Part DECIDING WHICH MARKETS TO ENTER

Transcription

1 II Part DECIDING WHICH MARKETS TO ENTER Introduction to Part II After considering the initia phase (Part I, The decision whether to internationaize) the structure of this part foows the process of seecting the right internationa market. First of a, Chapter 5 presents the most important internationa marketing research toos for anaysing the interna and externa environment. Then the poitica and economic environment (Chapter 6) and the sociocutura environment (Chapter 7) are used as inputs to the process from which the output is the target market(s) that the firm shoud seect as a basis for deveopment of the internationa marketing mix (see Part IV). The structure of Part II is shown in Figure 1. Ë

2 Part II Deciding which markets to enter Figure 1 The structure and process of Part II As Figure 1 shows, the research toos presented in Chapter 5, and the forces in Chapters 6 and 7, provide the environmenta framework that is necessary for the foowing: the seection of the right market(s) (Chapter 8); the subsequent deveopment of the goba marketing mix. The discussion foowing Chapters 6 and 7 wi be imited to the major macroenvironmenta dimensions affecting market and buyer behaviour and thus the goba marketing mix of the firm.

3 5 Goba marketing research Contents 5.1 Introduction 5.2 The changing roe of the internationa researcher 5.3 Linking goba marketing research to the decision-making process 5.4 Secondary research 5.5 Primary research 5.6 Onine (internet) primary research methods 5.7 Other types of marketing research 5.8 Setting up an internationa MIS 5.9 Summary Case studies 5.1 Teepack Speziamaschinen GmbH 5.2 Tchibo 5.3 Video case study: Burke Learning objectives After studying this chapter you shoud be abe to do the foowing: Expain the importance of having a carefuy designed internationa information system. Link goba marketing research to the decision-making process. Discuss the key probems in gathering and using internationa market data. Distinguish between different research approaches, data sources and data types. Discuss opportunities and probems with quaitative market research methods. Understand how onine surveys are carried out. Understand the reevance of the Word Wide Web as an important data source in goba marketing research. 5.1 Introduction Information is a key ingredient in the deveopment of successfu internationa marketing strategies. Lack of famiiarity with customers, competitors and the market environment in other countries, couped with the growing compexity and diversity of internationa markets makes it increasingy critica to coect information in reation to these markets. 153

4 Part II Deciding which markets to enter In contrast to a researcher concerned with ony one country, an internationa market researcher has to dea with a number of countries that may differ consideraby in a number of important ways. Therefore many internationa marketing decisions are concerned with priorities and aocation of resources between countries. The prime function of goba marketing is to make and se what internationa buyers want, rather than simpy seing whatever can be most easiy made. Therefore what customers require must be assessed through marketing research and/or through estabishing a decision support system, so that the firm can direct its marketing activities more effectivey by fufiing the requirements of the customers. The term marketing research refers to gathering, anaysing and presenting information reated to a we-defined probem. Hence the focus of marketing research is a specific probem or project with a beginning and an end. Marketing research differs from a decision support system (DSS) or marketing information system (MIS), which is information gathered and anaysed on a continua basis. In practice, marketing research and DSS/MIS are often hard to differentiate, so they wi be used interchangeaby in this context. At the end of this chapter a proposa for setting up an internationa MIS wi be presented. 5.2 The changing roe of the internationa researcher The roe of internationa market research is primariy to act as an aid to the decision maker. It is a too that can hep to reduce the risk in decision making caused by the environmenta uncertainties and ack of knowedge in internationa markets. It ensures that the manager bases a decision on the soid foundation of knowedge and focuses strategic thinking on the needs of the marketpace rather than the product. Earier marketing research was regarded as a staff function and not a ine function. Marketing researchers had itte interaction with marketing managers and did not participate in marketing decision making. Likewise, externa providers of marketing research had itte interaction with marketing managers. However, as we have moved into the new miennium this ine of demarcation between marketing research and marketing, and thus the distinction between marketing researchers and marketing managers, is becoming thinner and thinner. As the ine and staff boundary burs marketing managers are becoming increasingy more invoved in marketing research. This trend towards making marketing research more of a ine function, rather than a staff function, is ikey to continue and even acceerate in the near future where sense and respond wi increasingy characterize firms approach to business. Thus the traditiona marketing researcher in a commercia firm narrowy focused on the production of presentations and reports for management wi become a rare breed. The transition of marketing researchers to researchers-cum-decision makers has aready begun. Indeed some of the most effective researchers of customer satisfaction are not ony participating in decision making but are aso depoyed as part of the team to impement organizationa changes in response to customer satisfaction surveys. The avaiabiity of better decision toos and decision support systems is faciitating the transition of research managers to decision makers. Senior managers can now directy access interna and externa secondary data from computers and internet sites around the word. 154

5 Chapter 5 Goba marketing research In this miennium good marketing researchers wi be good marketing managers, and vice versa. 5.3 Linking goba marketing research to the decision-making process Primary data Information that is coected first-hand, generated by origina research taior-made to answer specific research questions. Secondary data Information that has aready been coected for other purposes and thus is readiy avaiabe. Goba marketing research shoud be inked to the decision-making process within the firm. The recognition that a situation requires action is the initiating factor in the decision-making process. Even though most firms recognize the need for domestic marketing research this need is not fuy understood for goba marketing activities. Most SMEs conduct no internationa market research before they enter a foreign market. Often decisions concerning entry into and expansion in overseas markets and the seection and appointment of distributors are made after a subjective assessment of the situation. The research done is usuay ess rigorous, ess forma and ess quantitative than in LSEs. Furthermore, once an SME has entered a foreign market, it is ikey to discontinue any research of that market. Many business executives therefore appear to view foreign market research as reativey unimportant. A major reason that firms are reuctant to engage in goba marketing research is a ack of sensitivity to cross-cutura customer tastes and preferences. What information shoud the goba marketing research/dss provide? Tabe 5.1 summarizes the principa tasks of goba marketing research, according to the major decision phases of the goba marketing process. As can be seen, both interna (firm-specific) and externa (market) data are needed. The roe of a firm s interna information system in providing data for marketing decisions is often forgotten. How the different types of information affect the major decisions are thoroughy discussed in the different parts and chapters of this book. Besides the spit between interna and externa data, the two major sources of information are primary data and secondary data: 1 Primary data. These can be defined as information that is coected first-hand, generated by origina research taior-made to answer specific current research questions. The major advantage of primary data is that the information is specific ( fine grained ), reevant and up to date. The disadvantages of primary data are, however, the high costs and amount of time associated with its coection. 2 Secondary data. These can be defined as information that has aready been coected for other purposes and is thus readiy avaiabe. The major disadvantage is that the data are often more genera and coarse grained in nature. The advantages of secondary data are the ow costs and amount of time associated with its coection. For those who are uncear on the terminoogy, secondary research is frequenty referred to as desk research. The two basic forms of research (primary and secondary) wi be discussed in further detai ater in this chapter. If we combine the spit of interna/externa data with primary/secondary data, it is possibe to pace data in four categories. In Figure 5.1 this approach is used to categorize indicator variabes for answering the foowing marketing questions. Is there a market for the firm s product A in country B? If yes, how arge is it and what is the possibe market share for the firm? Note that in Figure 5.1 ony a imited number of 155

6 Part II Deciding which markets to enter Tabe 5.1 Information for the major goba marketing decisions Goba marketing decision phase 1 Deciding whether to internationaize 2 Deciding which markets to enter 3 Deciding how to enter foreign markets 4 Designing the goba marketing programme 5 Impementing and controing the goba marketing programme Information needed Assessment of goba market opportunities (goba demand) for the firm s products Commitment of the management to internationaize Competitiveness of the firm compared to oca and internationa competitors Domestic versus internationa market opportunities Ranking of word markets according to market potentia of countries/regions Loca competition Poitica risks Trade barriers Cutura/psychic distance to potentia market Nature of the product (standard versus compex product) Size of markets/segments Behaviour of potentia intermediaries Behaviour of oca competition Transport costs Government requirements Buyer behaviour Competitive practice Avaiabe distribution channes Media and promotiona channes Negotiation styes in different cutures Saes by product ine, saes force customer type and country/region Contribution margins Marketing expenses per market indicator variabes are shown. Of course the one-market perspective in Figure 5.1 coud be expanded, to cover not ony country B (as in Figure 5.1) but a range of countries, e.g. the EU. As a rue, no primary research shoud be done without first searching for reevant secondary information, and secondary data shoud be used whenever avaiabe and appropriate. Besides, secondary data often hep to define probems and research objectives. In most cases, however, secondary sources cannot provide a the information needed and the company must coect primary data. In Figure 5.1 the most difficut and costy kind of data to obtain is probaby the strengths weaknesses profie of the firm (interna and primary data). However, because it compares the profie of the firm with those of its main competitors, this quadrant is a very important indicator of the firm s internationa competitiveness. The foowing two sections discuss different forms of secondary and primary research. With many internationa markets to consider it is essentia that firms begin their market research by seeking and utiizing secondary data. 156

