Testing consumers ’ acceptance for an extra-virgin olive oil with a naturally increased content in polyphenols: The case of ultrasounds extraction

A B S T R A C T Innovation is fundamental for all agri-food companies to increase competitiveness. Being extra-virgin olive oil (EVOO) a traditional food product (TFP), the main obstacle to innovation is its traditional nature. This study evaluated consumers ’ acceptance for an EVOO with a naturally increased content of poliphenols, as it has been extracted through ultrasounds. This product has been compared with a set of emerging innovations that may be introduced in the next future. To this end, a choice experiment was carried out bent on the estimation of a Latent Class Model (LCM). A nationally-representative sample of EVOO consumers were involved in a web-based interview. The LCM analysis highlighted three segments of consumers: (1) innovative; (2) traditionalist; (3) cautious . Results showed that there is cluster of consumers willing to accept this innovation, therefore its introduction on the market appears to be possibly successful.


Introduction
Phenolic compounds are believed to play a key role in human longevity achieved in the modern era (Lutz, Fuentes, Ávila, Alarcón, & Palomo, 2019).Among other substances contained in foods constituting a healthy diet, phenolic compounds are believed to reduce risk factors for cardiovascular diseases and aging-related diseases (Dauchet, Amouyel, & Dallongeville, 2009;Wang et al., 2014).Although evidence on the effects on health played by phenolic compounds is not available for all the foods concerned, the European Food Safety Authority officially confirmed such healthy effects through approving the use of a health claim on extra-virgin olive oil (EVOO) related to its content in hydroxytyrosol and its derivatives, the latter being an element able to prevent oxidative and inflammatory processes (Caporaso et al., 2015;EFSA 2011;Peyrol, Riva, & Amiot, 2017;Roselli, Clodoveo, Corbo, & De Gennaro, 2017).
In this context, there is an innovation in EVOO production that appears to be of particular interest: extraction assisted by ultrasound.This innovation allows an improvement in the efficiency of the EVOO production process, representing a benefit for producers (Clodoveo, 2012).In addition, EVOO can be obtained which is higher in phenolic compounds compared to that extracted with traditional technology (Clodoveo et al., 2016).While phenolic compounds tend to give EVOO a bitter taste, in the above case this does not happen: the taste of the final product is lower in bitterness than traditional products (Clodoveo et al., 2017).
EVOO is conventionally categorized as a Traditional Food Product (TFP) 1 (Caputo, Sacchi, & Lagoudakis, 2018).Attitudes to innovation for this category of products appear conflicting, judging from mixed responses to consumer surveys (Guerrero et al., 2009;Vecchio, Lombardi, Cembalo, Caracciolo, & Cicia, 2016).Innovations that appear more successful are those conferring specific benefits on consumers, such as improvements in nutritional and health properties or convenience, without detracting from the traditional nature of the product (Almli et al., 2011).Furthermore, among consumers, there are some groups who show greater acceptance, such as those who display openness to new foods (Hersleth, Lengard, Verbeke, Guerrero, & Naes, 2011).https://doi.org/10.1016/j.jff.2020.103940In this study, we aimed to assess whether an innovation on a TFP, namely ultrasound-extracted EVOO, would be able to capture consumer acceptance on the market.Most innovative products, despite appearing potentially successful, fail when they actually arrive on the market (Dijksterhuis, 2016;van Kleef, van Trijp, & Luning, 2005).Innovation failure can have serious repercussions, wiping out investments and wasting valuable resources for the Italian EVOO industry which has undergone several adverse events in recent years (Cornara et al., 2017).Therefore, our study aimed to answer the following research question: Will consumers accept the innovative EVOO extracted with ultrasound technology, once it has come onto the market?
A previous exploratory investigation on ultrasound-extracted EVOO found a potentially positive response for the possibility of buying the product.This suggests that acceptance on the Italian market may be possible, although the response from consumers may depend on different individual traits, such as quality perception, sensory preferences and individual level of education (Roselli, Cicia, et al., 2018).In this context, this study aimed to take a step forward in understanding the topic.
For this purpose, a survey was set up with a choice experiment.The investigated innovation was compared to others in order to simulate the competition actually occurring on the market.Since convenience characteristics (e.g.easy preparation and handling of the product) have been highlighted as the key competitive advantage for innovative TFPs (Guerrero et al., 2009), we compared the acceptance of ultrasonically assisted extraction with that assisted by microwave technology, another technology able to improve the efficiency of the EVOO extraction process.Two innovations that extend EVOO shelf-life were also taken into account: nitrogen packaging and bag-in-box.
Other important attributes defining the quality of EVOO and its characteristics were then considered.In order to identify the most important individual traits of the consumers, according to their degree of acceptance, the survey also collected demographic and psychological traits of respondents.
The article is structured as follows: the theoretical framework is presented in Section 2, and the survey is described in Section 3; after the results in Section 4, the conclusions convey the main insights obtained.

