Research and Advances
Computing Applications

A Web Site Design Model For Financial Information

Using the Internet to efficiently market companies to a variety of investors.
  1. Introduction
  2. Research Questions
  3. Analysis and Results
  4. Discussion
  5. Conclusion
  6. References
  7. Authors
  8. Footnotes
  9. Tables

It is generally accepted that the Internet has spawned a marketing revolution, providing improved methods for communicating with and selling to customers. Furthermore, as one might expect, the manner in which firms use this new medium differs markedly depending on the nature of their products and their customer base [7]. However, a company’s marketing is not limited to its products, nor is its Web site usually dedicated entirely to consumer use. Many corporations also use their Web sites to disseminate information about their financial performance. In effect, they are using Internet technology to market their company to investors. Our study of the financial content of 203 Web sites found evidence that detailed, objective financial data is associated with companies that have more sophisticated financial consumers. On the other hand, less extensive, more subjective financial data is associated with less sophisticated consumers. This pattern is consistent with theories developed to explain product marketing.

Maheswaran and Sternthal [5] summarize marketing literature that predicts which types of information are most likely to persuade different groups of information users. Consumers that have substantial knowledge of a product genre tend to respond best to objective information about product attributes. For example, Maheswaran and Sternthal found that computer experts were more motivated to process information and make product judgments when presented with attribute information such as “It has a large memory capacity expandable to 512K by bank switching” [5]. In contrast, novice computer users were drawn to more subjective, less extensive information about product benefits: “It has a large memory capacity adequate to handle heavy duty word processing better than the existing word processors” [5]. Research provides empirical evidence supporting the relation between subject expertise (high versus low) and effective routes to persuasion (attribute versus benefit information) [3, 5, 9].

In the context of financial information, literature from the National Investor Relations Institute (NIRI) suggests that investor relations officers explicitly view their activities as a means of marketing their firms’ securities by disseminating financial information [6]. Furthermore, successful marketing involves identifying and satisfying the needs and desires of specific information consumers.

While it is true that Web sites may not be constructed primarily to post financial data, their use for communicating with investors is growing throughout the world. A 1998 survey of investor relations (IR) officers [8] found that launching a Web site was the top priority innovation planned for 1999 for those located in the U.K. and Asia. It was the second priority for those located in Continental Europe. Overhauling an existing Web site was the top priority for IR departments located in the Americas.

Not only do investor relations personnel consider Web sites to be an important means of relaying information to the public, but as more individuals choose their own investments [2, 4], the investing public is coming to rely greatly on corporate Web sites and other Internet-based data sources for cheap, timely investing information.

The role of this new technology in distributing financial information has attracted generally positive attention from the SEC and stock exchange officials, who have left the financial content of corporate Web sites largely unregulated. Firms may present any combination of financial reports, stock price data or other investor-related information, provided that it is not fraudulent and that any new information is also released via news wires or SEC filings. Not surprisingly, in this environment the types and amounts of financial information found at Web sites vary widely across companies.

Back to Top

Research Questions

When applied to financial information presented at corporate Web sites, the marketing literature suggests that objective, attribute-based information is intended for users that possess high levels of financial expertise. We consider professional stock analysts to be information consumers that fall into this category, and so we expect the number of analysts that follow a firm is positively associated with the presentation of objective financial information items at Web sites.

On the other hand, results from this literature suggest that subjective, benefit-focused information is expected to be targeted toward less financially sophisticated users. On average, we expect that individuals investing for their own accounts (retail investors) possess less financial sophistication than professional analysts. We use the total number of shareholders (controlling for the size of the company) as a proxy for the company’s retail investors, and expect to find that companies with relatively more retail investors present more subjective financial information items at their corporate Web sites.

To estimate the objective or subjective content level of each Web site, we considered whether or not certain financial information items were present at each site. The financial information items are listed in Table 1. The list covers a broad range of financial information, from stock price data, to accounting information, to officer speeches. It was developed from interviews with IR directors, IR literature, and visits to a test sample of firms.

Each information item is assigned to one of three groups: Objective, Subjective, or Mixed content. The basic criterion for an item to be considered objective is the information relates a plain fact or is an opinion from an objective third party. Data subject to possible management “spin” is not included. The items classified as objective include the current and historical stock prices, links to third-party stock price sites, information about stock transfer agents, calendars of upcoming financial events (such as earnings release dates), information about dividend reinvestment plans (DRIPS) and information about/reports by analysts who follow the firm (items 1–7 in Table 1).

Conversely, subjective information items are those most likely to contain management interpretation. The information items with the most subjective content include managers’ speeches (item 8) and discussions of the advantages of owning stock, (item 9). The latter are considered to be subjective because historical price increases usually are presented without baselines for comparison with alternative investments.