7 Chapter 5 Goba marketing research Figure 5.1 Categorization of data for assessment of market potentia in a country 5.4 Secondary research Advantages of secondary research in foreign markets Secondary research conducted from the home base is ess expensive and ess time consuming than research conducted abroad. No contacts have to be made outside the home country, thus keeping commitment to possibe future projects at a ow eve. Research undertaken in the home country about the foreign environment aso has the benefit of objectivity. The researcher is not constrained by overseas customs. As a preiminary stage of a market-screening process secondary research can quicky generate background information to eiminate many countries from the scope of enquiries. Disadvantages of secondary research in foreign markets Probems with secondary research in foreign countries are as foows: Non-avaiabiity of data. In many deveoping countries secondary data are very scarce. These weak economies have poor statistica services many do not even carry out a popuation census. Information on retai and whoesae trade is especiay difficut to obtain. In such cases primary data coection becomes vita. 157

8 Part II Deciding which markets to enter Reiabiity of data. Sometimes poitica considerations may affect the reiabiity of data. In some deveoping countries governments may enhance the information to paint a rosy picture of the economic ife in the country. In addition, due to the data coection procedures used, or the personne who gathered the data, many data ack statistica accuracy. As a practica matter, the foowing questions shoud be asked to judge effectivey the reiabiity of data sources (Cateora, 1993, p. 346): Who coected the data? Woud there be any reason for purposey misrepresenting the facts? For what purpose was the data coected? How was the data coected (methodoogy)? Are the data internay consistent and ogica in the ight of known data sources or market factors? Data cassification. In many countries the data reported are too broady cassified for use at the micro eve. Comparabiity of data. Internationa marketers often ike to compare data from different countries. Unfortunatey the secondary data obtainabe from different countries are not readiy comparabe because nationa definitions of statistica phenomena differ from one country to another. The term supermarket, for exampe, has a variety of meanings around the word. In Japan a supermarket is quite different from its UK counterpart. Japanese supermarkets usuay occupy two- or threestorey structures; they se daiy necessities such as foodstuff, but aso cothing, furniture, eectrica home appiances and sporting goods, and they have a restaurant. In genera the avaiabiity and accuracy of recorded secondary data increase as the eve of economic deveopment increases. However, there are many exceptions: India is at a ower eve of economic deveopment than other countries but has accurate and compete deveopment of government-coected data. Athough the possibiity of obtaining secondary data has increased dramaticay the internationa community has grown increasingy sensitive to the issue of data privacy. Readiy accessibe, arge-scae databases contain information vauabe to marketers but considered privieged by the individuas who have provided the data. The internationa marketer must therefore aso pay carefu attention to the privacy aws in different nations and to the possibe consumer response to using such data. Negecting these concerns may resut in research backfiring and the corporate position being weakened. In doing secondary research or buiding a decision support system there are many information sources avaiabe. Generay these secondary data sources can be divided into interna and externa sources (Figure 5.1). The atter can be cassified as either internationa/goba or regiona/country-based sources. Interna data sources Interna company data can be a most fruitfu source of information. However, it is often not utiized as fuy as it shoud be. The goba marketing and saes departments are the main points of commercia interaction between an organization and its foreign customers. Consequenty a great dea of information shoud be avaiabe, incuding the foowing: Tota saes. Every company keeps a record of its tota saes over a defined time period: for exampe, weeky records, monthy records and so on. Saes by country. Saes statistics shoud be spit up by countries. This is party to measure the progress and competence of the export manager or the saesperson (sometimes to infuence earnings because commission may be paid on saes) and party to measure the degree of market penetration in a particuar country. 158

9 Chapter 5 Goba marketing research Saes by products. Very few companies se ony one product. Most companies se a range of products and keep records for each kind of product or, if the range is arge, each product group. Saes voume by market segment. Such segmentation may be geographica or by type of industry. This wi give an indication of segment trends in terms of whether they are static, decining or expanding. Saes voume by type of channe distribution. Where a company uses severa different distribution channes it is possibe to cacuate the effectiveness and profitabiity of each type of channe. Such information aows marketing management to identify and deveop promising channe opportunities, and resuts in more effective channe marketing. Pricing information. Historica information reating to price adjustments by product aows the organization to estabish the effect of price changes on demand. Communication mix information. This incudes historica data on the effects of advertising campaigns, sponsorship and direct mai on saes. Such information can act as a guide to the ikey effectiveness of future communication expenditure pans. Saes representatives records and reports. Saes representatives shoud keep a visit card or fie on every ive customer. In addition, saes representatives often send reports to the saes office on such matters as orders ost to competitors and possibe reasons why, as we as on firms that are panning future purchasing decisions. Such information coud hep to bring improvements in marketing strategy. Externa data sources A very basic method of finding internationa business information is to begin with a pubic ibrary or a university ibrary. The Internet can aso hep in the search for data sources. The Internet has made avaiabe thousands of databases for inteigence research (i.e. research on competitors). In addition, eectronic databases carry marketing information ranging from the atest news on product deveopment to new thoughts in the academic and trade press and updates in internationa trade statistics. However, the Internet wi not totay repace other sources of secondary data. Cost compared to data quaity wi sti be a factor infuencing a company s choice of secondary data sources. Internationa/goba sources (web addresses) The inks to the internationa data sources may be reached at hoensen. Secondary data used for estimation of foreign market potentia Secondary data are often used to estimate the size of potentia foreign markets. In assessing current product demand and forecasting future demand reiabe historica data are required. As previousy mentioned, the quaity and avaiabiity of secondary data are frequenty inadequate. Nevertheess estimates of market size must be attempted in order to pan effectivey. Despite imitations there are approaches to forecasting future demand in a market with a minimum of information. A number of techniques are avaiabe (see Craig and Dougas, 2000), but here ony two are further expained: ead ag anaysis and estimation by anaogy. 159

10 Part II Deciding which markets to enter Lead ag anaysis Determinants of demand and the rate of diffusion are the same in two countries, but time separates the two. Estimation by anaogy A correation vaue (between a factor and the demand for the product) for one market is used in another internationa market. Lead ag anaysis This technique is based on the use of time-series data from one country to project saes in other countries. It assumes that the determinants of demand in the two countries are the same, and that ony time separates them. This requires that the diffusion process and specificay the rate of diffusion is the same in a countries. Of course this is not aways the case, and it seems that products introduced more recenty diffuse more quicky (Craig and Dougas, 2000). Figure 5.2 shows the principe behind the ead ag anaysis with an iustrative exampe in the DVD market. By the end of 2003 it is assumed that 55 per cent of the US househods wi have at east one DVD in their home, whereas it is assumed that ony 20 per cent of the Itaian househods wi have a DVD. We define the time-ag between the American and the Itaian DVD market as two years. So if we were to estimate the future penetration of DVDs in Itaian househods (and as a consequence aso demand) we coud make a parae dispacement of the S-formed US penetration curve by two years, as iustrated in Figure 5.2. This aso shows how rapidy new products today are diffused from market to market. The difficuty in using the ead ag anaysis incudes the probem of identifying the reevant time ag and the range of factors that impact future demand. However, the technique has considerabe intuitive appea to managers and is ikey to guide some of their thinking. When data are not avaiabe for a reguar ead ag anaysis, estimation by anaogy can be used. Estimation by anaogy Estimation by anaogy is essentiay a singe-factor index with a correation vaue (between a factor and demand for a product) obtained in one country appied to a target internationa market. First a reationship (correation) must be estabished between the demand to be estimated and the factor, which is to serve as the basis for the anaogy. Once the known reationship is estabished the correation vaue then attempts to draw an anaogy between the known situation and the market demand in question. Figure 5.2 Lead ag anaysis of penetration of DVDs (Digita Versatie Discs) in the USA and Itay (iustrative exampes) 160

