Leveraging consumer neuroscience and analytics for crisis-driven international marketing strategies: insights from the COVID-19 pandemic

The pandemic has affected consumer behaviour on a global scale, revealing the need for marketers to develop flexible strategies that target different segments during this crisis (Christofi et al., 2024). In particular, COVID-19 has prompted a reconsideration of the ways and means we consume. It has affected current paradigms worldwide by reshaping purchase choice elements and changing consumers’ behaviour (Alaimo et al., 2020, 2022; Nemes et al., 2021). The challenges faced during the pandemic require a diverse range of innovative approaches. According to Ohme et al. (2020), declarations and actions do not always align, making predicting future behaviour based solely on self-reported measures ineffective. Understanding behaviour is crucial for the success of any strategy related to the pandemic.

This paper aims to explore how deeply rooted, non-conscious consumer attitudes, particularly those shaped by a crisis like a pandemic, can be captured using Consumer Neuroscience tools and translated into actionable segments for international marketing strategy by utilising data collected in a study using iCode. This study targets Portugal as a case country for several key reasons. Firstly, Portugal implemented an early and stringent lockdown in March 2020, characterised by a noteworthy level of public compliance and trust in institutions compared to many other European nations (Eurofound, 2020). Secondly, during the initial wave of the pandemic, Portugal maintained relatively low case fatality rates. It was regarded as a “success story” in managing the crisis, providing a pertinent context for examining public health perceptions and institutional effectiveness (OECD, 2021). These factors made Portugal an informative setting for examining how public attitudes toward crisis context translate into consumer segments relevant for international marketing.

By integrating Consumer Neuroscience, segmentation gains a new perspective beyond declarations, investigating deep-seated, non-conscious beliefs that individuals may not even be fully aware of – beliefs that traditional opinion surveys often do not capture (Vrontis and Christofi, 2021). This study brings innovative insights into the influence of culture on attitudes during the crisis, which can be relevant for practitioners and academics. Traditional marketing research tends to rely on self-reported data, but these measures can be unreliable when consumer attitudes fluctuate and are emotionally charged in times of crisis. Going beyond explicit, declarative data appears to be a powerful avenue through which a better, context-sensitive understanding can be achieved. However, studies incorporating implicit measurement in the context of the pandemic are lacking. There is, therefore, an increasing need for innovative methodologies to comprehend consumer behaviour beyond what can be articulated through surveys or verbal responses. This research suggests that using Consumer Neuroscience can give marketers valuable insights into non-conscious consumer behaviour and offer a more accurate understanding of how consumers genuinely feel and behave during global crises by incorporating an additional layer of attitude accessibility and confidence.

The paper is organised into five sections as follows. Section 2 reviews the relevant literature on crisis-driven consumer behaviour, segmentation models, and Consumer Neuroscience to establish a conceptual foundation. Section 3 presents the methodology and analytical approach. Section 4 discusses the main findings and their implications for international marketing strategies, while Section 5 elaborates on the theoretical and practical contributions. Finally, Section 6 concludes the study by summarising its key insights, limitations, and directions for future research.

International marketing has continuously dealt with the impacts of worldwide crises, including natural disasters, economic recessions, and political instability (Salunkhe et al., 2023). Christofi et al. (2024) identify five key areas reflecting significant contemporary changes resulting from the COVID-19 pandemic to improve international marketing theory in the context of pandemics and other external crises. These areas impact international marketing practices and include changes in consumer behaviour, advancements in digitalisation and artificial intelligence, disruptions in supply chains, changes in communication and corporate social responsibility, and the development of dynamic international marketing capabilities (Vázquez-Martínez et al., 2021). Together, they reveal the need to adapt the crisis marketing strategies quickly. Loxton et al. (2020) found that fear-driven behaviours, such as hoarding food, hygiene products, and medical supplies, became common responses to the uncertainty caused by crises. According to Maslow’s Hierarchy of Needs, prioritising basic needs over discretionary items was a natural reaction during the pandemic. Psychological factors such as anxiety and stress also had a significant impact on consumer behaviour. Di Crosta et al. (2021) studied how heightened emotional states and uncertainty influenced consumer decision-making. This distinction between necessities and non-necessities became more profound as consumers aimed to minimise perceived risks and avoid unnecessary spending. In response, marketers changed the focus to promoting essential products while addressing consumers’ emotional needs by emphasising psychological comfort, safety, and security. Additionally, the pandemic fostered a growing preference for supporting local producers. For example, Butu et al. (2020) found that many consumers favoured purchasing locally sourced products, such as fresh vegetables, reflecting a broader societal movement toward community support in times of crisis. This trend allowed international marketers to emphasise local sourcing, sustainability, and community engagement in their messaging. By aligning with consumers’ desires to support local economies, brands strengthened their positioning. They built consumer trust, particularly when many individuals were concerned about global supply chains and the economic impact of the crisis. Notably, a significant increase in the adoption of digital technologies has also been observed (Christofi et al., 2024). The COVID-19 pandemic accelerated digital banking and online shopping as consumers sought safer and more convenient ways to manage their finances and make purchases (Alber and Dabour, 2020). This tendency extended to retail, where consumers increasingly turned to online shopping (Verhoef et al., 2023). To meet the evolving needs of consumers, marketers needed to focus on creating seamless, transparent, and reliable online experiences to build customer satisfaction and loyalty. In parallel, the economic fallout of the pandemic further altered consumer behaviour. As incomes decreased and financial uncertainty grew, consumers became more sensitive to economic factors and adjusted their purchasing behaviours accordingly (Rosário and Dias, 2023). Offering value-driven messaging, such as promotions, discounts, or special offers, has become essential for maintaining consumer engagement during periods of financial uncertainty (Hesse et al., 2021). Deeper research on consumer segmentation would help refine targeted marketing strategies and ensure that campaigns are tailored to meet different consumers’ specific needs (Dutta et al., 2024).

