Consumers' preferences for processed meat: a best–worst scaling approach in three European countries

Processed meat products are a staple part of the typical European diet. Product packaging can include a considerable amount of information and, with other intrinsic and extrinsic attributes, substantially influence consumers' preferences and purchasing decisions. This study investigates 14 product attributes of processed meat products using a cross-country analysis. Based on an online survey conducted in Hungary (n = 410), Italy (n = 268), and Serbia (n = 402), an object-case best–worst scaling approach was applied. Results reveal both international and country-specific characteristics of preferences. Best-Worst scores reveal that taste and best-before date are among the most significant considerations in all three countries, while brand is among the attributes considered least important. Comparisons indicate significant differences according to country and socioeconomic characteristics. The study provides managerial implications.


Introduction
The consumption of animal-based proteins became more prominent in the twentieth century in developed countries (Bazoche et al. 2023).In many cultures, meat and meat products are an important part of the human diet (Sares-Jaske et al. 2022).Recent studies on consumers' preferences for processed meat products have identified a wide set of product attributes that might influence consumers (among others, Boncinelli et al. 2021;Hong et al. 2023;Török et al. 2022;Yeh and Hartmann 2021), indicating the complexity of attitudes toward this food product category.
As suggested by Mata et al. (2023), in Europe, there are great differences between consumers in terms of their consumption of processed meat.In our study, we analyzed patterns in consumers' attitudes toward processed meat products in a multinational context, including individuals with different European dietary backgrounds (from the Balkan, Eastern European, and Mediterranean regions).
For consumers, food quality is a multi-dimensional construct (Acebron and Dopico 2000); as such, it is hard to measure (Grunert et al. 1995).Like foods in general, processed meats have extrinsic and intrinsic product attributes that may influence their perceived quality level, and consumers often have difficulty evaluating the latter (Grunert et al. 2004).Extrinsic product attributes are related to the product (e.g., price level or brand).However, they are not a part of it in a physical sense, while intrinsic attributes are very specific and inseparable core characteristics (e.g., taste and nutritional value) (Brecic et al. 2017;Grunert et al. 2004;Malekpour et al. 2022).In our study, we investigate the stated importance of fourteen different product attributes -both extrinsic and intrinsic -that might influence consumers' attitudes toward processed meat products.The choice of all these attributes is supported by the literature and is introduced in the theoretical background chapter.
Consequently, this study aims to investigate consumers' perceptions of and preferences for processed meat products using the object case best-worst scaling (BWS) approach in three European countries.This methodology has often been used in recent food-related consumer studies (e.g., Carins et al. 2022;Chrysochou et al. 2022;Lerro et al. 2020;Rolfe et al. 2023;Thomson and Coates 2021); however, to the best of our knowledge, BWS studies have not investigated consumers' preferences for processed meat products in a multi-European-country context.Therefore, the contribution of our study to the literature is threefold.First, investigating consumers' preferences for processed meat, one of the most important protein sources in developed countries, is of utmost importance.We have included fourteen product attributes relevant to meat consumers in our survey, providing an opportunity for a broad assessment of their relevance.Second, our investigation of processed-meat-related preferences in a multi-national context is novel, providing comparable results for the three European countries with different dietary backgrounds.Previous studies mainly involved non-European and single-country approaches.Third, we applied BWS methodology in a new empirical context, enlarging the scope of this widely accepted and applied methodological approach.Our multi-dimensional segmentation of selected European countries and their consumer groups might serve as a reasonable basis for shaping the agenda of industry and policy stakeholders.The paper is structured as follows.After the literature review and validation of the selection of the attributes involved in the research, Sect."Methodology" describes the dataset that was constructed and the methodology that was applied for the BWS survey.Sect."Results" provides the results for all three countries; afterward, differences among countries and specific consumer groups are highlighted.The final section discusses the results in relation to prior literature and concludes.