Product attributes
In order to investigate the importance of the innovation according to consumers, we defined a set of attributes which constitute EVOO quality.The most important EVOO attributes were collected from the relevant literature on the topic in question (Del Giudice, Cavallo, Caracciolo, & Cicia, 2015).Together with these attributes, price was added, the levels being determined according to the prices as transmitted by IRI Infoscan data: the chosen prices were those at the 20th, 40th, 60th and 80th percentile of the whole distribution.This allowed us to represent the majority of prices that consumers actually encounter while shopping for groceries.
The first relevant attribute to be considered is product origin.It is a multi-faceted quality cue, linked to concepts such as typicality2 and ethnocentrism3 , embedding a set of other attributes, such as safety and traceability (Caporale, Policastro, Carlucci, & Monteleone, 2006;der Lans, Ivo, Van Ittersum, De Cicco, & Loseby, 2001;Menapace, Colson, Grebitus, & Facendola, 2011).Furthermore, it appears to be one of the main elements considered by consumers while shopping for EVOO (Cavallo, Caracciolo, Cicia, & Del Giudice, 2018;Del Giudice, Cavallo, Caracciolo, & Cicia, 2015).According to the existing legislation for the designation of origin on the EVOO label (EC Regulation No. 29/2012), three levels for the above attribute are used: "100% from Italy", "from EU countries" and "from extra-EU countries".
Among the attributes determining EVOO purchases, we also considered two certifications: organic and Protected Designation of Origin (PDO).Organic production is a method to obtain food and other agricultural products based on the use of processes with minimal impacts on the environment and the health of humans, plants and animals (EC Regulation No. 834/2007), while the PDO label, according to EC Regulation No. 1151/2012, is a geographical indication which can be used to differentiate food products originating from a specific place or region and whose quality is more attributable to its geographical origin.The aim of such labeling schemes is to reduce information asymmetry, signaling to consumers the presence of a bundle of quality characteristics, linked either to the region or area of origin (Aprile, Caputo, & Nayga, 2012), or to the organic production method (Liang, 2016).As a result, consumers may have higher quality expectations for certified products compared to products without such labels and usually show a higher willingness to pay for them (Caracciolo et al., 2018;Castriota & Delmastro, 2015;Cicia, Del Giudice, & Scarpa, 2002;Rickard, McCluskey, & Patterson, 2015).
Also, as an intrinsic attribute, organoleptic attributes have been added to the study.EVOO taste can essentially be represented by four sensory attributes, as defined by EC Regulation No. 2568/91 and subsequent amendments: fruity, pungent4 , sweet and bitter.Although the reaction of consumers is heterogeneous towards EVOO taste (Delgado & Guinard, 2011), an important trend has been highlighted elsewhere: despite bitter and pungent notes are not preferred by the majority of consumers (Del Giudice, Cavallo, Caracciolo, & Cicia, 2015), a preference for the latter two sensory features is shown by expert consumers and can be considered an indicator of technological quality of EVOO (Del Giudice, Cavallo, & Vecchio, 2018;Tuorila, Recchia, Monteleone, & Langstaff, 2014;Vecchio, Cavallo, Cicia, & Del Giudice, 2019).
In order to investigate the reaction of consumers to ultrasound-extracted EVOO, we considered the performance of this innovation compared to similar ones.It was, thus, possible to separate the individuals' preference for novelty, in general, from the preference for this single innovation.For this purpose we chose three other innovations: (a) microwave treatment: microwave radiation used in olive oil extraction plants during the malaxation phase.Although it allows the olive paste to be conditioned in a shorter time than the traditional system and avoids problems associated with process discontinuity, the resulting EVOO has a lower content of phenolic compounds (Kalogianni, Georgiou, & Hasanov, 2019;Tamborrino, Romaniello, Zagaria, & Leone, 2014).(b) nitrogen packaging: EVOO packed in a material with low oxygen permeability and air removed in the headspace, either by fully filling EVOO bottles or replacing air with nitrogen, an inert conditioning gas (Sanmartin et al., 2018).(c) bag-in-box packaging: EVOO packed in a new container type intended for household use in order to retain its original quality longer than that of traditional containers (Garrido- Delgado, del Mar Dobao-Prieto, Arce, & Valcárcel, 2015;Lolis, Badeka, & Kontominas, 2019) The final set of innovation levels is thus balanced: two innovations increase the efficiency of the EVOO extraction process (benefiting producers) and two extend the shelf-life of the final product (benefiting consumers).Although the content of phenolic compounds does not change for packaging innovations, by extending shelf-life their content persists for longer (Sanmartin et al., 2018).