As digital marketing tools evolve, innovations such as AI, machine learning, and personalised messaging could be increasingly important in fostering consumer trust and loyalty (Akter et al., 2023). Although existing research provides valuable insights into crisis-driven consumer behaviour, several gaps require further attention. While some changes observed during the pandemic can be temporary, others, such as continued reliance on digital services or changes in consumer values, can have lasting effects. While the attitudes explored in this study, such as trust in healthcare, prosocial behaviour, and personal or financial concerns, may seem general or societal, they play a crucial role in shaping consumer responses to marketing efforts. In times of crisis, these broader psychological orientations significantly influence consumers’ trust in institutions, their receptiveness to messages, and their perceptions of risk (Di Crosta et al., 2021; Hesse et al., 2021). This perspective aligns with the findings of Christofi et al. (2024), which highlight the importance of both emotional and institutional trust in determining the effectiveness of international marketing strategies. Consequently, examining these attitudes provides marketers with a more comprehensive understanding of the emotional and cognitive landscape that informs strategic decision-making. Consumer Neuroscience can help better understand these long-term shifts, critical for marketers looking to adapt their strategies beyond the immediate crisis (Damasio, 2018; Ohme et al., 2024).

This section on segmentation literature related to crisis-driven marketing focuses on psychological factors, communication behaviours, and their implications. According to Di Crosta et al. (2021), this approach helps to adjust marketing messages with psychological drivers of consumers and can significantly improve the effectiveness of communication efforts. Marketing messages informed by segmentation analysis tend to align more closely with consumers’ current emotional realities, leading to more profound engagement during crises. In addition to psychological factors, communication behaviours, particularly electronic word of mouth (eWOM), have recently received significant attention. Hussain et al. (2020) found that segmentation based on consumer motivations and perceived risks can help marketers develop strategies that enhance consumer trust and loyalty in e-commerce. This is particularly important during crises, when consumers often rely on social ties and peer recommendations to inform their purchasing decisions. During the COVID-19 pandemic, psychographic segmentation has proven to be particularly valuable in understanding consumer behaviour in the tourism sector. Matiza and Kruger (2021) identified distinct tourist typologies, such as dogmatic, sceptical, and apprehensive consumers, each responding differently to risk and safety issues.

The impact of remote work and teleworking is another relevant aspect that Jalal et al. (2019) approached. During the pandemic, the authors explored how boundary violations between work and personal life affected teleworkers’ stress and well-being. They suggest that understanding gender differences and coping mechanisms can support strategies targeted to teleworkers. In turn, Hermes et al. (2022) discovered that consumer personality traits, such as impulsiveness and trust, also shape online shopping behaviour, revealing that it is possible to customise offers to the specific needs and motivations by separating consumers into hedonic and prudent types. Additionally, Ahmadi et al. (2022) investigated how different cultures may respond to crises in unique ways and showed how cultural dimensions influenced stockpiling behaviours during the pandemic. Despite the existing literature, the intersection of cultural values and psychological factors in consumer segmentation warrants further research.