Theoretical background
Previous studies have discovered that information concerning the country of origin should be taken into account during the assessment of the purchasing process of many traditional processed meat products, given that it plays an important role (Balogh et al. 2016;Resano et al. 2007).This claim is particularly valid for Europe, where the European Union system of geographical indications (GIs) clearly distinguishes between products of a distinct quality or reputation due to their origin (Bellassen et al. 2022).Although among countries and sociodemographic groups consumers' awareness of these GIs is quite heterogeneous (Verbeke et al. 2012), their role is increasingly relevant within the European economy as they account for an estimated 7% of the total sales value of the European food and drink sector (European Commission 2021) and processed meat products are a distinct product group of European GIs (Török and Moir 2018).Moreover, among the GI foods on the EU register, meat products are present in the second largest proportion (after cheeses) and are associated with one of the highest price premiums (Jantyik and Török 2020;Török et al. 2020).
Visual appearance (the color of the meat and the packaging, in particular) plays an important role for consumers in their selection of processed meat products since it influences perceptions of the product by defining "expectations, as they represent the first contact between an individual and a product" (de Almeida et al. 2017, p. 390).Consumers' preferences for colors of meat are often similar for specific products (e.g., cherry-red for beef); however, different packaging methods may influence consumers' acceptance (Grebitus et al. 2013).
For meat products, brand might be an obvious tool for differentiation and signaling superior quality (Grunert et al. 2004).Besides product differentiation, the producers' brand might be associated with a price premium, particularly for processed meat products; however, the growing share of retailers' private label brands might erode such premiums, putting additional pressure on manufacturers' brands (Ahmad and Anders 2012).
For food products, consumers often associate freshness with the best-before date (Samotyja and Sielicka-Różyńska 2020).Consequently, consumers' correct interpretation of this label impacts food waste and is relevant from a marketing perspective (Thompson et al. 2018).Compared to fresh meat, processed meat products have a longer shelf life (Schumann and Schmid 2018), and the food innovation of processed meat products involves significant emphasis on extending this (Hygreeva and Pandey 2016).
The price of processed meat products is a particularly important product characteristic, as food consumers are generally considered price-sensitive (Ackerman and Tellis 2001;Dhar and Hoch 1997).Price can therefore be considered one of the most important product attributes that is investigated in processed-meat-related studies (among others, see Ahmad and Anders 2012;Baba et al. 2016;de Araujo et al. 2022;Di Vita et al. 2019, 2022;Hung and Verbeke 2018;Hussein and Fraser 2018;Loginova and Irek 2022).
With increasing consumer awareness about nutritional content and interest in obtaining more information about the health benefits of food products, one of the challenges for the food industry is producing products with high nutritional value that also have desirable sensory properties and appearance (Amani et al. 2017).Nowadays, consumers are highly demanding regarding the health benefits of food and other products (Badar et al. 2021).The demand for high-quality meat products has increased, and to attract these consumers, the processed meat sector is intensely focused on developing low-fat and healthy meat products (Badar et al. 2021).Several adverse health-related features of processed meat products may be overcome through production reformulation -e.g., the reduction of unhealthy constituents, such as saturated fats, salt, and nitrites (Hung et al. 2016;Wolfer et al. 2018).The addition of natural antioxidants, strengthened with functional and health-promoting bioactive ingredients such as dietary fiber (Ranucci et al. 2018) are a current trend in the food industry and the subject of several scientific studies (Karre et al. 2013;Martínez et al. 2011).Consumer concerns about the health characteristics of processed meats have increased in recent years (Shan et al. 2017;Tobin et al. 2014).Concerning food products, it has also been reported that the use of health-positive framing elicits more robust responses from consumers than the use of health-loss-avoiding frames (Dolgopolova et al. 2022).In connection with this, consumers and the meat industry have become more aware and knowledgeable regarding the benefits of healthier diets (Mora-Gallego et al. 2016) partly due to advances in comprehending the relationship between diet and health (López-López et al. 2010).
In addition to the attributes listed above, we included traditional methods used in the production/processing of the product because, in the context of a changing food consumption culture, the interest of consumers in traditional foods has increased, particularly in European countries (Di Monaco and Cavella 2015).Traditionally processed foods constitute an important element of European culture, identity, and heritage, contributing to the development and sustainability of rural areas, involving substantial product differentiation opportunities for producers, and increasing the diversity of food choices for consumers (Guerrero et al. 2009;Iaccarino et al. 2006).Traditionally processed food is defined as food (products) produced or processed in a non-industrial environment that is traditional or unique, characterized by specific production methods with a limited degree of mechanization (Kupiec and Revell 1998); these can also be referred to as specific place and producer/production-related factors (Kuznesof et al. 1997).This description implies that traditional food involves wide diversity and a strong identity (Rason et al. 2007).However, studies that have explored consumer perceptions of reformulated processed meat products have produced contradictory results: some have concluded that consumers have positive perceptions of reformulated meat products (e.g., Hung et al. 2016;Schnettler et al. 2018), whereas others have concluded the opposite (Shan et al. 2017;Żakowska-Biemans et al. 2016).Additionally, as far as traditional food is concerned, innovation associated with this food category is greater for those products that have visible and relevant benefits, such as improving nutritional values (Grunert et al. 2011).
At the same time, increasing concerns for farm-animal welfare and the citizens' rising awareness of production methods have led some companies to invest in a new concept of sustainable meat production that includes animal-friendly production practices.For example, in the pork supply chain, this trend has affected several aspects of pig farming, including substituting traditional methods of castrating male pigs (used to prevent boar taint in meat) with other methods such as immunocastration (Mancini et al. 2017).
The increase in the demand for foods produced using organic methods includes greater demand for organic processed meat products (Sindelar et al. 2007).In this area, being defined as organic means avoiding the use of the chemicals traditionally used for preservation (Haugaard et al. 2014;Sullivan et al. 2012).However, consumer studies often reveal that the share of meat in the diets of typical organic consumers is smaller (Kesse-Guyot et al. 2013).
We also included the attribute GMO free, as recent public concerns about GMO foods have resulted in the promotion of the GMO-free product characteristic (Robinson and Leonhardt 2018).Although European and US GMO regulations are contested, and even among European consumers there is no positive or negative consensus about the desirability of GMO foods (Popek and Halagarda 2017).
Because meat contains large amounts of bio-compounds, consumers have a great passion for its taste (Amani et al. 2017).Consumer demand for convenience and good tasting food has ensured that processed meat remains a dietary staple (Grunert 2006).Concerning the sensory attributes of processed meat, the prominent role of taste in consumers' food choices is well-established and considered the most important sensory attribute (de Almeida et al. 2017;Shan et al. 2017).Sensory analyses of cured and processed meats have been carried out in relative depth and several approaches have been used to evaluate the quality parameters of the latter (Di Vita et al. 2017;Resurreccion 2004).Font-I-Furnols and Guerrero (2014), focusing on the sensory analysis of meat products, highlighted the importance of flavor intensity and saltiness.
One response of the food sector to the lower level of trust in food due to food-related scandals in recent decades might be the shortening of food supply chains.Relatedly, (personal) knowledge of the producer might increase confidence through face-to-face interactions (Renting et al. 2003).However, consumer attitudes toward artisan-type, locally produced meat products are yet to be covered in the literature.In addition, food consumers do not appear to value the various implications of farmer's ownership of food brands, so farmer-owned labels might not be an effective means of increasing trust (Grashuis and Su 2023).
The concept of fair trade is usually applied to commodity-like food products (like cocoa or coffee) produced by small-scale farmers.However, in the case of meat products, fair trade has been investigated in the context of game meat authenticity (Fajardo et al. 2010) and (the identification of) meat product adulteration (Jawla et al. 2021).
In addition to the country of origin, in our study, the product's region of origin refers to a closer relationship between the consumer and the product in terms of physical distance and cultural embeddedness.Previous research reported that meat products are one of the predominant product categories involved in local food activities (Ilbery et al. 2006), and consumers often consider meat processed by local small-scale operators safer (Telligman et al. 2017).