11 Chapter 5 Goba marketing research Exampe We want to estimate the market demand for refrigerators in Germany. We know the market size in the United Kingdom but we do not know it in Germany. As neary a househods in the two countries aready have a refrigerator, a good correation coud be number of househods or popuation size in the two countries. In this situation we choose to use popuation size as the basis for the anaogy: Popuation size in the United Kingdom: 60 miion Popuation size in Germany: 82 miion Furthermore we know that the number of refrigerators sod in the United Kingdom in 2002 was 1.1 miion units. Then by anaogy we estimate the saes to be the foowing in Germany: (82/60) 1.1 miion units = 1.5 miion units A note of caution Generay caution must be used with estimation by anaogy because the method assumes that factors other than the correation factor used (in this exampe popuation size) are simiar in both countries, such as the same cuture, buying power of consumers, tastes, taxes, prices, seing methods, avaiabiity of products, consumption patterns and so forth. Despite the apparent drawbacks to anaogy it is usefu where internationa data are imited. 5.5 Primary research Quaitative and quantitative research Quantitative research Data anaysis based on questionnaires from a arge group of respondents. Quaitative research Provides a hoistic view of a research probem by integrating a arger number of variabes, but asking ony a few respondents. If a marketer s research questions are not adequatey answered by secondary research it may be necessary to search for additiona information in primary data. These data can be coected by quaitative research and quantitative research. Quantitative and quaitative techniques can be distinguished by the fact that quantitative techniques invove getting data from a arge, representative group of respondents. The objective of quaitative research techniques is to give a hoistic view of the research probem, and therefore these techniques must have a arge number of variabes and few respondents (iustrated in Figure 5.3). Choosing between quantitative Figure 5.3 The trade-off in the choice between quantitative and quaitative research 161

12 Part II Deciding which markets to enter Tabe 5.2 Quantitative versus quaitative research Comparison dimension Objective Type of research Fexibiity in research design Sampe size Choice of respondents Information per respondent Data anaysis Abiity to repicate with same resut Interviewer requirements Time consumption during the research Quantitative research (e.g. a posta questionnaire) To quantify the data and generaize the resuts from the sampe to the popuation of interest Descriptive and/or casua Low (as a resut of a standardized and structured questionnaire: one-way communication) Large Representative sampe of the popuation Low Statistica summary High No specia skis required Design phase: high (formuation of questions must be correct) Anaysis phase: ow (the answers to the questions can be coded) Quaitative research (e.g. a focus group interview or the case method) To gain an initia and quaitative understanding of the underying reasons and motives Exporatory High (as a resut of the persona interview, where the interviewer can change questions during the interview: two-way communication) Sma Persons with considerabe knowedge of the probem (key informants) High Subjective, interpretative Low Specia skis required (an understanding of the interaction between interviewer and respondent) Design phase: ow (no exact questions are required before the interview) Anaysis phase: high (as a resut of many soft data) and quaitative techniques is a question of trading off breadth and depth in the resuts of the anaysis. Other differences between the two research methodoogies are summarized in Tabe 5.2. Data retrieva and anaysis of quantitative respondent data are based on a comparison of data between a respondents. This paces heavy demands on the measuring instrument (the questionnaire), which must be we structured (with different answering categories) and tested before the survey takes pace. A respondents are given identica stimui: that is, the same questions. This approach wi not usuay give any probems, as ong as the respondent group is homogeneous. However, if it is a heterogeneous group of respondents it is possibe that the same question wi be understood in different ways. This probem becomes especiay intensified in cross-cutura surveys. Data retrieva and anaysis of quaitative data, however, are characterized by a high degree of fexibiity and adaptation to the individua respondent and his or her specia background. Another considerabe difference between quaitative and quantitative surveys is the source of data: Quantitative techniques are characterized by a certain degree of distance as the construction of the questionnaire, data retrieva and data anaysis take pace in separate phases. Data retrieva is often done by peope who have not had anything to do with the construction of the questionnaire. Here the measuring instrument (the questionnaire) is the critica eement in the research process. Quaitative techniques are characterized by proximity to the source of data, where data retrieva and anaysis are done by the same person, namey, the interviewer. Data retrieva is characterized by interaction between the interviewer and the respondent, where each new question is to a certain degree dependent on the 162

13 Chapter 5 Goba marketing research previous question. Here it is the interviewer and his or her competence (or ack of the same) which is the critica eement in the research process. Quaitative techniques impy a ess sharp separation between data retrieva and anaysis/interpretation, since data retrieva (e.g. the next question in a persona interview) wi be dependent on the interviewer s interpretation of the previous answer. The researcher s persona experience from fiedwork (data retrieva) is generay a considerabe input into the anaysis phase. In the foowing section the two most important quaitative research methods are presented. Trianguation: mixing quaitative and quantitative research methods Quantitative and quaitative research methods often compement each other. Combined use of quantitative and quaitative research methods in the study of the same phenomenon is termed trianguation (Denzin, 1978; Jick, 1979). The trianguation metaphor is from navigation and miitary strategy, which use mutipe reference points to ocate an object s exact position. Simiary, market researchers can improve the accuracy and vaidity of their judgements by coecting both quantitative and quaitative data. Sometimes quaitative research methods expain or reinforce quantitative findings and even revea new information. Sometimes it is reevant to use quaitative data coected by, for exampe, in-depth interview of a few key informants as exporatory input to the construction of the best possibe questionnaire for the coection of quantitative data. In this way trianguation can enrich our understanding of a research question before a structured and formaized questionnaire is designed. Research design Figure 5.4 shows that designing research for primary data coection cas for a number of decisions on research approaches, contact methods, samping pan and research instruments. The foowing pages wi ook at the various eements of Figure 5.4 in further detai. Research probem/objectives Companies are increasingy recognizing the need for primary internationa research. As the extent of a firm s internationa invovement increases, so does the importance and compexity of its internationa research. The primary research process shoud begin with a definition of the research probem and the estabishment of specific objectives. The major difficuty here is transating the business probem into a research probem with a set of specific researchabe objectives. In this initia stage researchers often embark on the research process with ony a vague grasp of the tota probem. Symptoms are often mistaken for causes, and action determined by symptoms may be oriented in the wrong direction. Research objectives may incude obtaining detaied information for better penetrating the market, for designing and fine-tuning the marketing mix, or for monitoring the poitica cimate of a country so that the firm can expand its operations successfuy. The better defined the research objective is, the better the researcher wi be abe to determine the information requirement. Research approaches In Figure 5.4 three possibe research approaches are indicated: observation, surveys and experiments. 163

14 Part II Deciding which markets to enter Figure 5.4 Primary data coection: research design Observation This approach to the generation of primary data is based on watching and sometimes recording market-reated behaviour. Observationa techniques are more suited to investigating what peope do than why they do it. Here are some exampes of this approach: Store checks: a food products manufacturer sends researchers into supermarkets to find out the prices of competing brands or how much shef space and dispay support retaiers give its brands. To conduct in-store research in Europe, for exampe, store checks, photo audits of sheves and store interviews must be schedued we in advance and need to be preceded by a fu round of introductions of the researchers to store management and personne. Mechanica observations are often used to measure TV viewership. Cash register scanners can be used to keep track of customer purchases and inventories. Observationa research can obtain information that peope are unwiing or unabe to provide. In some countries individuas may be reuctant to discuss persona habits or consumption. In such cases observation is the ony way to obtain the necessary information. In contrast, some things are simpy not observabe, such as feeings, attitudes and motives, or private behaviour. Long-term or infrequent behaviour is aso difficut to observe. Because of these imitations, researchers often use observation aong with other data coection methods. 164