Consumer Neuroscience has become decisive for understanding consumer behaviour in international marketing contexts (Song et al., 2025). For example, even breathing patterns during retail shopping can influence store product perception (Johannknecht and Kayser, 2022). Neuroscientific techniques and big data analytics offer a more nuanced understanding of consumer engagement, decision-making processes, and marketing strategy optimisation in the AI and digital transformation era (Zámečník, 2024). Zaltman (2003) posits that understanding the cognitive processes behind consumer behaviour is relevant for effective marketing strategy. Neuroscientific methods, such as metaphor elicitation and implicit association techniques, help uncover consumers’ subconscious minds, whereas traditional marketing methods are ineffective. By integrating these insights into marketing strategies, brands can create campaigns that resonate with consumers on a deeper psychological level. Plassmann et al. (2015) reinforce this idea by discussing applying neuroscientific methods to enhance marketing theory and practice. They outline specific approaches, including collaboration between neuroscience and marketing, to improve product development, pricing strategies, and brand communication. In turn, Hubert et al. (2018) argue against simplistic interpretations of consumer behaviour, advocating for a more comprehensive view informed by biological insights. This approach allows a deeper understanding of consumer preferences, contributing to marketing strategies and theoretical frameworks. Alsmadi and Hailat (2021) further highlight how neuroscience can uncover insights into consumer preferences, particularly by understanding the brain’s reward system. Moreover, integrating big data analytics with Consumer Neuroscience is becoming increasingly important in digital marketing and brick-and-mortar retail. While Beer (2018) emphasises the need to interpret big data in marketing contexts, suggesting that effective marketing strategies require a thorough understanding of both data and consumer insights, Vasilopoulou et al. (2023) support this by identifying the potential for digital consumer behaviour. To explore the importance of innovative neuroimaging techniques for understanding and influencing consumer behaviour in physical retail settings, Moses et al. (2021) explored the use of functional near-infrared spectroscopy (fNIRS) to study consumer behaviour in brick-and-mortar retail environments. Their study shows how fNIRS can capture neural activation in response to merchandising strategies, providing valuable insights into shopper behaviour and how consumers make purchasing decisions at points of sale. Other technologies employ neurophysiological measures to understand better the cognitive processes driving consumer behaviour. Electroencephalography (EEG) enables real-time measurement of consumer reactions to marketing stimuli (Cherubino et al., 2019). It provides a more nuanced understanding than traditional self-report methods, enhancing the predictive capabilities of marketing strategies (Lin et al., 2018). Integrating other physiological measures such as skin conductance and eye tracking further uncovers unconscious mechanisms that influence consumer behaviour (Chiavone Alves and Pasquotto Andreoli, 2024). The neuromarketing marketplace size in 2023 was estimated at approximately $3.3 billion, with related jobs rising. The integration of neuroscience and marketing has transformed the promotion world and the bond between businesses and consumers. These advanced methodologies significantly improve marketing strategies by offering insights into consumer decision-making processes, but several knowledge gaps remain. For example, while various neuroscientific techniques have been explored, there is a limited understanding of how Response Time measurement can be integrated to provide a holistic view of consumer behaviour.

Market segmentation is a critical strategy for international marketing, enabling businesses to effectively tailor their offerings to meet the diverse needs of global consumers (Poulis, 2024). Recent studies have shown that segmentation can greatly enhance marketing efficacy, especially in multicultural and diverse markets (Hernani-Merino et al., 2021). For example, food tourism segmentation allows campaigns that cater to varying consumer preferences and behaviours, increasing customer satisfaction and loyalty, while driving economic growth in the tourism sector (Berry and Kumar, 2024; Fernández-Cavia et al., 2024). Including local cultural elements in global brands has increased purchase likelihood, particularly in emerging markets. For instance, some features of a global brand transfer to local cultural elements and relations between global brands and local cultural elements. In turn, aligning marketing messages with consumer cultural identity and preferences improves brand acceptance and loyalty (Nie and Wang, 2021). In research on physical activity marketing, four distinct segments have been identified: Young Disinterested, Successful Enthusiasts, Vulnerables, and Happy Retirees. These segments exhibit unique needs and motivations, suggesting that marketers can optimise resource allocation and develop targeted communication strategies for each group. Successful Enthusiasts may benefit from campaigns highlighting physical activity benefits, whereas Young Disinterested may need more engaging and relatable messaging to attract their attention (Rundle-Thiele, 2015).

Ethnocentrism, xenocentrism, and cosmopolitanism are other important factors influencing consumer preferences across cultural contexts (Zdravković and Gašević, 2024). Understanding these constructs enables more effective audience segmentation. Marketers can develop campaigns that resonate with ethnocentric consumers by focussing on local production values while appealing to cosmopolitan consumers seeking global brands with local adaptations (Anaza et al., 2021). Additionally, the segmentation of art collectors based on personality traits helps marketers create personalised marketing efforts and curated experiences tailored to each group’s motivations and preferences, enhancing engagement and market performance (Pitt et al., 2020). Despite these insights, several knowledge gaps remain, including the intersection of cultural identity, digital engagement, consumer motivations, and the crisis-driven international marketing strategy segmentation.