Characteristics of consumers
In a context where ratings are highly subjective and preferences are heterogeneous, it is important to allow for the psychological characteristics of consumers.There are several individual traits which are able to shape perceptions and lead to different behaviors (Cavallo & Piqueras-Fiszman, 2017;Piqueras-Fiszman, Ares, & Varela, 2011;Piqueras-Fiszman & Spence, 2015).To this end, and for better segmentation, via the survey we collected some information on the individual traits characterizing the respondents.
The respondents were characterized according to their habits in terms of EVOO purchasing.The Italian EVOO market represents a particular case in which direct farm purchase has an important share, especially in the most important production areas where consumers are highly familiar with the product (Pomarici & Vecchio, 2013).Some psychological dimensions were also measured.First of all, the survey helped to categorize respondents according to their degree of neophobia towards new technologies through the Abbreviated Food Technology Neophobia Scale (AFTNS) which was rated on a Likert scale from 1 to 7 (Schnettler et al., 2017).Indeed, as the core of our study was to investigate how consumers react to a specific new technology, knowledge of how the individual engages with new technologies in food, in general, could be a valuable measure to understand and predict consumer behavior (Cavallo & Materia, 2018).
Two sub-scales were taken from the Health and Taste Attitude Scale (HTAS) by Roininen et al. (2001) in order to ascertain the personal relationship with food and its benefits.As regards the choice of food, consumers are constantly faced with the dilemma of choosing food with either the highest health potential, or the best taste.These two choices are seen as opposite alternatives, being the long-or short-term consequences of food consumption, respectively (Raghunathan, Naylor, & Hoyer, 2006).This scale helps to understand which of the two represents the stronger reason for the personal choice of food.The investigated traits were: to what extent the individual considers health while shopping for food (GHI -General Health Interest) and to what extent food is considered only as a source of pleasure (FP -food as Pleasure).They were also evaluated on a Likert scale from 1 to 7 (Roininen, Lähteenmäki, & Tuorila, 1999).
The last personal trait recorded by the survey was the Food Decision Involvement Scale (FDI) by Levi, Chan, and Pence (2006).Given that the degree of effort devoted to purchases can depend, among other factors, on the degree of involvement felt with the category of products, the survey aimed to measure the degree of involvement with the individual's food decisions.The reason lies in the different mental processes and trade-offs occurring in the mind of the consumer who is making a piecemeal decision vs. the consumer who is making a heuristic decision (Petty & Cacioppo, 1986).
Finally, the main socio-demographic characteristics of respondents were collected.Previous studies have shown that EVOO preferences are partially linked to demographics (Giannoccaro, Carlucci, Sardaro, Roselli, & De Gennaro, 2019;Roselli, Giannoccaro, Carlucci, and De Gennaro, 2018).In particular, we asked for details regarding the household member responsible for purchasing, such as gender, age and education, as well as income, size of household and place of residence.

The survey
A consumer survey was carried out in Italy, the largest olive oil consumer country in the world, during November-December 2017.A market research agency recruited a nationally representative sample of consumers who were involved in a web-based interview.The inclusion criteria for the target population were that: (i) the interviewee was the household member responsible for food purchasing, and (ii) he/she had bought EVOO at least once in the year prior to the survey.Ultimately, a total of 1003 responses were recorded.Participants were recruited with stratified quota sampling based on the geographical area, municipality size, age, gender and education, in order to ensure the representativeness of the sample at national level.
The composition of the final sample based on the participants' sociodemographics is summarized in Table 1.It appeared to be quite balanced with some deviations with respect to the Italian adult population: the older and least educated respondents were underrepresented compared to the Italian adult population as a whole.Although this is a common limit of Internet surveys, it was assumed that this discrepancy does not have a significant impact on the results.