Six information items, 10–15 in Table 1, are classified as mixed objective and subjective content because they contain both factual information and management discussion. An example of this type of item is accounting reports, which include both factual financial data and management’s interpretation of the numbers.

Back to Top

Analysis and Results

To evaluate the relationships between information clienteles and information, we surveyed the Web sites of two of the largest high-tech industries (as defined by four-digit SIC codes). The sample size was necessarily limited by the nature of the data collection, so focusing on specific industries helped control for possible industry effects. Also, complete industries provide a sample of firms with widely varying ages, sizes, and stages of maturity.

High-technology industries were selected because they are likely to include a relatively large number of companies that maintain Web sites. A list of these companies was obtained from the 1996 S&P Compustat PC Plus database: Semiconductors and Related Devices (SIC 3674), and Biotechnology (SIC 2836). A total of 203 firms are in our final sample.1

The Web sites in our sample were visited separately by two research assistants who examined the sites and recorded the presence or absence of each item on our list. Discrepancies in their results were resolved by a third visit to the site by an author. The frequency with which each item was found is shown in Table 2. Financial news press releases are the most prevalent items (found at 62% of the sites), and transcripts or audio versions of speeches are least common (at only 3% of the sites). Based on these site surveys, an Objective and a Subjective score was calculated for each company based on the presence or absence of each item at each Web site. For example, a company’s Objective score is the sum of the number of items 1–7 found at the site. These scores were used in regression models to see if sophisticated information users (analysts) and unsophisticated information users (retail investors) have different associations with information items found at the corporate Web sites.2 Since prior research has shown that larger companies usually present more financial information in any context [1], a measure of the company’s assets is included in the models to control for company size.3

The results of the statistical tests confirmed our expectations. As shown in Table 3, panel A, there is a highly significant association between the Objective score and the number of analysts (but not with retail investors). Likewise, there is a highly significant relationship between the Subjective score and the number of retail investors (but not analysts). Both Objective and Subjective scores tend to increase as company size increases, consistent with results from prior research.4

Our expectations do not consider the mixed content score explicitly, because it falls between the objective and subjective categories explored in the marketing literature. Our results indicate that Mixed content increases with company size, but not with the number of analysts or retail investors. So, we added a third information clientele that might fall between the expert and less sophisticated user categories: the business press. We found that higher press following (measured by the number of articles about the company appearing in the Wall Street Journal during 1997) is indeed associated with higher mixed scores (results not shown).5

Because any method of classifying Web site financial items as objective or subjective is itself somewhat subjective, and because many of the information items contain mixed content, we considered an alternative classification system. Mainly, we supplemented the Objective/Subjective framework with explicit consideration of whether each item would most likely be used by analysts or retail investors. This determination was based on the Maheswaran and Sternthal insight that detailed, attribute-based information is likely to be preferred by expert users while less sophisticated users favor more benefit-oriented, summarized presentations [5]. The primary difference between this classification scheme and the original is this one allows the mixed content items to be considered within the general context of the marketing theories. Each information item is assigned to one of two groups: Detailed or Summarized.

Table 4 summarizes the allocation of the various financial data items to Detailed Score and Summarized scores. The original Objective/Subjective classification is shown for comparison. Most of the Mixed items (Full Annual Report, Quarterly Reports, SEC Filings, EDGAR Link, and News) are assigned to Detailed because they contain extensive and detailed information, even though some of it is subjective (for example, the annual report items contain a great deal of information, including the typically subjective Letter to Shareholders).

Annual report extracts are assigned to Summarized because they are briefer and more concise versions of the Full Annual Report. Current stock price is a component of both scores. It is objective information, earning it a place in the Detailed Score, but it is also brief and concise. Also, Dividend reinvestment plan information (formerly a component of Objective) is assigned to Summarized because dividend reinvestment plans are generally intended to simplify investing for retail investors.

The statistical associations of the Detailed and Summarized scores with user type are shown in panel B of Table 3; the results reinforce the original findings. The Detailed score is associated with higher analyst following (but not retail investors) and the Summarized score is significantly associated with a higher number of retail investors (but not with analyst following). Both are positively associated with firm size.

Back to Top


The advent of the Web has provided corporate management with a unique method of communicating directly with investors and potential investors. The relationships between Web site financial content and information clienteles documented in these analyses suggest that investor relations departments may indeed attempt to present their companies’ financial results in a manner consistent with marketing theory. In effect, they appear to be marketing their company to a specific investor base.

Previous regulatory efforts have largely been aimed at equalizing the information set available to all groups of users. Thus, the possibility that investor relations professionals may discriminate between groups has important implications. If one assumes the data presented at Web sites is determined in response to consumer demand, our results suggest the increasing the volume of detailed, objective information is not necessarily useful to the growing number of investors with less financial training and sophistication. On the contrary, our results suggest retail investors prefer more high-level data that provides some interpretation of the underlying facts.