15 Chapter 5 Goba marketing research Experiments Experiments gather casua information. They invove seecting matched groups of subjects, giving them different treatments, controing unreated factors and checking for differences in group responses. Thus experimenta research tries to expain causeand-effect reationships. The most used marketing research appication of experiments is in test marketing. This is a research technique in which a product under study is paced on sae in one or more seected ocaities or areas, and its reception by consumers and the trade is observed, recorded and anaysed. In order to isoate, for exampe, the saes effects of advertising campaigns, it is necessary to use reativey sef-contained marketing areas as test markets. Performance in these test markets gives some indication of the performance to be expected when the product goes into genera distribution. However, experiments are difficut to impement in goba marketing research. The researcher faces the task of designing an experiment in which most variabes are hed constant or are comparabe across cutures. To do so represents a major chaenge. For exampe, an experiment that intends to determine a casua effect within the distribution system of one country may be difficut to transfer to another country where the distribution system is different. As a resut experiments are used ony rarey, even though their potentia vaue to the internationa market researcher is recognized. Surveys The survey research method is based on the questioning of respondents and represents, both in voume and in vaue terms, perhaps the most important method of coecting data. Typicay the questioning is structured: a forma questionnaire is prepared and the questions are asked in a prearranged order. The questions may be asked verbay, in writing or via a computer. Survey research is used for a variety of marketing issues, incuding the foowing: customer attitudes; customer buying habits; potentia market size; market trends. Unike experimenta research, survey research is usuay aimed at generating descriptive rather than casua data. Unike observationa research, survey research usuay invoves the respondent. Because of the importance and diversity of survey research in goba marketing, it is on this particuar aspect that we now concentrate. Contact methods The method of contact chosen is usuay a baance between speed, degree of accuracy and costs. In principe there are three possibiities when choosing a contact method: persona (face-to-face) interviews, teephone interviews and mai surveys. Each method has its own strengths and weaknesses. Tabe 5.3 gives an overview of these. Mai Mai surveys are among the east expensive. The questionnaire can incude pictures something that is not possibe over the phone. Mai surveys aow the respondent to answer at their eisure, rather than at the often inconvenient moment they are contacted for a phone or persona interview. For this reason, they are not considered as intrusive as other kinds of interviews. However, mai surveys take onger than other 165

16 Part II Deciding which markets to enter Tabe 5.3 Strengths and weaknesses of the three contact methods Questions/questionnaire Mai Internet/e-mai Teephone Persona Fexibiity (abiity to carify probems) Poor Fair Good Exceent Possibiity of in-depth information Fair Poor Fair Exceent (use of open-ended questions) Use of visua aids Good Exceent Poor Good Possibiity of a widey dispersed sampe Exceent Exceent Exceent Fair Response rates Poor Fair Good Fair Asking sensitive questions Good Poor Poor Fair Contro of interviewer effects Exceent Fair Fair Poor (no interviewer bias) Speed of data coection Poor Exceent Exceent Good Costs Good Exceent Exceent Poor kinds. You wi need to wait severa weeks after maiing out questionnaires before you can be sure that you have obtained most of the responses. In countries of ower educationa and iteracy eves, response rates to mai surveys are often too sma to be usefu. Internet/e-mai surveys These can coect a arge amount of data that can be quantified and coded into a computer. A ow research budget combined with a widey dispersed popuation may mean that there is no aternative to the mai/internet survey. E-mai surveys are both very economica and very fast. It is possibe to attach pictures and sound fies. However, many peope disike unsoicited e-mai even more than unsoicited reguar mai. Furthermore, it is difficut to generaize findings from an e-mai survey to the whoe popuation. Peope who have e-mai are different from those who do not, even when matched on demographic characteristics, such as age and gender. Teephone interviews In some ways these are somewhere between persona and mai surveys. They generay have a response rate higher than mai questionnaires but ower than face-to-face interviews, their cost is usuay ess than with persona interviews, and they aow a degree of fexibiity when interviewing. However, the use of visua aids is not possibe and there are imits to the number of questions that can be asked before respondents either terminate the interview or give quick (invaid) answers to speed up the process. With computer-aided teephone interviewing (CATI), centray ocated interviewers read questions from a computer monitor and input answers via the keyboard. Routing through the questionnaire is computer controed, heping the process of interviewing. Some research firms set up terminas in shopping centres, where respondents sit down at a termina, read questions from a screen and type their answers into the computer. Persona interviews Persona interviews take two forms individua and group interviewing. Individua interviewing invoves taking with peope in their homes or offices, in the street or in shopping arcades. The interviewer must gain the cooperation of the respondents. Group interviewing ( focus-group interviewing) consists of inviting six to ten peope to gather for a few hours with a trained moderator to tak about a product, service or 166

17 Chapter 5 Goba marketing research Samping pan A scheme outining the group (or groups) to be surveyed in a marketing research study, how many individuas are to be chosen for the survey, and on what basis this choice is made. organization. The moderator needs objectivity, knowedge of the subject and industry, and some understanding of group and consumer behaviour. The participants are normay paid a sma sum for attending. Persona interviewing is quite fexibe and can coect arge amounts of information. Trained interviewers can hod a respondent s attention for a ong time and can expain difficut questions. They can guide interviews, expore issues and probe as the situation requires. Interviewers can show subjects actua products, advertisements or packages, and observe reactions and behaviour. The main drawbacks of persona interviewing are the high costs and samping probems. Group interview studies usuay empoy sma sampe sizes to keep time and costs down, but it may be hard to generaize from the resuts. Because interviewers have more freedom in persona interviews the probem of interviewer bias is greater. Thus there is no best contact method it a depends on the situation. Sometimes it may even be appropriate to combine the methods. Samping pan Except in very restricted markets it is both impractica and too expensive for a researcher to contact a the peope who coud have some reevance to the research probem. This tota number is known statisticay as the universe or popuation. In marketing terms, it comprises the tota number of actua and potentia users/ customers of a particuar product or service. The popuation can aso be defined in terms of eements and samping units. Suppose that a ipstick manufacturer wants to assess consumer response to a new ine of ipsticks and wants to sampe femaes over 15 years of age. It may be possibe to sampe femaes of this age directy, in which case a samping unit woud be the same as an eement. Aternativey, househods might be samped and a femaes over 15 in each seected househod interviewed. Here the samping unit is the househod, and the eement is a femae over 15 years od. What is usuay done in practice is to contact a seected group of consumers/ customers to be representative of the entire popuation. The tota number of consumers who coud be interviewed is known as the sampe frame, whie the number of peope who are actuay interviewed is known as the sampe. Samping procedure There are severa kinds of samping procedure, with probabiity and non-probabiity samping being the two major categories: Probabiity samping. Here it is possibe to specify in advance the chance that each eement in the popuation wi have of being incuded in a sampe, athough there is not necessariy an equa probabiity for each eement. Exampes are simpe random samping, systematic samping, stratified samping and custer samping (see Mahotra (1993) for more information). Non-probabiity samping. Here it is not possibe to determine the above-mentioned probabiity or to estimate the samping error. These procedures rey on the persona judgement of the researcher. Exampes are convenience samping, quota samping and snowba samping (see Mahotra (1993) for more information). Given the disadvantages of non-probabiity sampes (resuts are not projectabe to the tota popuation, and samping error cannot be computed) one may wonder why they are used so frequenty by marketing researchers. The reasons reate to the inherent advantages of non-probabiity samping: 167

18 Part II Deciding which markets to enter Non-probabiity sampes cost ess than probabiity sampes. If accuracy is not critica non-probabiity samping may have considerabe appea. Non-probabiity samping can be conducted more quicky than probabiity samping. Non-probabiity samping, if executed propery, can produce sampes of the popuation that are reasonaby representative (e.g. by use of quota samping) (Mahotra, 1993, p. 359). Sampe size Once we have chosen the samping procedure the next step is to determine the appropriate sampe size. Determining the sampe size is a compex decision and invoves financia, statistica and manageria considerations. Other things being equa the arger the sampe, the ess the samping error. However, arger sampes cost more money, and the resources (money and time) avaiabe for a particuar research project are aways imited. In addition the cost of arger sampes tends to increase on a inear basis, whereas the eve of samping error decreases at a rate ony equa to the square root of the reative increase in sampe size. For exampe, if sampe size is quadruped data coection costs wi be quadruped too, but the eve of samping error wi be reduced by ony onehaf. Among the methods for determining the sampe size are the foowing: Traditiona statistica techniques (assuming the standard norma distribution). Budget avaiabe. Athough seemingy unscientific this is a fact of ife in a business environment, based on the budgeting of financia resources. This approach forces the researcher to consider carefuy the vaue of information in reation to its cost. Rues of thumb. The justification for a specified sampe size may boi down to a gut feeing that this is an appropriate sampe size, or it may be a resut of common practice in the particuar industry. Number of subgroups to be anaysed. Generay speaking the more subgroups that need to be anaysed, the arger the required tota sampe size. In transnationa market research, samping procedures become a rather compicated matter. Ideay a researcher wants to use the same samping method for a countries in order to maintain consistency. Samping desirabiity, however, often gives way to practicaity and fexibiity. Samping procedures may have to vary across countries in order to ensure reasonabe comparabiity of nationa groups. Thus the reevance of a samping method depends on whether it wi yied a sampe that is representative of a target group in a certain country, and on whether comparabe sampes can be obtained from simiar groups in different countries. Contact medium/measurement instrument Designing the questionnaire A good questionnaire cannot be designed unti the precise information requirements are known. It is the vehice whereby the research objectives are transated into specific questions. The type of information sought, and the type of respondents to be researched, wi have a bearing upon the contact method to be used, and this in turn wi infuence whether the questionnaire is reativey unstructured (with open-ended questions), aimed at depth interviewing, or reativey structured (with cosed-ended questions) for on the street interviews. In cross-cutura studies open-ended questions appear usefu because they may hep to identify the frame of reference of the respondents. Another issue is the choice between direct and indirect questions. Societies have different degrees of sensitivity to 168