Informative messages
To test consumer's reaction to novel elements in EVOO production, in our survey we used a set of innovations that the consumer had to consider as available options in the choice experiment.For this purpose, we provided information about each innovation considered in the study.Information content was created based on previous research and following guidelines about how to formulate product information in experiments.To create balanced information, we formulated each description controlling for its content, framing, linguistic style and order of presentation (Buda & Zhang, 2000;Rosenblatt, Dixon, Wakefield, & Bode, 2019;Soliha & Dharmmesta, 2012).To control for content, we presented the same product attributes for each innovation (i.e., production cost, taste, quantity of extracted oil, amount of healthy substances, shelf-life, characteristics of olive orchards and their landscape value, olive harvesting and packaging).We also used the same style of language for each information item regarding production.To control for the framing effect, moreover, in each information item we presented any gains or losses, and the similarity to the traditional production.To control for the order of presentation, the description of each production was presented in randomized order.Below is a transcription of the presented information.
Below is a list of innovations that can be applied to the traditional production of extra-virgin olive oil.These innovations can produce differences in the main characteristics of the traditional production of extravirgin olive oil.Such characteristics concern the olive orchards and their landscape values, olive harvesting, oil extraction, quantity of oil extracted, presence of healthy substances, bitterness and spiciness of taste, packaging, shelf-life, and production cost.
Please read carefully the description of each innovation and then answer some questions about them.
Extraction assisted by ultrasound.Compared to traditional extravirgin olive oil production, ultrasound EVOO extraction uses an ultrasound-assisted treatment apparatus.Therefore, the production cost and both product bitterness and spiciness are lower than those of the traditional product.Moreover, the quantity of oil extracted, the amount of healthy substances, and the shelf-life exceed those of traditional EVOO.In this case, olive harvesting and packaging do not differ from those of the traditional production.
Extraction assisted by microwaves.Compared to traditional extra-virgin olive oil production, such EVOO extraction uses microwave-assisted treatment apparatus.With this technique, more oil is extracted, and the production cost is lower.However, the amount of healthy substances and shelf-life are lower than measured in traditional production.In this case, the olive harvest, packaging, and both the bitterness and spiciness of taste do not differ from those of the traditional production of extra-virgin olive oil.
Bottling with nitrogen.Compared to traditional extra-virgin olive oil, nitrogen bottling conserves healthy substances throughout shelflife, which is longer than that of traditional EVOO.However, the production cost is higher.The olive harvest, oil extraction and quantity of extracted oil, the amount of healthy substances, packaging, and both the bitterness and spiciness of taste do not differ from those of traditional EVOO production.
Packaging with bag-in-box.Compared to traditional extra-virgin olive oil, bag-in-box packaging may be described as a box with a dispenser and an inner envelope that prevents contact of the product with oxygen.Hence healthy substances are preserved throughout product shelf-life, which exceeds that of traditional extra-virgin olive oil.However, the production cost is higher.The olive harvest, EVOO extraction and quantity of the extracted oil, the amount of healthy substances, and both the bitterness and spiciness of taste do not differ from those of traditional EVOO production.

Data and design
Traditionally, in choice experiments, attributes and levels are specified in advance according to the researcher's knowledge, but for consumers there may be no noticeable difference between some of them, which can lead to heteroskedasticity problems (Caputo, Van Loo, Scarpa, Nayga, & Verbeke, 2018).To allow for this situation, we used a D-optimal design, which identifies the most informative points within a probability distribution and improves parameter estimates (Kanninen, 2002).D-optimality entails maximization of the determinant of the Fisher information matrix, which is equivalent to minimizing the asymptotic joint confidence sphere surrounding the parameter estimates (Carson et al., 1994).
In order to account for the modulation of the D-optimal design, a pre-test was conducted on a sample of 150 respondents.The experimental design was created according to an algorithm which maximizes the D-efficiency of the design based on the covariance matrix of the conditional logit model (Carlsson & Martinsson, 2003;Cook & Nachtrheim, 1980;Zwerina, Huber, & Kuhfeld, 1996).The design with the highest D-Efficiency coefficient (1.25) was chosen after several iterations.
The choice experiment was divided into two blocks, each with four rounds of choice.In each round, four profiles were presented, together with the "no-choice" option.Table 2 reports the selected attributes with related levels and Fig. 1 shows an example of choice card.