The present study used the iCode Test, a browser-based response time testing solution. It is a tool that enables the evaluation of implicit consumer reactions and beliefs by combining two levels of measurement within a single survey. Specifically, for questionnaires administered using the iCode Test, both declarative responses and the reaction time required to provide an answer are recorded. In addition, the iCode Test employs a specific survey administration protocol (see Section 3.2.2) as well as a data processing algorithm (see Section 3.4). This approach allows for assessing attitude strength per Fazio’s attitude accessibility paradigm, constituting the theoretical foundation of the iCode Test. In particular, the model suggests that attitudes can be understood through two dimensions: (1) explicit, declared opinions and (2) implicit, accessible attitudes. Opinions that are strong and accessible are expressed with high confidence, indicated by quicker response times. In contrast, weaker and less accessible attitudes are expressed with hesitation, reflected in slower response times (Fazio et al., 1989). Fazio’s research indicates that the correlation between attitudes and behaviour is notably stronger among individuals who respond more quickly, suggesting that robust attitudes significantly influence behaviour (Fazio and Williams, 1986; Fazio et al., 1989). Thus, explicit and implicit answers in an attitude test reflect two different aspects of evaluation. Explicit answers involve conscious, deliberate declarations of attitudes, measured using a binary Yes/No scale, where respondents openly state their position on an issue. In contrast, implicit measure captures the underlying strength or accessibility of those attitudes through response times. Explicit measures (Yes/No answers) can be biased by social norms, evaluation apprehension, and rational “cost-benefit” thinking (Dovidio and Fazio, 1992). This means people often report what is expected rather than what they truly feel, especially in socially sensitive contexts. Self-reported attitudes frequently have a weak correlation with actual behaviour (r = 0.38, Kraus, 1995). Therefore, explicit answers alone may be insufficient for predicting future actions. Unlike traditional surveys that rely on self-reported data, the iCode Test provides a more profound insight into non-conscious consumer attitudes, revealing insights that consumers may not fully recognise about themselves.

For this study, a dedicated questionnaire was developed to explore people’s opinions on the COVID-19 pandemic and administered via the iCode Test platform. We tested 10 statements about life during the pandemic, grouped into five categories (two statements each): following guidelines, prosocial behaviour, evaluation of the healthcare system, worries about others and worries about self (see Table 1). The participant’s task was to answer on a 3-point scale: yes, hard to tell and no.

Surveys administered via the iCode Test platform consist of two sections: calibration and the main task. The calibration phase was divided into two parts. In the first one, participants were asked to choose the answer option that matched the one displayed on a bar at the top of the screen. This part of the procedure was designed to familiarise the respondents with the response scale and the arrangement of buttons on the screen. Nine such screens were included. In the second part, respondents were presented with a series of statements describing the purpose of the study and the research process. Nine screens were presented, each with a different statement. After a few statements were shown, respondents were asked to indicate which of the statements was presented last. If the response was incorrect, a message asked the respondent to work more carefully (“Please work carefully”). This aimed to increase participants’ readiness to focus on the displayed statements.

In terms of the main task, the participants were asked to express their opinions on 10 statements (see Table 1), which were presented in random order. The statements were shown on separate screens, and the answers were given on a 3-point scale: yes, hard to tell, no (see Figure 1. Participants’ declarative responses, as well as the time needed to provide the answer, were recorded.

To make sure that the registered response time for both yes and no answers is not biased (i.e. influenced by the position of the mouse cursor), a control screen (Advanced Button) was introduced. The Advanced Button screen appeared between all the main task screens. After the answer to each statement was given, a white screen with a blue button in the middle was presented. The task of the respondents was to click on the blue button. This ensured that the distance to both yes and no buttons was the same each time.

The study was conducted online using a non-probability convenience sampling method. Participants were recruited through social media platforms (Facebook, Instagram and LinkedIn), institutional mailing lists, and community forums. An open invitation link to the study was shared, and participants voluntarily completed the online questionnaire during the national lockdown period in Portugal (April 18 to May 6, 2020). To ensure data quality, participation was restricted to Portuguese residents aged 18 and older, and the study design prevented multiple submissions from the same IP address. The study was anonymous; no personal data of the respondents was collected. We managed to recruit 291 Portuguese citizens during that lockdown period. A sample of this magnitude was sufficient for the planned segmentation analysis. In particular, it was expected to identify between three and five segments and follow a rule-of-thumb approach to estimate the minimum number of participants, i.e. at least 50 participants per planned segment (Dolnicar et al., 2018; Hair et al., 2019). Demographics of the tested sample are presented in Table 2.