19 Chapter 5 Goba marketing research certain questions. Questions reated to the income or age of the respondent may be accepted differenty in different countries. Thus the researcher must be sure that the questions are cuturay acceptabe. This may mean that questions, which can be asked directy in some societies, wi have to be asked indirecty in others. Formuation (wording) of questions Once the researcher has decided on specific types of questions the next task is the actua writing of the questions. Four genera guideines are usefu to bear in mind during the wording and sequencing of each question: The wording must be cear. For exampe, try to avoid two questions in one. Seect words so as to avoid biasing the respondent. For exampe, try to avoid eading questions. Consider the abiity of the respondent to answer the question. For exampe, asking respondents about a brand or store that they have never encountered creates a probem. Since respondents may be forgetfu, time periods shoud be reativey short. For exampe: Did you purchase one or more coa(s) within the ast week? Consider the wiingness of the respondent to answer the question. Embarrassing topics that dea with things such as borrowing money, sexua activities and crimina records must be deat with carefuy. One technique is to ask the question in the third person or to state that the behaviour or attitude is not unusua prior to asking the question. For exampe: Miions of peope suffer from hemorrhoids. Do you or does any member of your famiy suffer from this probem? It is aso a feasibe soution to ask about embarrassing topics at the end of the interview. The impact of anguage and cuture is of particuar importance when wording questions. The goa for the goba marketing researcher shoud be to ensure that the potentia for misunderstandings and misinterpretations of spoken or written words is minimized. Both anguage and cutura differences make this issue an extremey sensitive one in the goba marketing research process. In many countries different anguages are spoken in different areas. In Switzerand German is used in some areas and French and Itaian in others. And the meaning of words often differs from country to country. For exampe, in the United States the concept of famiy generay refers ony to the parents and chidren. In the southern part of Europe, the Midde East and many Latin countries it may aso incude grandparents, unces, aunts, cousins and so forth. When finay evauating the questionnaire, the foowing items shoud be considered: Is a certain question necessary? The phrase It woud be nice to know is often heard, but each question shoud either serve a purpose or be omitted. Is the questionnaire too ong? Wi the questions achieve the survey objectives? Pretesting No matter how comfortabe and experienced the researcher is in internationa research activities, an instrument shoud aways be pretested. Ideay such a pretest is carried out with a subset of the popuation under study, but a pretest shoud at east be conducted with knowedgeabe experts and/or individuas. The pretest shoud aso be conducted in the same mode as the fina interview. If the study is to be on the street or in the shopping arcade, then the pretest shoud be the same. Even though a pretest may mean time deays and additiona cost the risks of poor research are simpy too great for this process to be omitted. 169

20 Part II Deciding which markets to enter Data coection The goba marketing researcher must check that the data are gathered correcty, efficienty and at a reasonabe cost. The market researcher has to estabish the parameters under which the research is conducted. Without cear instructions the interviews may be conducted in different ways by different interviewers. Therefore the interviewers have to be instructed about the nature of the study, start and competion time, and samping methodoogy. Sometimes a sampe interview is incuded with detaied information on probing and quotas. Spot checks on these administration procedures are vita to ensure reasonabe data quaity. Data anaysis and interpretation Once data have been coected the fina steps are the anaysis and interpretation of findings in the ight of the stated probem. Anaysing data from cross-country studies cas for substantia creativity as we as skepticism. Not ony are data often imited, but frequenty resuts are significanty infuenced by cutura differences. This suggests that there is a need for propery trained oca personne to function as supervisors and interviewers; aternativey internationa market researchers require substantia advice from knowedgeabe oca research firms that can aso take care of the actua coection of data. Athough data in cross-country anayses are often of a quaitative nature the researcher shoud, of course, use the best and most appropriate toos avaiabe for anaysis. On the other hand, internationa researchers shoud be cautioned against using overy sophisticated toos for unsophisticated data. Even the best of toos wi not improve data quaity. The quaity of data must be matched with the quaity of the research toos. Probems with using primary research Most probems in coecting primary data in internationa marketing research stem from cutura differences among countries, and range from the inabiity of respondents to communicate their opinions to inadequacies in questionnaire transation (Cateora et a., 2000). Samping in fied surveys The greatest probem of samping stems from the ack of adequate demographic data and avaiabe ists from which to draw meaningfu sampes. For exampe, in many South American and Asian cities street maps are unavaiabe, and streets are neither identified nor houses numbered. In Saudi Arabia, the difficuties with probabiity samping is so acute that non-probabiistic samping becomes a necessary evi. Some of the probems in drawing a random sampe incude: no officiay recognized census of popuation; incompete and out-of-date teephone directories; no accurate maps of popuation centres, therefore no area sampes can be made. Furthermore, door-to-door interviewing in Saudi Arabia is iega. Non-response Non-response is the inabiity to reach seected eements in the sampe frame. As a resut opinions of some sampe eements are not obtained or propery represented. A good samping method can ony identify eements who shoud be seected; there is no guarantee that such eements wi ever be incuded. 170

21 Chapter 5 Goba marketing research The two main reasons for non-response errors are as foows: 1 Not being at home. In countries where maes are sti dominant in the abour force it may be difficut to contact a head of househod at home during working hours. Frequenty ony housewives or servants are at home during the day. 2 Refusa to respond. Cutura habits in many countries virtuay prohibit communication with a stranger, particuary for women. This is the case in the Midde East, much of the Mediterranean area and throughout most of South-East Asia in fact wherever strong traditiona societies persist. Moreover, in many societies such matters as preferences for hygienic products and food products are too persona to be shared with an outsider. For exampe, in many Latin American countries a woman may fee ashamed to tak with a researcher about her choice of brand of sanitary towe, or even hair shampoo or perfume. Respondents may aso suspect that the interviewers are agents of the government, seeking information for the imposition of additiona taxes. Finay, privacy is becoming a big issue in many countries: for exampe, in Japan the midde cass is showing increasing concern about the protection of persona information. Language barriers This probem area incudes the difficuty of exact transation that creates probems in eiciting the specific information desired and in interpreting the respondents answers. In some deveoping countries with ow iteracy rates written questionnaires are competey useess. Within some countries the probem of diaects and different anguages can make a nationa questionnaire survey impractica this is the case in India, which has 25 officia anguages. The obvious soution of having questionnaires prepared or reviewed by someone fuent in the anguage of the country is frequenty overooked. In order to find possibe transation errors marketers can use the technique of back transation, where the questionnaire is transated from one anguage to another, and then back again into the origina anguage. For exampe, if a questionnaire survey is going to be made in France, the Engish version is transated into French and then transated back to Engish by a different transator. The two Engish versions are then compared and, where there are differences, the transation is checked thoroughy. Measurement Reiabiity If the same phenomenon is measured repeatedy with the same measurement device and the resuts are simiar then the method is reiabe (the how dimension). Vaidity If the measurement method measures what it is supposed to measure, then it has high vaidity (the what dimension). There are three types of vaidity: construct, interna and externa. The best research design is useess without proper measurements. A measurement method that works satisfactoriy in one cuture may fai to achieve the intended purpose in another country. Specia care must therefore be taken to ensure the reiabiity and vaidity of the measurement method. In genera, how you measure refers to reiabiity and what you measure refers to vaidity. If we measure the same phenomenon over and over again with the same measurement device and we get simiar resuts then the method is reiabe. There are three types of vaidity: construct, interna and externa. Construct vaidity estabishes correct operationa measures for the concepts being studied. If a measurement method acks construct vaidity it is not measuring what it is supposed to. Interna vaidity estabishes a causa reationship, whereby certain conditions are shown to ead to other conditions. 171