Latent class model
Econometric analysis was carried out through the latent class model (LCM) (Lazarsfeld & Henry, 1968), which simultaneously allows sample segmentation and segment-specific estimation of parameters.In particular, LCM captures preference heterogeneity across classes, i.e. consumer segments, but assumes homogeneous parameters within each class.
In formal terms, LCM investigates preference heterogeneity through both estimation of the probability of individual i belonging to class c, and the structure of preferences within each class (Swait, 1994).For class c, individual i's utility (U) for alternative j in choice situation s is specified by class-specific preference coefficients (α cj , β c ), defined as: where α cj is an alternative-specific constant (ASC) for consumers in class c, which captures the mean effect of unobserved utility for EVOO choice j.Vector x describes attributes, while the unobserved component of utility ε is independently and identically distributed (IID) extreme value type 1.The probability of individual i choosing alternative j is conditional on his/her belonging to class c, such that (Greene & Hensher, 2003): Respondents' class membership is defined by a vector z i of their observable characteristics.If H ic is the prior probability of membership of class c for individual i, and for which θ c are parameters indicating the impact of z i on class membership: then the probability of individual i choosing alternative j is: Finally, the values of coefficients that maximize log-likelihood through maximum likelihood estimation are calculated as (Greene & Hensher, 2003): WTP was computed according to WTP a = −(β a /β COST ), where WTP of attribute a is the negative ratio between the coefficient of the attribute β a and the cost coefficient β COST .The confidence intervals were estimated by the Krinsky and Robb simulations based on 1000 replications (Creel & Loomis, 1991;Haab & McConnell, 2002;Park, Loomis, & Creel, 1991).

Results
LCM analysis highlighted three segments of consumers (Table 3).Each consumer segment was profiled according to all the socio-economic and psychographic variables collected in the study, as explained above (Table 4).

Cluster 1 -innovative
The first consumer segment, identified as Cluster 1, accounted for 32% of the respondents.These consumers, compared to the other two clusters, are those who showed the highest preference for the EVOO of Italian origin and with the PDO label.They also appreciated the organic  As regards preferences for organoleptic this segment liked EVOO with sweet and fruity attributes and disliked pungency.Focusing on the possible acceptance for the proposed innovations, this group showed the most openness to innovations.Indeed, this was the only segment that showed a positive rating for EVOO extracted by ultrasound or packed under nitrogenflushed atmosphere.By contrast, their opinion toward bag-in-box packaging appeared neutral, the coefficient being extremely low (0.03).However, the cluster showed the highest aversion to EVOO extracted by microwave.With regard to socioeconomic characteristics (Table 4), a few differences were detected with respect to the other two groups, the most important being middle age and the highest share of respondent with medium-high income.The psychometric profile showed the lowest for the AFTNS scale but no significant difference from Cluster 3, the highest scores for the GHI and the FP scales but both similar to Cluster 2, and the highest scores for the FDI scale but close to the score of Cluster 3.

Cluster 2 -traditionalist
The second consumer segment, called Cluster 2, comprised 25% of the respondents.They revealed a liking for Italian EVOO and a disliking for EVOO of extra-EU origin, but both are lower than in Cluster 1.They exhibited the highest inclination for organic label EVOO, compared to the other two clusters.Instead, PDO certification did not affect their choices.Preferences for sensory attributes were similar to those of Cluster 1 with a slightly higher propensity for a fruity flavor and a slightly lower propensity for sweet EVOO.They showed the least openness to the proposed innovations.Only bag-in-box packaging was generally accepted, while all other innovations were rejected.They were the most averse to EVOO packed under nitrogen-flushed atmosphere and EVOO extracted by ultrasound.Their aversion to microwaves was also very high.This segment included elderly women, but all other socioeconomics were not relevant.The psychometric profile showed the highest scores for the AFTNS; the scores for the GHI and FP scales were similar to Cluster 1; the score for the FDI scale was similar to that of Cluster 3.

Cluster 3 -cautious
The third segment, called Cluster 3, consisted of 43% of respondents.These consumers showed a slight positive propensity for Italian origin of EVOO and a moderate aversion to origin.They liked the PDO label slightly more than Italian origin, but were uninterested in the organic label.They showed preferences for EVOO organoleptic attributes similar to the other two segments but with a somewhat lower intensity.The only accepted innovation is EVOO packed in bag in box, but aversion to other innovations is slight (microwave and nitrogen) or not statistically significant (ultrasound).This group included the youngest respondents, most prone to buying EVOO at supermarkets rather than on direct channels.Respondents in this cluster live mainly in urban or sub-urban areas.The psychometric profile showed an intermediate score for the AFTNS scale but not statistically differing from the least neophobic Cluster 1; the lowest scores for the GHI and FP scales; and an intermediate score for the FDI scale but not statistically differing from the other two segments.
Comparing the two innovative extraction techniques, the use of microwaves was not accepted by the whole sample, while ultrasound technology was preferred by Cluster 1 and rejected by Cluster 2; Cluster 3 appeared to be indifferent to the above innovation.The first consumer group was also interested in nitrogen-flushed packaging.Conversely, bag-in-box packaging was preferred by the other two segments.