Methodology
An online cross-country questionnaire using the BWS method was developed to explore consumers' preferences associated with processed meat products.First and foremost, the aim was to elicit the relative importance that participants award to items/attributes when purchasing processed meat.BWS is an attribute-based method that has attracted research attention in the agricultural and health economics literature in recent decades (Erdem et al. 2012;Louviere and Flynn 2010;Merlino et al. 2018).In the present study, fourteen processed-meat-related attributes of particular relevance to consumers' processed-meat purchase decisions were defined and included based on a review of the relevant literature and intensive discussion with the academic researchers on the project team.Table 1 provides an overview of these attributes.
Regarding the experimental design of the BWS questions, 240 BWS choice scenarios were generated for this study.An orthogonal frequency balanced design was developed to maximize BWS design efficiency.To prevent respondent fatigue, the choice scenarios were divided into 40 blocks, with the respondents being required to complete six BWS choice sets displaying five attributes for each version per block.
Figure 1 shows an example of the BWS questions that were used in the online survey.Respondents were randomly assigned one of 40 versions of the questionnaire to complete and asked to select the processed meat attribute that they found the most and least important when purchasing processed meat.All BWS block variations consisted of the same content, but the questions contained different combinations of product attributes.All country-specific questionnaires were initially designed in English but translated by the participating researchers into their local languages using the back-translation procedure to ensure linguistic equivalence.The content of the survey and the translations were discussed and reflected on by the researchers in the respective countries.

