Using Professional Association Newsletters as a Primary Data Source for Measuring Organizational Constructs

Learning Outcomes

By the end of this case, students should be able to

• Identify the types of data that can be gathered from industry and professional association publications
• Understand the opportunities and challenges involved in data collection from published sources such as industry or professional association publications
• Recognize potential opportunities and risks of using data collected by industry and professional associations in studies intended for academic publication

International Franchise Expansion

Franchising is a business model wherein the owner of a business, the franchisor, provides to a franchisee the rights to use the business’ brand, logo, and product or service, along with ongoing support and training in the operation of the business, in exchange for initial fees and continuing royalty payments (Dant & Grünhagen, 2014). Many “chain” businesses are franchise organizations, including restaurants, hotels, and retailers such as Dunkin Donuts, Hyatt Hotels, Burger King, Snap Fitness, Gap, Inc., Auntie Anne’s Pretzels, along with many others. Franchising is a way for a business to expand more rapidly at a lower cost and risk to the owners of the company, because most of the cost and some of the risk of establishing new locations is borne by the franchisee (Fladmoe-Lindquist & Jacque, 1995). Because of the lower cost and risk, many companies expand internationally through franchising, with franchise companies, such as Hilton, McDonalds, Pizza Hut, 7-Eleven, and KFC, being some of the earliest consumer brands to gain a strong global presence.

Over the past several decades, franchising has become an important mode of business growth for U.S. companies seeking to expand internationally. Much of the initial international expansion by U.S. franchise companies was into developed regions of the world, such as Canada and Western Europe (Elango, 2007). However, as we found in our status report on international franchise expansion (Hoffman, Watson, & Preble, 2016), in recent years, companies have focused to a greater extent on emerging markets, including India, China, Brazil, and Russia, where economic growth is higher, and there is a burgeoning set of customers with an income level high enough to support these companies. As a result, franchising has been growing at levels of up to 30% to 50% annually in these countries. For example, one now can visit a Papa John’s pizza restaurant in Russia, a Starbucks in Brazil, and eat at Taco Bell in India. However, the total franchise sector accounts for only 2% to 3% of sales in these emerging markets, which means there is much more room for growth in the franchise sector in developing economies.

While franchise companies wish to gain access to these fast developing markets to build their businesses, the countries, themselves, also desire the investment by U.S. franchisors. Franchising has become a driver of business growth in developing countries, because it aids economic development and provides job growth (e.g., Geromel, 2012). Thus, as the franchising sector has spread globally, many countries have modified their policies and regulations regarding foreign investment. Yet, even though many previously closed markets have become open to foreign investment by franchise companies, countries still vary in the degree to how welcoming their business environments are to new business development. It is this variation in country-level characteristics that we were interested in studying.

The main goal of our research was to understand how various aspects of a country’s business environment affect the expansion of franchise companies into that market. We developed a strong theoretical model but then faced the challenge of finding data sources to measure the variables in that model.

Data Sources

One of the challenges of studying franchise companies is in data collection. Because public companies are required to report financial and other data only at the aggregate corporate level, little public data are available on franchise companies at the individual franchise or brand level or even aggregated at the country level, particularly with respect to franchise expansion, our key dependent variable of interest.

In the past, researchers have surveyed franchisors regarding their intentions to expand internationally as an approach for predicting future expansion, yet this approach has several limitations. Surveys rely on key informants and their perceptions, often measuring franchisors’ intentions rather than their actual expansion activities. Survey studies also are dependent on response rates and suffer from non-response bias. To overcome these data collection problems, we sought other sources of data on the international expansion activities of franchise companies. We began studying various publications, industry groups, and professional associations affiliated with the franchise sector, and found that the International Franchise Association (IFA), the world’s largest franchise association, publishes an online e-newsletter, the IFA SmartBrief (ifa@smartbrief.com), that includes a summary of press coverage of notable activities of franchise companies. This newsletter is published online two to three times per week, and encompasses articles on many aspects of franchising, including announcements of companies’ international expansions, the key dependent variable in our theoretical model. In this way, we found information collected by a relevant industry association to be a valuable data source.

Given the vast number of professional and industry associations, it is likely that there is an organization that collects data relevant to research on many types of organizational phenomena. The researcher’s key task is to identify the associations appropriate for their particular research topic and to discover the forms and types of data those associations collect from their members.

Data Collection Process

Although it may seem that secondary data sources provide an “easy” way to collect data for a research study, that is not necessarily the case. Once we had identified a good source for our dependent variable, international franchise expansion, the data collection process turned out to be rather lengthy and labor-intensive. We monitored the IFA SmartBrief over a 10-year period, and collected all articles that announced specific franchisor expansions into foreign locales during that time. This process resulted in 440 press announcements regarding international expansion by 86 different U.S. franchise companies. We then coded a number of variables from each announcement, including the date of announcement, industry, number and name of country(s) of expansion, number of planned units, time frame of expansion, number of current franchised units in each market, method of entry, and broad rationale for expansion into the specific market. Thus, the data collection process took years.