Discussions and conclusions
This study started with the assumption that innovation of TFP has to be investigated in detail before the innovation arrives on the market, to prevent failure.For this purpose, we conducted a choice experiment to investigate consumer acceptance of the use of ultrasound technology for the EVOO production process, in order to provide producers with useful insights.In particular, the technology was compared to other new technologies in order to simulate a competitive scenario.
Our results showed that consumer attitudes towards emerging technologies in EVOO production varied considerably according to the type of technology, as well as among different consumer segments.Latent class analysis helped to group consumers into four segments with different opinions about EVOO purchase and quality.Among these groups, ultrasound extraction obtained a heterogeneous reaction among consumer segments, but the segment of "innovative" consumers, which accounted for 32% of respondents, appeared to highly appreciate ultrasonically assisted extraction, together with other certified quality labels of EVOO, such as geographical origin and organic method.
Overall, we obtained general indications about consumers' preferences and their distribution across groups of consumers.The most important element, according to consumer's opinion, seemed to be the Italian attribute: it was preferred by all the clusters highlighted by this study.The importance of this attribute has been extensively explained by previous research (Del Giudice, Cavallo, Caracciolo, & Cicia, 2015).
Similarly, homogeneous were preferences regarding the taste profile of EVOO: a sweet, fruity flavor is positively rated by consumers, while the opposite applies to pungent, bitter notes.This confirms previous investigations showing that consumers' aversion to bitterness and pungency affects preferences for EVOOs.This particularly seems to affect the market success of those that are rich in polyphenols (Caracciolo et al., 2020;Cavallo, Cicia, Del Giudice, Sacchi, & Vecchio, 2019).
With regard to innovations, the least neophobic cluster showed a preference for ultrasound-assisted EVOO extraction, while other clusters only seemed to accept the bag-in-box innovation.This is hardly surprising, since such packaging looks exactly like the boxes used for wine, which are extremely common on the Italian market.Hence more neophobic consumers have no aversion to such boxes.
With regard to the profile of the respondents in the first cluster, the one that best represents the target market of ultrasound-extracted EVOO, such consumers would appear to be not only the least neophobic of the sample, but also those most concerned about health and involved with food decisions.This suggests that they are those who devote more time and effort to their food choices (Kamrath, Bidkar, & Bröring, 2019).They are also those who would benefit most from the use of a product with an increased content in polyphenols.However, EVOOs richer in polyphenols currently have a high bitterness and pungency (Vitaglione et al., 2013), which are not liked by such consumers, as emerged from the questionnaire.This gap would be easily filled by ultrasound-extracted EVOO which, together with a high content in polyphenols, has a low level of bitterness (Clodoveo et al., 2017), bringing a tangible benefit to this cluster of consumers.
Our findings were as follows: first, the new technologies which tend to deviate considerably from traditional production practices without providing benefits in terms of improvement/preservation of product quality are widely rejected by consumers (e.g.microwave-assisted extraction), while the opposite applies to more familiar technologies (e.g.bag-in-box); secondly, other new technologies such as ultrasound-assisted extraction or nitrogen flushing), albeit deviating considerably from traditional production, provide tangible benefits and may be accepted by groups of consumers who are more willing to accept innovation than others.
The main managerial implications concern the finding that, despite a general consumer's reluctance to accept new technologies in EVOO production, a non-negligible consumer segment (32% of the sample) proved to be innovation-oriented, which suggests promising prospects for the development of new technologies in production.According to recent statistics, 90% of in Italy are regular consumers of EVOO, which corresponds to as many as fifty million consumers, an undeniably large number (Coldiretti, 2020).
However, the attempt to profile the different consumer segments in terms of socio-demographic and psychometric characteristics showed dissimilarities that can be considered only moderately significant.

Table 1
Socio-demographic characteristics of the sample.

Table 2
Attributes of the choice experiment.

Table 3
Estimation results of the latent class model., but less than Cluster 2. The highest aversion, instead, can be supposed towards EVOO extra-EU origin.