22 Part II Deciding which markets to enter Externa vaidity is concerned with the possibe generaization of research resuts to other popuations. For exampe, high externa vaidity exists if research resuts obtained for a marketing probem in one country wi be appicabe to a simiar marketing probem in another country. If such a reationship exists it may be reevant to use the anaogy method for estimating market demand in different countries. Estimating by anaogy assumes, for exampe, that the demand for a product deveops in much the same way in countries that are simiar. The concepts of reiabiity and vaidity are iustrated in Figure 5.5. In the figure, the bu s eye is what the measurement device is supposed to hit. Situation 1 shows hoes a over the target, which coud be due to the use of a bad measurement device. If a measurement instrument is not reiabe there are no circumstances under which it can be vaid. However, just because an instrument is reiabe, the instrument is not automaticay vaid. We see this in situation 2, where the instrument is reiabe but is not measuring what it is supposed to measure. The shooter has a steady eye, but the sights are not adjusted propery. Situation 3 is the idea situation for the researcher to be in. The measurement method is both reiabe and vaid. An instrument proven to be reiabe and vaid in one country may not be so in another cuture. The same measurement scaes may have different reiabiities in different cutures because of various eves of consumers product knowedge. Therefore it may be dangerous simpy to compare resuts in cross-country research. One way to minimize the probem is to adapt measurement scaes to oca cutures by pretesting measures in each market of interest unti they show simiar and satisfactory eves of reiabiity. However, as different methods may have varying reiabiities in different countries, it is essentia that these differences can be taken into account in the design of a muticutura survey. Thus, a mai survey coud be most appropriate to use in country A and persona interviews in country B. In coecting data from different countries it is more important to use techniques with equivaent eves of reiabiity than to use the same techniques across countries. Figure 5.5 Iustrations of possibe reiabiity and vaidity situations in measurement Source: McDanie and Gates, 1993, p

23 Chapter 5 Goba marketing research 5.6 Onine (internet) primary research methods Athough the Internet is sti confined to the boundaries of the persona computer screen this wi soon be a thing of the past; it is now cear that the Internet is definitey going to be a medium for the masses. Many researchers are amazed at how efficienty surveys can be conducted, tabuated and anaysed on the Web. Additionay, onine data coection ets marketers use compex study designs once considered either too expensive or too cumbersome to execute via traditiona means. Whie initia forays were fraught with technica difficuties and methodoogica hurdes recent deveopments have begun to expose the medium s immense potentia. The eariest onine toos offered itte more than the abiity to depoy paper-based questionnaires to internet users. Today, however, onine toos and services are avaiabe with a wide range of features at a wide range of prices. For the internationa market researcher the major advantages and disadvantages of onine surveys are the foowing (Grossnicke and Raskin, 2001). Advantages of onine surveys Low financia resource impications: The scae of the onine survey is not associated with finance, i.e. arge-scae surveys do not require greater financia resources than sma surveys. Expenses reated to sef-administered posta surveys are usuay in the form of outward and return postage, photocopying, etc., none of which is associated with onine surveys. Short response time: Onine surveys aow questionnaires to be deivered instanty to their recipients, irrespective of their geographica ocation. Fast survey execution aows for most interviews to be competed within a week or so. Saving time with data coection and anaysis: The respective questionnaire can be programmed so that responses can feed automaticay into the data anaysis software (SPSS, SAS, Exce, etc.), thus saving time and resources associated with data entry process. Furthermore, this avoids associated data transcription errors. Visua stimui can be evauated, unike CATI. Disadvantages of onine surveys Respondents have no physica addresses: The major advantage of posta over onine surveys is that respondents have physica addresses, whereas yet not everyone has an eectronic address. Especiay is this an internationa marketing research probem in geographica areas where the penetration of the Internet is not as high as in Europe and North America. For cross-country surveys the mutimode approach (i.e. a combination of onine and posta survey) compensates for the misrepresentation of the genera popuation. Guarding respondents anonymity: Traditiona mai surveys have advantages in guarding respondents anonymity. Sensitive issues, which may prevent respondents from giving sincere answers, shoud be addressed via the post rather than onine. Time necessary to downoad pages: Probems may arise with oder browsers, which faied to propery dispay HTML questionnaires, with the appearance of the questionnaires in different browsers (Internet Exporer, Netscape). 173

24 Part II Deciding which markets to enter Onine quantitative market research (e-mai and web-based surveys) Onine surveys can be conducted through e-mai or they can be posted on the Web and the URL provided (a password is optiona depending on the nature of the research) to the respondents who have aready been approached. When a wide audience is targeted the survey can be designed as a pop-up survey, which woud appear as a web-based questionnaire in a browser window whie users are browsing the respective websites. Such a web-based survey is appropriate for a wide audience, where a the visitors to certain websites have an equa chance to enter the survey. However, the researcher s contro over respondents entering the web-based surveys is ower than for e-mai surveys. One advantage of web-based surveys is the better dispay of the questionnaire, whereas e-mai software sti suffers from certain imitations in terms of design toos and offering interactive and cear presentation. However, these two modes of survey may be mixed, combining the advantages of each (Iieva et a., 2002). Onine quaitative market research There are many interesting opportunities to conduct internationa quaitative market research quicky and at reativey ow cost, without too much traveing invoved (Scho et a., 2002). Saving money on traveing costs, etc: Many quaitative researchers often have to trave to countries in which research is conducted, briefing oca moderators and viewing some groups or hoding interviews to get a grasp of the oca habits and attitudes. This eads to high traveing costs and increases the time needed to execute the fiedwork. It usuay takes one or two weeks to recruit the respondents, and one or two weeks before the anaysis can start. In onine research the respondents can be recruited and interviewed from any computer anywhere in the word. Neary everyone who is connected to the Internet knows how to use chat rooms and they speak Engish. Fiedwork may start two days after briefing, and the anaysis may start right after the ast interview on the basis of compete and accurate transcripts, with each comment inked to the respective respondent. Cross-country quaitative research: Internationa onine research is particuary interesting for mutinationa companies that se their products on a goba scae and are afraid to buid the goba marketing strategy on research which has been conducted in ony a few of these countries. Onine quaitative research coud serve as an additiona muticountry check. This is not intended to give insight into the psychoogy of customers but rather to check whether other countries or cutures may add to the genera picture, which has been made on the basis of quaitative face-to-face research. One of the imitations with, for exampe, onine focus groups is that they seem to generate ess interaction between members than the face-to-face groups. Discussions between respondents occur, but they are ess cear and coherent. 5.7 Other types of marketing research A distinction is made between ad hoc and continuous research. Ad hoc research An ad hoc study focuses on a specific marketing probem and coects data at one point in time from one sampe of respondents. Exampes of ad hoc studies are usage 174

