Web survey and the questionnaire design

The web survey and questionnaire have significant roles for IB professionals, but the concepts are not synonymous. A survey is a methodology for the quantitative description of a population. It collects, relates and clarifies the various primary data, measurements, and observations based on the target samples. On the other hand, a questionnaire consists of scales, appropriate tools, and a bundle of questions to frame the information. It is influential because of the construct measurements and research instruments. However, this study designed and structured the web survey considering the recommendations of Dillman (2000) and Slattery et al. (2011). Subsequently, it is important to recognize the various biases that result in ample flaws in the survey validity and reliability. Usually, the questionnaire’s structure, formatting, length, faulty scales, imperfect wording, incomplete and inconsistent data create various biases in the web survey (Collis

& Hussey, 2009; Mathews & Diamantopoulos, 1995; Slattery et al., 2011).

4.1.1 Sample and the data criteria

The sampling criteria comprise a significant aspect of the study design. There are two types of sampling criteria: probability and non-probability. This study relies on probability sampling to reduce the biases where the random sampling is also a part of that (Min, Park, & Kim, 2016). Principally, this study selected companies that engaged in cross-border M&A deals because there is an absence of corporate branding research in the CBM&A context, even though there is growth in CBM&A. For example, the total value of CBM&As was less than USD 100 billion in the 1980s, but it was USD 555 billion in 2011 (Bjorvatn, 2004; Kedia & Reddy, 2016).

Due to its massive growth and the complexity of the phenomena, the research on CBM&A is indeed persuasive (Bertrand & Zuniga, 2006; Very & Schweiger, 2001). Moreover, this study considered both related and unrelated acquisitions.

Related acquisitions are necessary for the similar complementary markets, efficiency gains, integration cost of the strategic fit, potential synergies and similar value chains. Related acquisitions are a major focus in the USA and EU because of deregulation, globalization, competitive intensification, relaxation of anti-trust legislation, and EU integration (Buckley et al., 2014; Capron, 1999).

Subsequently, the study sampled the acquiring firms instead of the target ones because the acquisition strategy and performance depend entirely on the acquirers (Ambrosini et al., 2011; Hongjiu et al., 2010).

The previous studies contemplated the acquiring firms from the OECD (Bertrand

& Zuniga, 2006; Gubbi et al., 2010), BRICS (Bertrand & Betschinger, 2012;

Chatterjee & Banerjee, 2013; Ma & Zhang, 2010) and APEC countries (Chunlai Chen & Findlay, 2003). However, these distinct clusters do not provide the entire picture of CBM&As around the world. Therefore, this study investigates acquiring firms that accomplished cross-border M&As between 1990 and 2014. The prior studies highlighted that the 1990s were a significant decade for cross-border M&A growth (Bertrand & Betschinger, 2012; Bertrand & Zuniga, 2006). On the flip side, the year 2014 was accentuated by the study because the acquiring firms take at least one or two years to realize their synergies and performance after the acquisition deals (Ambrosini et al., 2011; Chatterjee & Banerjee, 2013; Nguyen et al., 2012).

However, the financial information necessary to measure acquisition performance is not always publicly available because CBM&As are carried out not only by public limited companies, but also by private limited companies.

Performance measurements also differ in terms of the units and consolidated accounts due to the diversity of accounting systems used in different countries.

Therefore, this study considered both public and private limited companies because their performance was measured by the managerial rating (Ambrosini et al., 2011; Changqi & Ningling, 2010; Erel et al., 2012; Meglio & Risberg, 2011;

Schoenberg, 2006). Nevertheless, the size of the acquiring firm and the deal value are also important for the sample section. The prior studies selected the acquiring firms based on their business consolidation, taking into consideration their annual turnover (i.e., USD 25 million to 1 billion) (Homburg & Bucerius, 2006), total assets and market capitalization (i.e., USD 1.1 to 1.4 billion) (Gubbi et al., 2010). The range of acquisition deal value was from 5 million to 21.16 billion (Yang & Hyland, 2012).