Data collection
An online survey was conducted in the autumn of 2017 after pretesting using a nationwide online survey administered in three countries (Hungary, Italy, and Serbia) by a third-party contractor (Lightspeed Research Ltd.) using its consumer panel database.
To generate a more comprehensive overview, these countries were selected based on their heterogeneous food cultures (Central-European, Mediterranean, and Western-Balkan, respectively), with diets favoring processed meat products.A total of 1,488 individuals (HU = 505; IT = 488; RS = 495) participated in the study.We excluded from the survey those individuals who did not consume meat.

Data analysis
Our research applied a stated preference methodology using the BWS approach.This method investigates individual preferences in a hypothetical context by asking respondents to evaluate and choose the 'best' and the 'worst' options in multiple-choice situations, allowing the 'importances' of the attributes to be measured (Cohen 2003;Louviere et al. 2015).
According to the structure of the decision set and the level of complexity, three types of BWS can be distinguished: Case 1 (object case), Case 2 (profile case), and Case 3 (alternative case) (Flynn 2010).Our approach involved implementing Case 1 BWS, often called maximum difference scaling (although this name is misleading and not entirely accurate; see Marley and Louviere (2005)), which can be linked to the authors Finn and Louviere (1992).The initial step when employing Case 1 BWS is allocating attributes into decision sets, whose number should be defined according to the cognitive capacity of the respondents.Respondents are then asked to choose the attributes they consider the best and the worst of the options.This step is followed by data collection and assessing responses.The last step involves analyzing the dataset by calculating several statistical indicators (the 'counting approach') or modeling (Aizaki and Fogarty 2023).We used the so-called counting approach in our analyses, which is detailed below.
The survey dataset can be analyzed using statistical indicators both on an individual and aggregate level.First, the best-worst values for the individual (Eq. 1) and aggregate (Eq.2) levels are calculated, where n is the individual and k is the examined attribute.
The standardized form of these values is calculated using Eqs. 3 and 4.
where f denotes the frequency of attribute k that appears in the decision sets.

Results
In this section, our sample is first described, then the B-W scores are presented for the selected countries (Hungary, Italy, and Serbia).Following this, the B-W differences are analyzed among the selected countries.Finally, we describe our analysis of the B-W differences according to several sociodemographic aspects.

Description of the sample
After excluding incomplete questionnaires, a final sample of 1,080 meat consumers remained for analysis (see details in Table 2).In the samples, females and males participated equally, with an average age of 41-43 years.Non-rural respondents are the majority in all three samples who have completed at least upper secondary school.The main sample characteristics are reported in Table 2.