The first step of the analysis was to clean the response time data. Responses with latencies below 500 ms and above 10,000 ms were eliminated. Additionally, respondents who were very quick or slow in providing their answers in over 40% of responses were excluded from the analysis, as such response patterns were deemed indicative of insufficient attention to the task. The milliseconds were transformed using natural logarithms (Winer, 1971) and the extreme values were truncated. The final step was to standardise the values into a scale from 0 to 1,000, with the lowest values indicating the slowest response times and the highest values representing the fastest response times. This was done separately for each participant. Resulting values are referred to as implicit scores.

A cut-off value of 600 has been set as a threshold above which the implicit scores were treated as fast, confident responses. The proposed value has been calculated based on 15 studies, including approximately 3,500 respondents, and represents a unified value (benchmark) above which responses were significantly faster than average.

Based on that, two measures are distinguished: declarative answers (explicit) and confident answers (explicit responses with the implicit score equal to or higher than the threshold of 600). Explicit and implicit scores are not treated as contradictory; the implicit measure is used to supplement and deepen the explicit results.

Additionally, the gap ratio was introduced to further analyse the differences between declarative answers and the confidence of these answers. It is the percentage of people who expressed hesitation when answering (out of all respondents who said yes). This was calculated using the formula below:

Gap ratio = ((yes-confident yes)/yes) *100%

To perform cluster analysis, the explicit answers were combined with implicit scores into a single variable. To do so, the results for every respondent were transformed using the following formula:

EXP * IMP = Confidence Index

where EXP – declarative evaluation; IMP – standardised response time score (implicit).

Declarative evaluation was multiplied by the standardised response time score, creating a Confidence Index ranging from −1,000 to 1,000 (the declarative answer determined the valence; no answers were negative, yes answers were positive). The higher the value, starting from 0, the more positive and confident the answer; the lower the value, starting from 0, the more negative and confident the answer. This index represents what respondents express and reflects the confidence and intuition behind their responses, capturing underlying attitudes that may not be explicitly articulated.

After the above-mentioned transformations, three types of variables were obtained:

1. Declarative answers (nominal) – yes/hard to tell/no

2. Declarative answers filtered by response time (nominal) – confident/non-confident yes answers

3. Confidence Index (continuous): ranging from −1,000 to 1,000

The first step of the analysis was to compute frequencies and calculate the gap ratio using the declarative variables and the declarative variables filtered by response time. This was done to get a general overview of the participants’ responses.

In the second step, we aimed to uncover underlying patterns in participants’ responses that could not be identified through univariate analyses alone. For this purpose, we used the continuous confidence index and performed a cluster analysis using the 10 statements describing people’s opinions about the COVID-19 pandemic presented in Table 1. The statements served as foundational input for segmentation by capturing a broad array of crisis-related attitudes across five key categories: adherence to guidelines, prosocial orientation, trust in healthcare, concern for others, and concern for self. By segmenting the sample based on response patterns, we aimed to move beyond general trends and gain deeper insights into heterogeneity in pandemic-related experiences. To segment the respondents, we performed k-means cluster analysis. The number of clusters was determined using R 4.0.4 (R Core Team, 2021) and the NbClust package version 3.0 (Charrad et al., 2014). NbClust is a function that computes up to 30 indices, each recommending an optimal number of clusters for the given dataset. Extracted segments were then interpreted based on their prevailing response patterns across the 10 statements and described in terms of demographic variables.

When looking at the declarative results, 80% of answers expressed a favourable opinion (yes answers), 10% expressed a negative opinion (no answers), and 10% indicated a lack of opinion (hard to tell answers). Due to the significant dominance of affirmative responses, only yes answers will be presented.

At least 70% agreed with the tested statements in four out of five tested categories. Participants declared that they are following guidelines (93%), they are concerned about the health of others (93%), and at the same time, they positively evaluate the healthcare system (70%). The Portuguese are also willing to demonstrate prosocial behaviour (83%). The lowest number of affirmative answers was observed for worries about self (63%) (see Figure 2). According to Friedman’s test, the percentage of affirmative responses differed significantly across the tested dimensions, χ²(4) = 233.11, p < 0.001. All pairwise comparisons were significant except for worries about self vs. evaluation of the healthcare system and following guidelines vs. worries about others.

Analysis of the merged explicit and implicit answers (confident yes) shifts the ranking of categories. The confidence for worries about others and following guidelines is the highest (69 and 66%, respectively), followed by worries about oneself (47%). The confidence scores for the healthcare system evaluation and the eagerness for demonstrating prosocial behaviour are the weakest, with values of 27 and 32%, respectively (see Figure 2). The tested dimensions differed significantly regarding confidence scores as indicated by Friedman’s test, χ²(4) = 289.47, p < 0.001. Only two pairwise comparisons were nonsignificant: evaluating the healthcare system vs. prosocial behaviour and following guidelines vs. worrying about others.