25 Chapter 5 Goba marketing research and attitude surveys, and product and concept tests via custom-designed or muticient studies. More genera marketing probems (e.g. tota market estimates for product groups) may be examined by using Dephi studies (see beow). Custom-designed studies These are based on the specific needs of the cient. The research design is based on the research brief given to the marketing research agency or interna marketing researcher. Because they are taior-made such surveys can be expensive. Muticient studies These are a reativey ow-cost way for a company to answer specific questions without embarking on its own primary research. There are two types of muticient study: 1 Independent research studies. These are carried out totay independenty by research companies (e.g. Frost and Suivan) and then offered for sae. 2 Omnibus studies. Here a research agency wi target specified segments in a particuar foreign market and companies wi buy questions in the survey. Consequenty interviews (usuay face to face or by teephone) may cover many topics. Cients wi then receive an anaysis of the questions purchased. For omnibus studies to be of use the researcher must have ceary defined research needs and a corresponding target segment in order to obtain meaningfu information. Dephi studies This type of research approach ceary aims at quaitative rather than quantitative measures by aggregating the information of a group of experts. It seeks to obtain answers from those who possess particuar in-depth expertise instead of seeking the average responses of many with ony imited knowedge. The area of concern may be future deveopments in the internationa trading environment or ong-term forecasts for market penetration of new products. Typicay key informants are seected and asked to identify the major issues in the area of concern. They are aso requested to rank their statements according to importance and expain the rationae behind the ranking. Next the aggregated information is returned to a participants, who are encouraged to state ceary their agreements or disagreements with the various rank orders and comments. Statements can be chaenged and then, in another round, participants can respond to the chaenges. After severa rounds of chaenge and response a reasonaby coherent consensus is deveoped. One drawback of the technique is that it requires severa steps, and therefore months may eapse before the information is obtained. However the emergence of e-mai may acceerate the process. If done propery the Dephi method can provide insightfu forecast data for the internationa information system of the firm. Continuous research (ongitudina designs) A ongitudina design differs from ad hoc research in that the sampe or pane remains the same over time. In this way a ongitudina study provides a series of pictures that give an in-depth view of deveopments taking pace. The pane consists of a sampe of respondents who have agreed to provide information at specified intervas over an extended period. There are two major types of pane: Consumer panes. These provide information on their purchases over time. For exampe, a grocery pane woud record the brands, pack sizes, prices and stores used 175

26 Part II Deciding which markets to enter for a wide range of supermarket brands. By using the same househods over a period of time measures of brand oyaty and switching can be achieved, together with a demographic profie of the type of person or househod who buys particuar brands. Retaier panes. By gaining the cooperation of retai outets (e.g. supermarkets) saes of brands can be measured by aser scanning the bar codes on goods as they pass through the checkout. Athough brand oyaty and switching cannot be measured in this way retai audits can provide accurate assessments of saes achieved by store. A major provider of retai data is the A.C. Niesen Company. Saes forecasting A company can forecast its saes either by forecasting the market saes (caed market forecasting) and then determining what share of this wi accrue to the company or by forecasting the company s saes directy. Techniques for doing this are deat with ater in the chapter. The point is that panners are ony interested in forecasts when the forecast comes down to individua products in the company. We sha now examine the appicabiity and usefuness of the short-, medium- and ong-term forecasts in so far as company panners are concerned and sha then ook at each from individua company departmenta viewpoints. Short-term forecasts. These are usuay for periods up to three months ahead, and as such are reay of use for tactica matters such as production panning. The genera trend of saes is ess important here than short-term fuctuations. Medium-term forecasts. These have direct impications for panners. They are of most importance in the area of business budgeting, the starting point for which is the saes forecast. Thus if the saes forecast is incorrect the entire budget is incorrect. If the forecast is over-optimistic then the company wi have unsod stocks, which must be financed out of working capita. If the forecast is pessimistic then the firm may miss out on marketing opportunities because it is not geared up to produce the extra goods required by the market. More to the point is that when forecasting is eft to accountants they wi tend to err on the conservative side and wi produce a forecast that is ess than actua saes, the impications of which have just been described. This serves to re-emphasise the point that saes forecasting is the responsibiity of the saes manager. Such medium-term forecasts are normay for one year ahead. Long-term forecasts. These are usuay for periods of three years or more depending on the type of industry being considered. In industries such as computers three years is considered ong term, whereas for stee manufacture ten years is a ong-term horizon. Long-term forecasts are worked out from macro-environmenta factors such as government poicy, economic trends, etc. Such forecasts are needed mainy by financia accountants for ong-term resource impications, but such matters of course are boards of directors concerns. The board must decide what its poicy is to be in estabishing the eves of production needed to meet the forecast demand; such decisions might mean the construction of a new factory and the training of a workforce. Forecasts can be produced for different horizons, starting at an internationa eve and then ranging down to nationa eves, by industry and then by company eves unti we reach individua product-by-product forecasts. This is then broken down seasonay over the time span of the forecasting period, and geographicay right down to individua saesperson areas. It is these atter eves that are of specific interest to saes management, or it is from this eve of forecasting that the saes budgeting and remuneration system stems. Figure 5.6 shows an exampe of trend forecasting. 176

27 Chapter 5 Goba marketing research Figure 5.6 An exampe of trend forecasting The unit saes and trend are drawn in as in Figure 5.6. The trend ine is extended by sight (and it is here that the forecaster s ski and intuition must come in). The deviations from trend are then appied to the trend ine, and this provides the saes forecast. In this particuar exampe it can be seen that the trend ine has been extended sowy upwards, simiar to previous years. The technique, as with many simiar techniques, suffers from the fact that downturns and upturns cannot be predicted, and such data must be subjectivey entered by the forecaster through manipuation of the extension to the trend ine. Scenarios Stories about pausibe aternative futures. Convergent forces Factors driving deveopments in the same direction. Divergent forces Forces driving deveopments apart from each other. Scenario panning Scenarios are stories about pausibe aternative futures (Wright, 2005). They differ from forecasts in that they expore possibe futures rather than predict a singe point future. Figure 5.7 shows two different scenarios A and B where the outcome measured on two dimensions is infuenced by both convergent and divergent forces. Figure 5.7 shows that the diverging and converging factors have to be baanced. Time fows from the eft to the right. The courses of the scenarios pass through a number of time windows, each made up of the key dimensions the scenario writers want to highight. In Figure 5.7 two time windows are shown: One in two years from now and another one in five years from now. The two dimensions coud be e.g. wordwide market share and wordwide market growth for one of the company s main products. The convergent forces woud mean that Scenario A and B woud come nearer to each other over time. The divergent forces woud have the opposite effect. Exampes of convergent forces woud be: high degree of macroeconomic stabiity in key internationa markets; increasing standardization of products across borders. An exampe of a divergent force woud be cutura diversity among target markets. Scenario panning aows us to consider a range of aternative futures, each of which is dramaticay different from the other and from the current operating environment. Rather than rey on a singe most ikey forecast it is possibe to compare and contrast aternative opinions on how your industry may evove. 177

28 Part II Deciding which markets to enter Figure 5.7 Deveopment of scenarios A and B over time Since it is externay oriented scenario panning is very effective at identifying growth strategies for the company as we as potentia threats to its market position. Scenarios can aso hep to identify the specific externa industry changes that are causing faing market share or margins. Guideines for scenario panning Estabish a core panning team. Anaysing the strategic impications of scenarios is best done in teams. The creative dynamics of an effective group are ikey to provide the types of breakthrough that wi make the scenario process worthwhie. What seems obvious to one person wi be surprising to another. A good rue of thumb is to have five to eight peope in the panning group. Get a cross-section of expertise. Incude the heads of a functiona areas saes, marketing, operations, purchasing, information technoogy, personne, etc. We aso recommend incuding individuas beyond the top executives. This injects new perspectives on your company or your ine of trade. This is a great time to invove the rising stars and innovative thinkers in the organization. Incude outside information and outside peope. Focus on injecting interesting and chaenging perspectives into the discussion. In a group composed soey of insiders it wi be hard to achieve breakthrough insights. Outsiders may be customers, suppiers or consutants. If possibe, invove an executive from another ine of trade or even from outside whoesae distribution. However, many executives fee uncomfortabe etting outsiders participate in the panning process of their companies. 5.8 Setting up an internationa MIS Once research has been conducted, the data coected and anaysed, the next step is to incorporate this information into management decision making. More and more businesses are now concerned with increasing the productivity of their marketing efforts, especiay in their marketing research departments. 178