This study selected the acquiring firms based on their transaction value (i.e., around 50 million to 10 billion) and annual turnover (i.e., around 49 million to more than 50 billion). Afterward, the size of the target was restrained by the proportionate sales of the target compared to the acquirer at the time of acquisition (Ambrosini et al., 2011; Homburg & Bucerius, 2005, 2006).

The next issue is whether the consumer and industrial product market should be examined in cross-border M&A research. It is self-evident that branding research is frequently carried out in the consumer market (Rui & Lan, 2011). On the other hand, the industrial market is also convincing (Baumgarth, 2010; Han & Sung, 2008; Muylle et al., 2012) because the B2B firms also extend their businesses into the consumer market (Burnaz & Bilgin, 2011). Therefore, this study considered the industrial and consumer market, which are equally substantial in the international trade (Glynn, 2012b; Herbst et al., 2012). Moreover, the study determines which industries should be examined. The prior studies focused on specific industries such as the finance, distribution, engineering, chemical, pharmaceutical (Ambrosini et al., 2011), machinery, logistic and electrical industries (Bauer & Matzler, 2014). On the other hand, a few studies combined specific industries into the manufacturing and service sectors (Collins et al., 2009; Lampela. & Giachettib, 2013; Qing & Qiu, 2013). Splitting these sectors is complicated because manufacturing firms also provide post-sales services.

Therefore, the study focused on both the manufacturing and service sectors, as they are equally important; for example, the service sector in India is better than the manufacturing one, while the manufacturing sector is better than the service one in the USA in terms of their GDP contribution (Gubbi et al., 2010).

In the web survey, few companies mentioned the specific name of the industry rather than their business sector because industry and sector are often used interchangeably even though there are some differences. For example, industryis defined as a specific group of companies in a similar business, while a group of firms in a particular economy indicates the sector. Additionally, the service sector refers to intangible services that are not easily identifiable (e.g., advertising, banking, software, IT). On the other hand, the manufacturing sector involves tangible and physical products that are grouped in terms of the process and discrete manufacturing. Process manufacturing produces the ingredients and formula (e.g., pharmaceutical drugs, soda pop) while discrete manufacturing makes the products from several parts (e.g., automobiles, appliances, and electronics). In line with these characterisations, this study grouped distinct industries into the manufacturing and service sectors.

The industries were the software, trade and retail, ICT, retail & production, financial, insurance, marketing & advertising, credit management services, sales

& marketing, construction materials, oil and gas, mining, media and publishing, mining and mineral processing, and pharmaceutical industries. Several other industries were also included: exhibitions, document technology, production, biotech/diagnostics, energy, life science tools, engineering, and construction (Langager, 2016; Schieltz, 2016).

For cross-border M&A information, this study mainly used the “Thomson One Banker” database (Nguyen et al., 2012; Yang & Hyland, 2012). Additionally, a few more databases were used: the Finnish business magazine “Talouselama,” “the Orbis,” “Forbes Global 2000 in 2015,” “Fortune 1000 in 2015”, and the

“Financial Times 500 in 2015.” Finally, the study sampled 20,000 acquiring firms from around the world. Subsequently, the knowledgeable and experienced top executives such as the CEO, president, vice-president, chairman, directors, global brand manager and investment relationship manager of those companies were specified. Lastly, the author made a database in Microsoft Excel with the executives’ name, position, email, contact number, gender specification, and company details. The information was gathered from press releases, Thomson, Fonecta, Google, company websites, software and LinkedIn (Ambrosini et al., 2011; Ellis, Reus, & Lamont, 2009; Homburg & Bucerius, 2006; Jun et al., 2014;

Nguyen et al., 2012).

4.1.2 Survey administration

There are several types of surveys, such as simple (i.e., snapshot), panel, longitudinal and rotating. Also, a survey can be classified as prospective and retrospective. This study considered only the retrospective and simple web survey. On the other hand, there are several types of survey administration; for instance, telephone, mail, web survey and interview. This study only used the web survey technique to reduce shortcomings in analyzing, coding, real-time participation and apprehending survey data (Poncheri, Lindberg, Thompson, &

Surface, 2008). The web survey technique is superior to mail and telephone surveys. It reduces executives’ unwillingness, inability and nonresponse biases.