The significant discrepancy between yes and confident high gap ratio scores reflects yes answers. The highest numbers were obtained for two categories: readiness to engage in prosocial behaviour (62%) and evaluation of the healthcare system (61%). For the other three categories, the gap ratio oscillates between 26 and 29% (see Figure 2). The gap ratio is used to capture the discrepancy between explicit declarations and implicit attitudes, and this discrepancy significantly varies depending on the tested dimension, as shown by Friedman’s test, χ²(4) = 162.06, p < 0.001. All pairwise comparisons were significant except for the following four: worries about others vs. worries about self, worries about others vs. following guidelines, worries about self vs. following guidelines, and evaluation of the healthcare system vs. prosocial behaviour.

To preserve the full variability of the data and enable accurate computation of interindividual distances, cluster analysis was conducted using the confidence index, as it was the only continuous variable. This approach facilitates the identification of distinct and interpretable subgroups based on nuanced patterns of fears and behaviours.

The analysis was initiated by determining the optimal number of clusters. This was done using R 4.0.4 (R Core Team, 2021) and the NbClust package version 3.0, which provides up to 30 indices of cluster validity (Charrad et al., 2014). The two-cluster solution was indicated the most frequently (10 times), followed by the four-cluster solution, indicated by five indices. To ensure greater granularity of the analysis, we chose the second-best option, that is, the four-cluster solution, splitting the tested sample into: Health Worriers (34%), Overwhelmed with Worries (34%), Unconcerned (10%), and Financial Worriers (22%).

Four of the 10 statements used in the cluster analysis did not significantly differentiate the clusters. Significant differences were observed between the clusters (see Table 3) for the following six statements, showcasing that they differ in their worries and health behaviours concerning COVID-19. Each segment can also be described using demographic characteristics. Table 4 illustrates the demographics for the segments, showing that the groups differ in the distribution of specific demographics. In particular, significant differences were found with respect to education and income. Details are described in the cluster description part below.

• Cluster 1 Health worriers

This segment is one of the two largest groups and comprises 34% of the sample. These individuals were generally worried about health, but not only their own. They expressed the most significant concern about the most vulnerable, especially older family members. At the same time, they did not worry about their finance. Most of this segment is formed by people with higher education who are employed full-time. This group (similarly to the Unconcerned) has a relatively higher income than the other two segments.

• Cluster 2 Overwhelmed with worries

This is the second group with the most significant percentage of the sample, and similarly to cluster 1, it consists of 34% of the sample. These individuals were the most affected by the pandemic and expressed high worries about both health and finances. This group, in comparison to other segments, expressed the highest worries about their own health and at the same time the lowest belief that in case of a Covid infection, they will get good medical care. Health worries seem to promote proactive behaviour, and these individuals, together with cluster 1, are the most actively encouraging others to follow restrictions and guidelines. In terms of demographics in this segment, we can find more people living in villages, with a high school education and students than in other segments. This segment also has a lower income than Health Worriers and Unconcerned.

• Cluster 3 Unconcerned

This smallest cluster consists of 10% of the tested sample. These individuals do not express any worries about health or finances. At the same time, they express one of the highest beliefs that in case of a COVID infection, they will get appropriate medical help. Even though they comply with COVID recommendations and restrictions, they are the least eager to encourage others to do the same. This is a relatively older and better-educated segment. A vast majority of this group are full-time employees.

• Cluster 4 Financial worries

22% of the sample falls into this cluster. This group consists of people who express two main worries – about their finances and the health of their older family members. At the same time, they are not worried about their own health. This segment focused on practical solutions, emphasising a desire for stability and tangible benefits rather than emotional responses. This group comprises most women and is the youngest segment with lower education than other segments, especially Health Worriers and Unconcerned. There are also more unemployed people in this segment than in other groups. This segment represents a lower-income group (similarly to Overwhelmed with worries).

COVID-19 brought an unprecedented challenge for international marketers, highlighting that traditional segmentation models that rely on demographics or self-reported behaviours often fail to capture the dynamics of consumer emotions and beliefs in times of crisis. This research demonstrates that segmentation based on implicit attitudes offers deeper psychological insights that can be effectively integrated into international marketing strategies:

1. Tailored Messaging and Emotional Segmentation

Marketing communication during crises should effectively address not only the rational concerns of consumers but also their emotional and psychological states (Di Crosta et al., 2021; Hesse et al., 2021). This research indicates that different consumer segments, such as Health Worriers and Financial Worriers, respond uniquely to health-related and economic messaging. This observation aligns with Maslow’s hierarchy of needs and supports the findings of Loxton et al. (2020), which suggest that fear and uncertainty significantly influence purchasing priorities. Marketers can utilise these consumer profiles to tailor their message framing and tone for diverse international audiences.