A more cynical view might be that the companies are trying to obscure information from naive investors. However, evidence that companies do appear to be forthcoming with detailed, objective data when their clientele base possesses greater financial expertise, plus the fact that detailed data is available from other Internet sources,6 argues against this conclusion.

In either case, the recent surge in investing activity, and the related demand for access to information, has drawn more attention to the distribution and availability of financial data. At the same time, the Internet has provided corporations with a way to dramatically increase the availability of their own perspectives on events and results. Web sites permit much greater flexibility for corporate communication with the masses than previously was feasible.

Back to Top


Our study suggests companies select the data items to be presented on their Web sites determined by the relative sophistication of their particular user base. Specifically, our analyses indicate the information provided at Web sites varies with companies’ levels of analyst following and retail ownership. Higher levels of analyst following are associated with relatively objective, more extensive data, and higher levels of retail ownership are associated with relatively subjective, more abbreviated information. Overall, our analysis provides some support for the notion that firms use their sites to communicate directly with specific information clienteles.

Corporate Web sites are still in the early stages of evolution. Clearly, firms are only beginning to use the capabilities of the Internet to communicate with investors. Microsoft, for example, presents its financial reports in various international reporting formats. Intel provides financial statements that can be easily loaded into Excel for additional analysis. Many companies now have their conference calls online and provide transcripts of the calls soon thereafter (for example, see There are also predictions that in the next few years, firms will use the Internet to disseminate their financial results in real time. It is likely that firms’ ability to target specific information constituencies will also evolve and mature. Their use of this unprecedented potential will likely have a great influence on future investing patterns and populations.

Back to Top

Back to Top

Back to Top

Back to Top


T1 Table 1. Information items and definitions.

T2 Table 2. Frequency of information items at Web sites.

T3 Table 3. Analysis results.

T4 Table 4. Detailed score/summarized score classifications.

Back to top

    1. Ashbaugh, H., Johnstone, K.M., and Warfield, T. Corporate reporting on the Internet. Accounting Horizons (Sept. 1999).

    2. Clements, J. Investors' game: The price is right. The Wall Street Journal (June 29, 1999).

    3. Conover, J.N. Familiarity and the structure of product knowledge. Advances in Consumer Research 9 (1982), 494–498.

    4. Ip, G. Individuals' role in stock market grows as the influence of institutions declines. The Wall Street Journal (Nov. 16, 1998).

    5. Maheswaran, D. and Sternthal, B. The effects of knowledge, motivation, and type of message on ad processing and product judgments. Journal of Consumer Research 17 (June 1990), 66–73.

    6. National Investor Relations Institute (NIRI). Standards of Practice for Investor Relations. National Investor Relations Institute, Vienna, VA, 1998.

    7. Palmer, J.W. and Griffith, D.A. An emerging model of Web site design for marketing. Commun. ACM 41, 3 (Mar. 1998), 45–51.

    8. Stewart, N. Global reach. Investor Relations (Oct. 1998), 24–28.

    9. Walker, B., Celsi, R., and Olson, J. Exploring the structural characteristics of consumer's knowledge. Advances in Consumer Research 14 (1987), 17–21.

    Financial support was provided by a Steve Berlin/CITGO grant, and by the Ernst and Young Center for Auditing Research and Advanced Technology at the University of Kansas.

    1The initial sample consisted of 231 firms. Shareholder data could not be located for 18 firms, and other data were missing for 10 firms, resulting in the final sample of 203 firms. Web sites could not be found for 40 of these firms. All information items for these firms were coded zero (that is, not provided). We also ran our tests excluding the 40 firms without sites (n=163). Results were similar.

    2Specifically, we use a Probit model to test our hypotheses. Probit is an adaptation of standard regression models, used if the dependent variable is not continuous.

    3The distribution of Total Assets is skewed, not normal. Since a normal distribution is best for our statistical tests, we convert assets to the natural log of assets for use in our models.

    4Significance tests are one-tailed for hypothesized relationships, two-tailed otherwise.

    5When Press is added to the models, the relationship between Mixed information and Press is positive and very significant (at the 0.01 level). The relationships between Mixed and the original variables (number of analysts, retail investors and company size) do not change. Nor do the relationships between Objective and number of analysts and Subjective and number of retail investors.

    6The SEC's EDGAR Web site is an example of a source of free, extremely detailed financial data.

Join the Discussion (0)

Become a Member or Sign In to Post a Comment

The Latest from CACM

Shape the Future of Computing

ACM encourages its members to take a direct hand in shaping the future of the association. There are more ways than ever to get involved.

Get Involved

Communications of the ACM (CACM) is now a fully Open Access publication.

By opening CACM to the world, we hope to increase engagement among the broader computer science community and encourage non-members to discover the rich resources ACM has to offer.

Learn More