Best-Worst scores in the selected countries
The frequency of best and worst scores for the attributes is shown in Table 3.In addition, the standardized B-W scores are presented and interpreted using a country-level breakdown in Fig. 2. Table 3 reports the sum of the most important (column "total best"), the sum of the least important (column "total worst"), and the difference between the most and least important for each processed meat attribute (column "best-worst score") in the three countries.The mostleast difference for each attribute is then ranked (column "rank").The results indicate that the taste of processed meat products is among the attributes most highly ranked in all three countries, placed first in Hungary and Italy and second in Serbia.In contrast, best-before date is ranked second in Hungary and third in Italy and Serbia.For the Hungarians, product price is also an attribute of fundamental importance (rank 3).Simultaneously, GMO-free and animal welfare-friendly production methods play an important role for consumers in Serbia and Italy, respectively.Among the attributes considered least important is the brand in all three samples, together with the respondent's knowledge of the producer (in Hungary) and fair trade (in Italy and Serbia).In contrast to the other two countries, organic production methods are considered less important in Hungary, while product region of origin is rated lowest in importance in Serbia.
Standardized B-W scores were also calculated (Fig. 2).On the x-axis, we can see the attributes according to the countries under study, while the y-axis shows the standardized B-W scores for each country and attribute.It is important to note that the standardized Best-Worst scores range between − 1 and 1.
Although the results underline that the taste of the product and best-before date play a relevant role everywhere for processed meat products, country-specific differences exist; therefore, pairwise comparisons were calculated and are described in the following section.

Differences between countries
Several country-specific differences are evident in Fig. 2, showing how the importance of the selected attributes differs among the studied countries.The differences regarding the evaluation of the attributes between the three selected countries were tested and are presented in Table 4.The nonparametric Kruskal-Wallis test indicates significant nationallevel differences for all fourteen attributes.The product's country and region of origin are more important for Italian consumers than in Serbia and Hungary.Furthermore, the bestbefore date and price attributes are evaluated more highly in Hungary, while GMO-free and organic in Serbia.On the other hand, the respondent's knowledge of the producer is considered more important in Italy and Serbia than in Hungary.

Differences among consumer groups
In the last step of our analysis, we also applied the Kruskal-Wallis test to identify if sociodemographic characteristics (gender, age, residence, level of education, and income) significantly affected the evaluation of the importance of the attributes.All results are presented in the Appendix (Table 6).
Significant differences between consumer groups were also identified.Gender plays a significant role in all three selected countries, as females consider best-before date (Hungary and Serbia), animal-welfare-friendly production (Hungary and Italy), GMOfree production (Hungary), and country/region of origin (Serbia) more important than male consumers.On the other hand, for men, price (Hungary and Serbia), together with brand, taste, and fair trade (Serbia), seem to be more important.In addition, age was found to be a relevant characteristic in all three countries, as Hungarians younger than 30 consider fair trade to be important.However, older consumers evaluate animal-welfare-friendly production methods as more important, and middle-aged consumers (age 40-49) rank the product's visual appearance highly (both in Italy).Traditional methods used in the production/processing of the product are more important for older Serbian consumers.
In Hungary and Serbia, income status is also determinative.In Hungary, knowledge of the producer is more important for consumers with a mid-level income than those with a higher income, while in Serbia, the best-before date is valued more by less affluent consumers.
In Hungary, educational level significantly influences the attitude to fair trade, as this is considered important for consumers with a secondary and lower education level.Finally, place of residence was found to be relevant only in Italy: products' nutritional value was evaluated as more important for consumers living in urban settings.