1. The Role of Trust and Prosocial Behaviour

Attitudes like trust in institutions and the willingness to engage in prosocial behaviour extend beyond mere sentiments, serving as strategic levers. For instance, consumers with confidence in the healthcare system tend to be more receptive to messaging emphasising institutional credibility. In contrast, those with lower levels of trust might respond more favourably to messages focused on peer influence (Ahmadi et al., 2022). This insight carries significant implications for cross-cultural marketing campaigns, where perceptions of government and authority can vary greatly (Zdravković and Gašević, 2024).

1. Bridging Neuroscience and International Strategy

This study enhances marketing research by incorporating neuroscience tools, such as implicit response-time testing, to create a more sophisticated segmentation model. Previous research (Plassmann et al., 2015; Ohme et al., 2020; Cherubino et al., 2019) has demonstrated that non-conscious attitudes often predict behaviour more accurately than self-reported data. In an international context, this methodological innovation enables marketers to identify subtle psychological differences across cultures that may not emerge through traditional surveys, thereby improving the targeting and effectiveness of global marketing campaigns.

1. Toward Culturally Attuned, Data-Driven Strategy

The proposed segmentation approach illustrates that even general societal attitudes (e.g. concern for others, fear of illness, or financial insecurity) can be systematically analysed and translated into actionable marketing strategies. This reinforces calls in recent literature (Christofi et al., 2024; Vrontis and Christofi, 2021) for developing dynamic, adaptive international marketing capabilities that are responsive not only to economic but also to psychological and cultural variables.

This study provides fresh insights at the intersection of behavioural science and international marketing strategy by introducing a new perspective combining declarative statements with implicit response times to reveal non-conscious consumer profiles, a significant methodological innovation. In light of the COVID-19 crisis, our findings identify four distinct consumer profiles in Portugal: Health Worriers, Overwhelmed with Worries, Unconcerned, and Financial Worriers. These nuanced profiles not only highlight variations in emotional and financial reactions to the pandemic but also reflect broader cultural and socioeconomic factors that are highly pertinent for international marketers navigating uncertain and volatile environments.

Building on the foundational work of Christofi et al. (2024) and Vázquez-Martínez et al. (2021), who call for dynamic capabilities and consumer-centric strategies in times of crisis, the results provide empirical support for shifting away from one-size-fits-all approaches. Instead, they demonstrate the strategic need of incorporating psychological segmentation, particularly when designing crisis-response strategies for diverse markets.

The use of implicit response time testing represents one more methodological advancement. As noted in prior literature (Fazio et al., 1989; Ohme et al., 2020; Cherubino et al., 2019), explicit declarations often fail to accurately reflect true behavioural intentions, particularly in emotionally charged situations such as global crises. This study illustrates that implicit attitudes, including hesitation and emotional dissonance, provide predictive insights into actual consumer behaviour, thereby serving as a valuable enhancement to the tools utilised by international marketers.

Our findings also align with the work of Ahmadi et al. (2022), who emphasised the cultural underpinnings of crisis behaviour, and Matiza and Kruger (2021), who identified distinct consumer typologies within the tourism sector. By basing our segmentation model on non-conscious cognitive responses, this study contributes a novel conceptual perspective to the international marketing strategy literature, which approaches emotional and cultural dimensions together. Theoretically, this study extends international marketing theory by integrating behavioural science and neuroscience into segmentation research. It advances existing consumer decision-making models under crisis conditions by incorporating attitude accessibility and implicit confidence as central constructs. These additions provide a more comprehensive understanding of how emotional and cognitive mechanisms influence global consumer behaviour. Furthermore, the findings contribute to cross-cultural consumer research by highlighting that implicit attitudes, often shaped by cultural norms, serve as latent drivers of market responses that traditional models may overlook.

As illustrated in Figure 3, general attitudes during a crisis are not isolated from market behaviour but serve as key antecedents that shape how consumers respond to marketing messages and brand engagement. Comprehending the attitudinal dimensions is crucial for formulating international marketing strategies that are both emotionally intelligent and culturally adaptive. This understanding facilitates the development of approaches that resonate with diverse consumer segments across varying cultural contexts.