In contrast, data collection for our independent variables, which were at the country level, was relatively straightforward as there are many sources of country data. However, when multiple data sources are available, it is important to use the data most appropriate for the research topic being studied. In our case, we needed a country data source that not only focused on business-related information but also included data on most countries in the world, within the same database. The World Bank conducts an ongoing research project, Doing Business, in which it gathers and publishes a number of types of data on doing business in various country markets around the world, so it served as source for our business regulation variable while other World Bank databases served as the sources for our other country variables.

Given that we were studying the country-level factors that affect the international expansion of U.S. franchise companies, we also needed to control for other possible variables that might affect franchise expansion. In particular, we controlled for several firm-level variables that previous research has shown affect expansion into international markets. We examined firm archives (websites, company annual reports, and filings with the Security and Exchange Commission) to collect data for variables such as the year the company was founded, existing number of international units, and current number of international markets. We also coded the firm’s ranking on the Entrepreneur magazine Franchise 500 (Entrepreneur, 2013), an annual ranking of franchise companies, for the year of announcement.

Practicalities of Using a Novel Data Source

One of the strengths of our research method is the use of a novel source for information about franchise companies’ international expansion activities—press announcements in the IFA SmartBrief newsletter. However, the use of a unique data source means that reviewers and editors may need reassurance that the source and the data are valid and reliable.

The use of press announcements is quite common in financial research (e.g., event effects on stock prices, mergers and acquisitions, etc.), but this data source is less common in the management and marketing fields. Thus, it was important that we include in our manuscript a detailed description of the data source, as well as the procedure we followed to code the variables from the press announcements. Because editors and reviewers might be skeptical of our data, the data collection section of our paper was a bit longer and more detailed than usual.

Because we were using a new data source, we also felt it was necessary to include evidence that the press announcements provided reliable data on our dependent variable, international expansions. To verify that the expansion plans announced in the press releases revealed true expansion actions rather than merely intentions, we conducted two reliability checks.

First, we selected a random sample of 60 announcements (13.7% of the sample) stratified by year, and checked to make sure the announced plans were accompanied by actual openings of new franchise locations. The majority of these announcements (62%) made note that the opening of the first franchise unit had already occurred or would occur within the next year. Another 33% of the announcements noted the signing of official contracts with overseas franchisees and detailed the date that the unit(s) would open. The remaining 5% were vague about when units would be opened, but they still reflected signed contracts. Given these results, we felt confident that the press announcements accurately reflected our dependent variable, international expansions.

Second, we followed up on a sample of 13% of the announcements made by 57 firms, stratified by years, and examined the associated companies’ websites to determine evidence of expansion plan implementation. The results indicated that 79% of the announced expansions had indeed been executed as planned, 47% were still in the process of a multi-year implementation plan, 19% had not implemented their plans as of the time of the reliability check, with the remaining not providing follow-up data. These reliability checks provided confirmation that the press announcements in the industry newsletter reflected actions rather than merely intentions with respect to international expansions, which helped us convince reviewers that we had a reliable data source.

Issues to Overcome When Using Industry Association Publications as a Data Source

We were successful in publishing two studies using the data we collected from the IFA e-newsletter, yet we encountered several issues in using for academic research data from a source intended for a different purpose. In this case, the IFA SmartBrief, the source of our data on franchise companies’ international expansion activities, has an intended audience of people and organizations directly involved in the franchise sector, such as managers, entrepreneurs, franchise companies and their executives, government entities related to international business or franchising, and non-governmental organizations that monitor or are involved in the franchise sector. The newsletter is not intended to be a source for academic research, so it had some limitations for our purpose.

One issue we encountered was that the sample from this database may not have included all of the international expansions made by franchise companies during the time period we studied. The newsletter includes only expansions by those firms that chose to make public announcements of their plans, whereas, some firms do not officially announce their expansions. It would be difficult to identify unannounced expansions for inclusion in the data set, although there most likely were at least some international expansions by U.S. franchisors that were not documented in this particular industry newsletter. Furthermore, not all franchise companies are members of the IFA, so their expansions may not be as likely to be included in the association’s newsletter. Nonetheless, the newsletter covers activities of most of the major U.S. companies in the franchise sector and as such, the announcements are certainly reflective of expansion activity of industry leaders, and provide a broad data set, points we needed to make clear to the reviewers of our papers.

