail and asked to participate by goint to the survey Web site. Leedy (1997) described sample sizes required to adequately represent different population sizes, stating that a population of 3,000 should have 341 respondents.

The ideal response (roughly 341 out of the 2,695 population) was thought to be a possible limitation to this study, as there is no-way to control the number of responses. Ultimately, this was not seen as a problem, as there was great interest by the population in this type of data. Additionally, the survey was accessible via the Web, making it (in most cases) more convenient to respond to. Finally, because the Internet was used for the initial participant contact, and, subsequent reminders (for those who had not yet completed the survey), a high response rate (387) was obtained.

Another limitation was in the accuracy of the participant e-mail addresses that were supplied by the Editor & Publisher research department. This department has been collecting the e-mail addresses of Web site publishers for the past several years, however the accuracy of these addresses was questionable. In the end, it was found that of the 2,965 e-mail address supplied by Editor & Publisher, 572 were invalid, leaving 2,393 potential respondents. See the results section of this dissertation for further comments on the respondents.

Other limitations include slowness of the Web (causing users to not complete the survey), accuracy and/or truthfulness of the answers, and the newness of the Web to some users, which could potential cause incomplete or unknown answers.


Several choices were available for conducting this survey, including U.S. postal mail, e-mail, phone interviews, or allowing the participant to fill out an electronic form on the Web. The latter was chosen for ease of use, reduced cost (in the case of postal mail), reduced time (in the case of phone interviews), and accuracy in writing data to a database. The use of e-mail, however, caused a problem in that a number of recipients considered the e-mail "spamming", or unsolicited e-mail. A number of recipients subsequently responded and asked to be taken off of the mailing list. Several others were apparently more upset, and contacted computer systems managers at Nova Southeastern University, threatening to reject any e-mail from the University that tried to pass through their servers. Needless to say, this caused much grief for the researcher, as well as Nova Southeastern University. Future researchers planning to use e-mail for solicitation should be aware of the potential spamming problem.

Although the survey instrument is somewhat long (61 questions), all questions were deemed necessary and were actually the end result of an initial set of more than 90 questions. Because of the great interest by the industry as a whole, it was felt that participants would not be adverse to the length of the survey. During survey validation, it was determined that the survey should take about 15-20 minutes, however some participants commented that it took much longer because they had to research answers to some of the questions.

Projected outcomes

While it is difficult to predict the future, certain assumptions were made based on previous research. For example, The Philadelphia Inquirer (Mann, 1997) has 16-17   full time people on staff, but is this the norm? For their archives, which include all editorial stories from the last 16 years, Mann stated that the Inquirer charges $6.95 per month, no matter how many stories are downloaded. Based on size, market, competition and other variables, it is valuable to see how newspapers compare to one another for such things as staffing and pricing.

From the 1996 Interactive Survey done by Editor & Publisher (Stoltman, 1997), it was found that 65% of newspapers have Web sites. Of the 35% who do not, 91% said they would have one within 6 months. For this new survey, it was estimated that 85-90% will now have Web sites. Other important factors include content, advertising, promotion, and most importantly, profitability.

It was expected that with the exception of a few electronic publishers, most newspaper Web sites are not making money. It was also expected that few newspaper Web sites are charging subscriptions, few are using push technologies, and some will be downsizing or ceasing Web operations. The most interesting variable was felt to be in the comparison of newspaper’s Web sites to magazine, radio and television Web sites.  Other projected outcomes include, but are not limited to the following:

               1.      Most U.S. newspapers, certainly those of medium to large size, are on the Web.

               2.      Newspaper Web sites are far ahead of their competitors (magazines, radio, television) in                           terms of archives, page views, staffing, content, and profitability.

               3.      Those Web sites with less content have a smaller audience (less page views), and thus lower pricing                           of banner ads, and, less profitability.

               4.      Few companies in the media industry provide Internet services. Those that do will have a larger                           technical staff, and should have more revenue.

               5.      A large percentage will have their Web site on-line less than 2 years.

               6.      Many Web sites operate their own server and require a larger technical staff to support the                           in-house technology.

               7.      The size of banner ads is standardizing as the industry moves forward.

               8.      The use of "Push Technologies" by the media industry as a whole is minimal. However, those                           that do use this technology are seeing a high rate of page views.

               9.      The Web product is having no effect on the core product.

               10.    Few Web sites are doing "shovelware" (putting their core product content on-line) and are                          instead opting to produce original content.

               11.   Web sites with large archives available are getting the most page views, are spending more to provide archives, and are making more money in the process.

               12.   Newspapers have more content than magazines, radio and television.

               13.   Very little, if any, outside measurement data is being provided to advertisers by these Web                          publishers.

               14.    Few Web sites offer on-line classifieds. Those that do have more expenses and should have                          higher revenue.

               15.   Outside of their core product, most Web sites do little promotion and have relatively small                          promotional budgets.

               16.    Newspapers have a larger overall staff size as compared to magazine, radio, and television,                          and the size will stay about the same in 1998.

               17.    Few Web sites charge for access or to retrieve archival information. Those that do charge                          will have lower page views resulting in lower banner rates.

               18.   Few sites require registration, yet many sites are in some way capturing user demographics,                         and will sell this information to third parties.

               19.   The average Web site will spend $50,000 - $100,000 per year for equipment and services                           (not including salaries).

               20.   Overall, there is little, if any, profit among the four groups.

               21.    Competition on the Web (newspaper versus television for example) will be similar to the competition                          they normally have outside the Web, which may be true due to a high rate of local users.

               22.   Few Web sites give the user an ability to customize their content.


The results of this dissertation will provide the newspaper industry with a sound background in the history of electronic publishing. Additionally, survey results and an ensuing analysis will provide a clear picture of what newspapers, as well as their electronic publishing competitors are doing in various markets around the country. Information from the survey results was compiled and analyzed to compare and contrast the various practices used in the production of their electronic editions, and to reveal any industry trends that may be taking shape.

Previous Chapter II

Next Chapter  (Chapter IV)

Return to Main Menu