1. Neuroscience-Based Segmentation as a Strategic Tool: This study introduces an innovative dual-level segmentation framework that integrates implicit cognitive data with explicit responses, effectively capturing crisis-induced behavioural shifts. Incorporating response time testing offers insights into the strength of attitudes, allows for a more apparent distinction between seemingly similar declarative responses, and encourages the creation of more tailored and targeted communication. Addressing a gap in the existing literature regarding the necessity for a multifaceted and refined segmentation framework equips marketers operating in culturally diverse markets with a valuable and actionable tool.

2. The model emphasises the influence of country-specific cultural norms and emotional drivers on consumer behaviour during crises, such as fear, trust, and financial anxiety. This underscores the need for context-aware, flexible international marketing strategies that are both emotionally attuned and culturally sensitive (Zdravković and Gašević, 2024; Nie and Wang, 2021), which is particularly critical in a crisis-driven environment.

3. Furthermore, this study advocates for a thorough exploration beyond surface-level assertions, proposing a novel conceptual framework to understand better consumer trust, compliance, and resilience under stress. These factors are vital for effective brand positioning, message framing, and developing loyalty-building strategies in the face of global disruptions. Additionally, considering the connection between attitude accessibility and behaviour, incorporating the dimension of attitude strength may enhance the predictive power of the segmentation framework.

This work advances international marketing strategy by illustrating how innovative, interdisciplinary methods can reveal hidden consumer drivers and support developing culturally resonant, psychologically grounded strategic decisions in global markets. From a decision support perspective, the segmentation framework developed here offers actionable insights for international marketers navigating crisis contexts. Practitioners can apply implicit testing tools like the iCode Test to identify underlying consumer anxieties and adapt real-time communication strategies. For example, campaigns targeting “Health Worriers” should emphasise reassurance and institutional trust, whereas messages directed toward “Financial Worriers” should highlight affordability, value, and security. Likewise, the “Overwhelmed with Worries” cluster suggests the need for empathy-based messaging and simplified digital experiences that reduce cognitive load. These insights can inform product positioning, crisis communication, and brand resilience planning across different cultural markets. In addition, the proposed methodology can support public policy communication by helping institutions frame messages that resonate with citizens’ implicit emotional states, ultimately enhancing compliance and trust.

This study aimed to understand how implicit attitudes toward health, prosocial behaviour, financial security, and institutional trust, elicited during the pandemic, could be leveraged to segment consumers in ways that enhance international marketing strategies. Utilising Consumer Neuroscience techniques, particularly response time testing (iCode), Portuguese consumers’ explicit and non-conscious attitudes were captured during the national lockdown. This data was transformed into a Confidence Index, facilitating a segmentation process that identified four distinct consumer profiles: Health Worriers, Overwhelmed with Worries, Unconcerned, and Financial Worriers.

This segmentation model provides a psychologically grounded, data-driven tool for tailoring international marketing strategies to consumers’ emotional and cognitive states in crises. It underscores the importance of moving beyond traditional self-report surveys and integrating neuroscience-informed methods to uncover underlying behavioural drivers, particularly in multicultural and high-stress environments. Furthermore, the findings illustrate that attitudes typically perceived as societal or institutional (e.g. healthcare trust) can have significant implications for international marketing communication, message framing, and strategic positioning.

However, the study presents some limitations related to a sample from only one country (Portugal), which constrains the generalisability of the findings. Additionally, the data were collected during a particular phase of the pandemic (April–May 2020), and attitudes may differ across various stages of crisis recovery.

Future research should apply this methodology to cross-country samples to examine how cultural, economic, and political contexts influence implicit consumer attitudes and segmentation patterns. Comparative studies would aid in validating the robustness of the model and refining its applicability for culturally adaptive marketing strategies. Longitudinal designs could also show how consumer segments evolve, supporting the dynamic and responsive decision-making process in international marketing planning. Theoretically, this research advances international marketing research by integrating neuroscientific tools into the framework of segmentation theory, thereby offering a multidimensional perspective on consumer behaviour in crisis situations. Practical implications of the study are based on the roadmap for organisations aiming to develop emotionally intelligent and culturally adaptive strategies. By merging explicit analytics with implicit neuroscience data, multinational companies can better anticipate behavioural changes and create communications that resonate with consumers’ subconscious motivations, an essential skill for navigating uncertainty in global markets. This method equips practitioners with a decision support tool that helps to design communication strategies aligned with consumers’ implicit emotional states and cultural contexts. By establishing a connection between neuroscience and marketing analytics, the proposed approach empowers marketers with data-driven tools to predict behavioural changes and foster resilience in global markets both during and after crises.

In conclusion, this study highlights the strategic value of integrating neuroscience, segmentation theory, and international marketing practice. As global markets continue to navigate the impacts of crises, leveraging these advanced tools can be relevant to stay ahead of changing consumer needs and preferences.