29 Chapter 5 Goba marketing research Figure 5.8 Internationa marketing information system Source: Schmidt and Hoensen (2006), p Internationa marketing information system An interacting organization of peope, systems, and processes devised to create a reguar, continuous fow in information essentia to the internationa marketer s probemsoving and decisionmaking activities around the word. A massive amount of data is avaiabe from a wide variety of sources. The trick is to transform that data, ranging from statistics and facts to opinions and predictions, into information that is usefu to the organization s marketing decision makers. The importance of a timey and comprehensive information system is becoming more evident with the increased need to deveop coser customer reationships, the increasing costs of making wrong marketing decisions, the greater compexity of the marketpace, and the eevated eve of competitor aggressiveness. The need for current and reevant knowedge may resut in the deveopment and impementation of information systems that incorporate data management procedures invoving generating new data or gathering existing data, storing and retrieving data, processing data into usefu information, and disseminating information to those individuas who need it. The internationa marketing information system is an interacting organization of peope, systems and processes devised to create a reguar, continuous and ordery fow of information essentia to the marketer s probem-soving and decision-making activities. As a panned, sequentia fow of information taiored to the needs of a particuar marketing manager, the internationa MIS can be conceptuaized as a four-stage process consisting of ocating, gathering, processing and utiizing information. Figure 5.8 iustrates the centra issues to be addressed in each of the four internationa MIS-stages. In this rather compete internationa MIS mode, input data fow into the system from three major sources: the micro-environment, the macro-environment and from functiona areas of the firm. The output information wi then be made avaiabe to management for anaysis, panning, impementation and contro purposes. The proposed mode meets the exigencies of the ever-expanding roe of the MIS professiona that has to provide timey, accurate and objective information for management to be abe to navigate its way through the compex and fast-changing word of business gobaization. Against the backdrop of a dynamic business environment, companies are increasingy deveoping their marketing information systems to provide managers 179

30 Part II Deciding which markets to enter with rea-time market information. Likewise, they are expanding from oca to nationa to goba operations whie consumers are becoming ever more seective in their product choices. 5.9 Summary The basic objective of the goba marketing research function is to provide management with reevant information for more accurate decision making. The objective is the same for both domestic and goba marketing. However, goba marketing is more compex because of the difficuty of gathering information about mutipe and different foreign environments. In this chapter, specia attention has been given to the information coection process and the use of marketing information. This coverage is far from being exhaustive, and the reader shoud consut marketing research textbooks for specific detais reated to particuar research topics. An internationa marketer shoud initiate research by searching first for any reevant secondary data. Typicay a great dea of information is aready avaiabe, and the researcher needs to know how to identify and ocate the internationa sources of secondary data. If it is necessary to gather primary data the internationa marketer shoud be aware that it is simpy not possibe to repicate esewhere the methodoogy used in one country. Some adaptation of the research method to different countries is usuay necessary. The firm shoud set up a decision support system or an internationa market information system (MIS) to hande the gathered information efficienty. This system shoud integrate a information inputs, both interna and externa. In addition, an internationa MIS can support managers in their marketing decision making by providing interinkage and integration between functiona departments or internationa divisions. However, in the fina anaysis, every internationa marketer shoud keep in mind that an information system is no substitute for sound judgement. CASE STUDY 5.1 Teepack Speziamaschinen GmbH: Organizing a goba survey of customer satisfaction Teepack ( is a speciaized manufacturer of tea bag machines for the word s bestknown brands of tea and herbs and fruit teas, such as Lipton, Pickwick, Twinings and Lyons/Tetey. Teepack is a sister company of Teekanne, the eading tea, herb, and fruit tea packing company in Germany, with the Teefix, Pompadour and Teekanne brands. (The Teekanne Group has production and saes subsidiaries in severa countries. There are about 1,300 empoyees in the Group). The invention of the automatic tea bag-packaging machine by Teepack in 1949 revoutionized the tea market with the doube-chamber tea bag. It meant that production voumes coud be increased dramaticay. Today the atest generation of these machines is capabe of production speeds of amost 400 tea bags per minute, i.e. some 4 biion per year. The tea bag produced on Teepack machines is the most sod doube-chamber tea bag in the word. Important benefits are that it has consideraby arger space between the two bag chambers and offers maximum tea bag stabiity and durabiity without adding gue or heat seaing. 180

31 Chapter 5 Goba marketing research than 2,000 of its packaging machine Constanta. Thanks to Teepack s packaging machines Lipton is the market eader of the internationa tea market. Up to 1957 Teepack had sod more than 100 tea bag packaging machines in the United States. Technica innovation resuted in Teepack engineers deveoping a new, even more efficient machine Perfecta. Since 1990 more than 200 Perfecta machines have been sod wordwide. Today Teepack has a market share of about 70 per cent of the goba doube chamber tea bag machine market. The popuarity of this practica tea bag has continued to grow, for exampe, in Germany 82 per cent of tea saes are in doube-chamber tea bags; in the United States the figure is about 90 per cent and in Europe, if you omit the United Kingdom, the figure is cose to 100 per cent. Even in the former UK coony, Austraia, the doube-chamber tea bag has amost convinced the consumers. Down under, saes of UK tea bags and the doube-chamber tea bag more or ess baance themseves out. For over 50 years Teepack GmbH has been the number one producer of doube-chamber tea bag packaging machines in the word and has sod more Questions (Pease visit before you answer the questions.) 1 How woud you forecast wordwide demand for tea bag machines? 2 Argue the case for the market anaysis method you woud choose if you had to evauate the competitiveness of Teepack Speziamaschinen on the goba tea bag packaging machine market. 3 In order to achieve better customer feedback, the top management of Teepack is interested in earning how to measure customer satisfaction. Propose a questionnaire design that contains some of the themes which it woud be reevant to incude in the questionnaire. 4 How woud you organise the interna database with the customers responses and the feedback of the questionnaire resuts to the customers? CASE STUDY 5.2 Tchibo: Expanding the coffee shops business system in the United Kingdom and the rest of Europe Tchibo Frisch-Röst-Kaffee GmbH (Hamburg, Germany) was founded in 1949 by Max Herz. Tchibo was originay set up as a mai order company. At that time Tchibo sent coffee by post. The origina mai-order coffee company has grown into a mutinationa enterprise, active in many more sectors than just traditiona coffee retaiing. For exampe, at the end of 2003 Tchibo was one of the top two onine shops in Germany. The first Tchibo speciaist coffee shop with coffee counter service opened in Hamburg in The idea was that customers woud have the chance to try the coffee before they bought a whoe packet. This idea has been consistenty deveoped ever since. The retaiing concept typica of Tchibo combines saes of roasted coffee with counter saes of coffee speciaities, surrounded by an attractive merchandise word that changes every week. Tabe 1 shows the Tchibo coffee shops in Europe. Tchibo is market eader in the German, Austrian, Czech, Hungarian and Poish househod roasted coffee market with its coffee brands Tchibo, Gaa von Eduscho and oca brands. This success is party based on the systematic deveopment of a business Ë 181

32 Part II Deciding which markets to enter system, which combines Tchibo roasted coffee and coffee bar saes with a rich variety of innovative consumer merchandise and services. The number of Tchibo Coffee shops is now around 800. In comparison, Starbucks has around 8,500 coffee shops in 37 countries. The product range in the Tchibo coffee business system is being deveoped on a continua basis and expanded by offering innovative weeky changing new products. Tchibo s uniqueness is emphasized by the fact that not a products are offered at the same time but that the assortment changes 52 times a year. The motto A new experience every week enabes Tchibo to surprise its customers every Wednesday with introduction of a new theme, made of around 25 products. Tabe 1 Tchibo Coffee shops in Europe in 2005 Country Germany around 450 United Kingdom around 100 Switzerand around 50 Austria around 200 Poand around 50 Tota around 800 Source: Tchibo and other pubic sources. Number of coffee shops In the United Kingdom, Tchibo has successfuy opened about 100 coffee shops mainy in the Greater London area. Questions Tchibo is panning to expand its business system in the United Kingdom. However, in order to deveop the right promotion to the right customer group, Tchibo asks you as an internationa marketing consutant to answer the foowing questions. 1 Which market anaysis shoud be made in the United Kingdom in order to target the right promotion campaign to the right customer group? 2 How woud you estimate the potentia market for coffee shops (in genera) in Europe? 3 How wi you use market anaysis methods for estimating the possibe European market share of Tchibo coffee shops? VIDEO CASE STUDY 5.3 downoad from hoensen Burke Burke Inc. ( based in Cincinnati, was founded in 1931 and has been empoyee-owned since Burke is an independent, fu-service marketing research and decision-support company, which offers a broad range of decision-support services for marketing, operations, and quaity and human resources, through acquiring, integrating, anayzing and appying knowedge across the entire business enterprise. Questions 1 What are the key stages of the marketing research process and how does Burke reate to each? 2 Why is defining the research probem a crucia part of the research process? 3 Approximatey 70% of Burke s data coection is done through teephone surveys. What is your opinion of the future of this research methodoogy? 4 How is marketing research done in an internationa environment different from nationa marketing research? For further exercises and cases, see this book s website at 182