Subsequently, the Webropol survey software was used to conduct the web survey, which was authorized by the University of Vaasa. In the second stage, the author put the instructions for participation on top of the questionnaire so that the respondents would understand how to answer the questions. Likewise, the author enabled various media options like smartphone, tablet, and computer because the selected senior officials are used to participating in several media.

In the survey structure, the study also maintained the survey design and proper wording, for example, the number of questions per page, survey logic, length, introduction, and conclusion. It also endeavored to use non-confusing, straightforward and concrete language to avoid biased phrases. Additionally, the study included one open-ended question at the end of the web survey that asked for executives’ opinions. After collecting the data, the author removed the “don’t know” and “skip” options and outliers. The missing data were kept counting the mean value with SmartPLS software (Bauer & Matzler, 2014; De Vellis, 2012;

Min et al., 2016).

4.1.3 The web survey process

Validity and reliability are the roots of psychometric measurement, which originated from classic test theory. Validity indicates the measurement reliability based on the survey administration. Therefore, the study pre-tested the questionnaire in January 2016 on seven executives and academic professionals to verify the construct, content, and face validity (Bauer & Matzler, 2014; Katsikeas et al., 2006; Min et al., 2016). After pre-testing the questionnaire, the contents and questionnaire instructions were reformulated. Also, the length of the questionnaire was set at 36 questions in three pages.

The most critical part of the web survey is to prepare the forwarding letter (i.e., cover letter or motivational letter). It is the pathway to communicating with the senior officials and letting them know how they can participate and why they should forward the survey link to other executives of their acquaintance in the same organization (Collis & Hussey, 2009). In addition, the author concentrated on the email’s subject because it makes the first impression among potential participants. In the invitation to the web survey, the author explained the research objectives, potential findings, global contribution, and the reason for participation. The author also put the software link in the cover letter. Afterward, the spreadsheet (i.e., data bank), outlook express (i.e., university email address) and Microsoft Word (i.e., forwarding letter) were merged to send the survey invitation. The reason behind this process is that the sent invitations were personalized with the participant’s name, position, organization, and gender (e.g., Dear Sir or Madam). Lastly, the study conducted a six-round web survey from March to November 2016. Although 10,000 companies received the survey invitations, only 800 companies contacted the author. Also, they had questions about the conditions of the survey, data security, the period of CBM&A, the business model, the analysis and summary of the research.

In some cases, the CEO’s private secretary arranged the appointments. In other cases, senior officials contacted the author directly. On the flip side, the author ensured the data security and confirmed the recipients; the analysis and the research summary would be sent to them. Finally, 124 acquiring firms participated in the web survey. In a few cases, two executives took part in filling out the survey questionnaire, for example, one official from the M&A department and another one from branding. They also confirmed their participation through email or over the phone.

The response rate of this study is about 15.5% (i.e., 124 out of 800 acquiring companies). It has been defined as the numbers of achieved units divided by qualified units in the web survey sample according to American Association for Public Opinion Research (AAPOR) (Fan & Yan, 2010). For the statistical summary, the response rate of 124 companies is pretty adequate because this is multidisciplinary research grounded on CBM&A and corporate branding (Bauer

& Matzler, 2014; Mathews & Diamantopoulos, 1995; Min, Park, & Kim, 2016;

Slattery et al., 2011). In the professed acquisition studies, the response rate was 20.23% (Bauer & Matzler, 2014), 14.47% (Bauer, Matzler, & Wolf, 2014) and 17.8% (Homburg & Bucerius, 2006). Wright (2015) also noted that in certain cases, the response rate as low as 10% provide the precise estimation. Lastly, the author transferred the collected data into IBM software and SmartPLS professional version 3 for statistical analysis.