We also ran into several potential problems while coding the expansion announcements. The IFA SmartBrief newsletter assembles articles from a number of industry sources as well as regional, national, and international news outlets, so the articles that contained franchise expansion announcements were not uniform, and quite often did not include all variables of interest to our studies. We attempted to code for the date of the announcement, country of expansion, number of units, time frame, industry, method of entry, and reasons for expansion, but the inconsistency of reporting resulted in an unusually large amount of missing data. Fortunately, we were able to capture the most important variables for most of the announced expansions, and enough data overall, such that we had a positive publication result.

The variability in the wording of the announcements required some interpretation by the person doing the coding, which could potentially have introduced coder error. The author who was the primary data coder is an expert in international franchising, with the knowledge necessary to make such interpretations. Nonetheless, to ensure coder reliability and accuracy, a second expert in franchise research independently coded a subset of the announcements, which resulted in nearly 90% agreement between the two sets of coded data, giving us, and the reviewers, confidence in the accuracy of our data.

As noted earlier, the IFA SmartBrief newsletter that served as the source of our data included franchise news of many types, not only announcements of international expansions. Identifying expansion announcements within the larger body of articles required one of the authors to monitor the newsletter to tag and save any articles that mentioned international expansion. It was a laborious process, although it was less onerous because of the author’s interest in franchising.

In sum, the use of the franchise association newsletter presented a number of hurdles to overcome, but none was too great to prevent the publication of our research. Although using this data source was time-consuming, it resulted in a unique set of rich data that allowed us to draw insightful conclusions that might not have been available through a different data source.

Practical Lessons Learned

Anyone who has done original research is well aware of how time-consuming it can be to design and implement an empirical project, and see it through to completion. Data collection is just one step in the research process, but it is a step that can make or break a project. In this case, our unique data source was key to our study, yet the data collection was so time-consuming that the study might never have seen the light of day, given that the data collection spanned 10 years.

Who begins a data collection process that will take 10 years, especially not knowing whether the data will end up being useful? In all honesty, we did not intentionally start a 10-year data collection. Unbeknownst to us, our coauthor on one of the papers discussed here, who subscribed to the IFA SmartBrief, began in late 2005 to download and retain in a file all items pertaining to announcements of international expansion of franchise companies. After about 6 years, he turned over to us piles of printed articles that we then coded into a spreadsheet. That coding process took so long that, by the time it was finished, he had another 2 years of articles to add to the spreadsheet, giving us 8 years of data. It took a while to analyze the data and write the first manuscript, at which time an editor asked for 2 additional years of newer data. We were able to comply, which resulted in our 10-year data set.

Fortunately, all authors of both papers are tenured research faculty, who are able to engage in long-term projects such as this one. This type of lengthy data collection process would not be feasible for someone doing a doctoral dissertation or even for a tenure-track faculty member who must complete their research studies more quickly. Indeed, our coauthor who initially started this data collection effort retired before the final publication of the first article based on the data set. However, industry and professional association publications still can be valuable sources of data for doctoral students and untenured researchers, as long as the publications are archived in some manner, so that multiple years of data are available for coding. In our case, we gathered the data in real time over 10 years, but one could just as easily, and more quickly, access previous editions of such newsletters from archival sources, if the organization maintains them.

Conclusion

Here we highlight industry and professional associations as a potential source of data. Given the large number of professional associations, trade groups, and government agencies that exist, this type of data source likely is available for a wide variety of research topics. While some of these associations may maintain databases that might be useful for empirical academic research, many others likely publish newsletters or other types of publications on a regular basis that can be rich sources of data on a number of topics. However, these data often are not in a form that fits neatly into a spreadsheet or database, ready for analysis, so researchers should be aware of the time and effort that may be needed to code the data into useable form.

The novel data source we used, franchise association newsletters, allowed us to study international franchise expansion in a way that had not been done in previous franchise research. As such, the unique data may have helped us get our empirical work published, although it did present some difficulties in both data collection and the journal review process. In the end, the solid results we were able to achieve by using the data outweighed the difficulties, and through several reliability check procedures, we were able to address reviewers’ concerns about the data to publish our work in two different academic journals, and we have high hopes to publish our third study in the future.

Exercises and Discussion Questions

• 1.
  What were the main advantages of the authors’ use of the IFA newsletters as a data source for this project? What were the main problems they encountered?
• 2.
  For what other types of research questions might industry association newsletters serve as a primary data source?
• 3.
  In your field of research, what industry associations or other organizations might have data sources that are useful for your research projects?
• 4.
  Many private and government organizations collect and maintain data on a variety of phenomena. Identify organizations related to your field of research and then identify two or three types of data these organizations collect and maintain. How might you design a study that uses one of these data sources to help you answer a research question?