Discussion
In developed countries, animal-based proteins in diets are of significant importance.However, the role of the daily consumption of meat products has come under scrutiny in recent years, mainly due to health and environmental concerns (Willett et al. 2019).Processed meat products are directly associated with unfavorable health and environmental impacts (Sares-Jaske et al. 2022).In 2015, the World Health Organization declared the excessive consumption of processed meat products "carcinogenic to humans" (WHO 2015), and from an environmental perspective, livestock production greatly contributes to greenhouse gas emissions and has other harmful effects on ecosystems (Clare et al. 2022).
Despite these recent trends, processed meat products are still an important part of the European diet (Cocking et al. 2020), and it is estimated that the demand for animalbased proteins will continue to increase globally (Willett et al. 2019).In response to health and sustainability concerns, plant-based proteins might represent a solution; however, consumer acceptance of these is still relatively low (Bazoche et al. 2023).In addition, cultured meat products are expected to become a substitute for traditional meat; however, several hurdles must be overcome before marketable products will be available (Broucke et al. 2023).Therefore, in the short term, the role of traditional processed meat products remains unquestioned.
This research explores the importance of product attributes associated with processed meat products in three European countries.The survey results of this research indicate that some general patterns are valid in all three countries.First, taste matters most for consumers everywhere in the purchase of processed meat products.Numerous pieces of earlier research reported similar conclusions.Among others, de Almeida et al. ( 2017) and Shan et al. (2017) for Brazilian and Irish meat consumers, respectively, while Malone and Lusk (2017), for a US sample, found that for meat product consumers, taste is of prominent importance.
Similarly to taste, best-before date also played an important role everywhere, first of all in Hungary.The role of best-before dates has recently been investigated in a food-wasterelated context (e.g., Li et al. 2020;Samotyja and Sielicka-Różyńska 2020); however, as a product attribute, this might also significantly influence consumer preferences (Aday and Yener 2014;Secondi 2019).Meat and processed meat products are usually considered perishable food (Genigeorgis 1986;Umaraw et al. 2020) with a limited shelf life; therefore, the freshness of these products is a crucial cue for consumers regarding their safety (Van Rijswijk et al. 2008).Our results also proved that the best-before date among Hungarian and Serbian consumers was significantly more important for females than males.Other studies have also found that females are much more liable than men to reject consuming expired processed meat products (e.g., Van Boxstael et al. (2014) in Belgium), and they check best-before dates more frequently, too (e.g., Achón et al. (2017) in Spain).
For processed meat products, brand might be an indicator of quality and a guarantee of authenticity and traceability (Bredahl 2004;Grunert et al. 2004), and eye-tracking experiments have proven that brand logos on packaged meat products are very attractive to consumers (Lombard et al. 2020).However, our results showed that brand was among the least important product characteristics, as noted by Rolfe et al. (2023).This also indicates that the level of brand loyalty was low in the selected three countries, as has been found for other countries (e.g., in Sweden-Lagerkvist 2013).It should be noted that for processed cured meat, geographical indications, such as Parma ham from Italy, are relevant attributes in terms of shaping product reputation; in such cases, a collective brand (e.g., the Consortium brand) is considered more important than a private/ company one (Arfini and Mancini 2015).
In parallel to the general patterns, we also identified national characteristics.Generally, price is the extrinsic attribute considered most important in relation to meat products (Davidson et al. 2003) and is often considered highly important (Lagerkvist 2013).This finding is particularly valid for Hungary: results indicate that Hungarian consumers are significantly more price-sensitive than those in Italy and Serbia.This finding aligns with previous national-level results showing that price is the main attribute Hungarian consumers evaluate when purchasing food (Szakály et al. 2014).Gender also determines attitude to price, as males usually pay more attention to prices than female consumers.
For Italians, animal welfare-friendly production methods were found to be more important than for consumers in other countries; females usually cared more about this characteristic.For Italian meat consumers, similar results were previously obtained (e.g., Napolitano et al. (2007) and Merlino et al. (2018) -both for beef meat), indicating that consumers consider animal welfare issues a great deal in Italy; this attribute is considered of similar importance to price.Our results also validated that animal welfare certification is more important for female Italian consumers than males (as in Hungary) (Blanc et al. 2020).Also, origin played a more dominant role for Italians, as both a product's country and its region of origin were considered of significantly greater importance.As noted before, the strong awareness of Italian consumers of geographical indications in general (Verbeke et al. 2012), as well as the relevance of regional and local GIs for processed meat purchase decisions in Italy, has been identified by several scholars (e.g., Arfini and Mancini 2015).
For Serbians, purchasing processed meat products that are GMO free and knowing the producer were significant characteristics, reflecting the latter's solid anti-GMO attitudes (Zdjelar et al. 2013) and the dominant position of small private farms in the meat product supply chain of the country (Zaric et al. 2011).
Our contribution to the literature is threefold.First, our research validates wider claims about a range of previously investigated processed meat attributes in a multicountry European environment.Similar investigations mainly focused on non-European countries (particularly the USA), following a single-country approach.Second, we applied BWS methodology, which has not been applied to this research topic, to highlight similarities and differences in the selected countries, allowing us to draw broader conclusions.Third, our multi-dimensional segmentation of consumers in selected European countries might serve as a reasonable basis for shaping the agenda of both industry and policy stakeholders.The attributes identified and analyzed herein are relevant in marketing the traditional processed meat products broadly available in European markets.Further, greater understanding of the relevance of these attributes might help comprehend consumers' attitudes toward processed meat products derived from alternative protein sources, which are currently subject to debate in Europe.
From an economic point of view, this study also has several implications.First, the importance to consumers of taste is unquestionable; therefore, meeting consumers' taste expectations is one of the keys to success.However, tastiness is an experience quality that varies not only between consumer segments but also between types of products.This puts significant pressure on producers and processors as product development in the food industry is complex and risky, with a substantial failure rate (Grunert et al. 2004).Second, highly valued best-before dates indicate that consumers also appreciate the extended shelf-life of processed meat products.Therefore, focusing on advanced packaging might have benefits in terms of profitability and food-waste-reduction (Soro et al. 2021).Third, the minimal importance of company brands suggests that a private labeling marketing approach to processed meats may be relevant, with all the associated benefits (larger retail margins and store loyalty, particularly (Riboldazzi et al. 2021)).Finally, consumers' national and socioeconomic characteristics (in all three countries) revealed that a uniform marketing approach should not be applied to consumers; focusing on a specific set of extrinsic attributes is required for developing and promoting meat products tailored to consumers (Bernues et al. 2003).
While the research has led to several robust findings, it also has some limitations.First, the survey was conducted in three European countries; therefore, no overall conclusions about European tendencies can be made.Another limitation is related to the structure of the samples.Due to the online data collection process, the samples are biased toward more educated, urban, and upper-middle-income respondents; however, the survey aimed at collecting data about meat consumers, about which population we need more information.Finally, the hypothetical nature of the experiment is another limitation of the research, as stated preferences were investigated in our approach.Introducing cheap talk at the beginning of the said preference exercise should have minimized hypothetical bias (Penn and Hu 2018).
The best-worst scaling methodology proved useful for investigating consumers' preferences for processed meat products in a cross-country analysis of three European countries.Results clearly identified dominant product attributes in all locations (taste and best-before date, in particular) and minor important characteristics (e.g., brand).Additional econometric tests identified significant differences among countries and several socioeconomic characteristics associated with processed meat product attributes.
These limitations might also serve as a basis for further investigations.Future research might expand the spatial scope (e.g., to involve countries from the Nordic, the Baltic, or Western-European states) or include other product attributes which shape consumers' attitudes toward processed meat products (such as alternative protein sources, in particular).

Conclusions
To the best of our knowledge, this study is the first to investigate consumer preferences toward processed meat products' attributes using the best-worse scaling approach in three European countries.Our study identifies general and country-specific characteristics, and thus might be useful from a managerial and policy perspective.Our research reveals new insights into this research field and contributes to studies on consumer preferences for processed meat products (de Almeida et al. 2017;Shan et al. 2017).In general, taste and best-before date were ranked as the most important attributes in all three countries, while brand was among those considered least important.The country-level comparisons revealed that Hungarians are very price-sensitive; Italians care about the product's country and region of origin and animal welfare-friendly production methods, while Serbians value GMO-free products and knowledge of the producer.The results also identified significant differences in socioeconomic characteristics: in general, for female consumers, best-before date and animal welfare are more important attributes, while for males, price matters more.

Appendix
See Tables 5 and 6.

Fig. 1
Fig. 1 Example of a BWS question (English-language version of survey deployed in Italian)

Table 1
Attributes examined in the object-case BWS questions

Table 2
Sample characteristics *Detailed income categories according to selected countries are presented in the appendix (Table5)

Table 3
Best-Worst scores in the selected countries *Darker background color refers to higher rank according to Best-Worst scores

Table 4
Differences between selected countries *Significant at a 5% level.**Significant at a 1% level.Bold indicates the country where the attribute was considered more important in the pairwise comparisons

Table 5
Household net income categoriesNational currencies of Hungary and Serbia are converted to euros at the exchange rate relevant to the time of the start of the survey

Table 6
Differences between consumer groups in Hungary, Italy and Serbia (Kruskal-Wallis tests)