Selling on your own land: emergence and self-organization of innovations among small-scale coffee producers to sell in the Mexican domestic market

1 Introduction

Coffee is one of the most significant agricultural commodities in global trade, with a value chain involving over 125 million people. In 2025, production reached 176.4 million 60-kg bags, with a value of 28.5 billion US dollars (OIC, 2025). Mexico accounts for 2.3% of global exports, cultivating coffee across 564,651 ha distributed across 14 states, with 358,575 production units, the vast majority of which are small-scale and family-operated (Instituto Nacional de Estadística y Geografía, 2022).

Despite this presence, the appropriation of generated value remains profoundly unequal. This prompted small-scale producer organizations (SPOs) to establish the fair-trade system in 1988 in order to challenge the rules of the conventional market (Coscione, 2018). However, increasing regulatory requirements have diluted its benefits and hindered the effective participation of SPOs (Doherty, Davies and Tranchell, 2012). Selling in the Mexican open market has thus become necessary, yet SPOs lack specialized sales and promotion units, limiting their capacity to understand market dynamics and make strategic decisions (Guerrero and Herrera, 2023).

The Localized Agri-Food Systems (SIAL) approach offers a pertinent analytical lens, as it examines food production, processing, marketing, and consumption systems as localized productive systems organized around specific local resources (Boucher and Pomeon, 2010; Muchnik et al., 2011). The SIAL framework studies how the collective activation of territorial resources related to food—origin, productive tradition, cultural identity—can become a source of competitiveness (Boucher and Fraire, 2012; Correa et al., 2006). SIALs are processes of territorial construction that seek to qualify products through their connection to origin and cooperation, elements that resonate with the value logics of specialty coffee shops—that is, those that market coffee with characteristics that consumers consider to distinguish it from other products and for which they are willing to pay a premium (Tolentino et al., 2018).

The domestic market offers opportunities for SPOs, as consumption of roasted and ground coffee is increasing, particularly in coffee shops (NUUP, 2018). Specialty coffee shops (or third-wave coffee shops) require value co-creation with producers (Boaventura et al., 2018), representing a window for SPOs to capture greater value. However, to capitalize on this window, SPOs must overcome their competitive disadvantages in sales, promotion, and market knowledge (Guerrero and Herrera, 2023).

The complex systems perspective can contribute useful elements to enable SPOs to improve their understanding of Mexico’s domestic coffee market and to identify and foster the emergence of innovations and learning processes for self-organization, articulating both generic and territory-specific resources (Boucher and Pomeon, 2010), thereby complementing the SIAL approach. The complex systems framework studies how the interactions of SPO agents with new actors in the market’s possibility space generate the emergence of innovations, self-organization, and coevolution (Ackoff, 1971; Bohórquez, 2013; Earls, 2013; Kauffman, 1995; Lara-Rosano, 2025). Of particular importance for SPOs is the concept of the edge of chaos, as it describes the pendular transition between the chaos of innovation and the order of self-organization (Kauffman, 1995; Kelly and Allison, 1999).

This research addresses two questions: What elements from the SIAL and complex systems approaches are important for SPOs to develop competitive advantages in their participation in the domestic coffee market? And how can SPOs articulate processes of emergence and self-organization of innovations to coevolve with the coffee shop market, activating territorial resources through intermediate projects, and what second-order learning emerges for the construction of the coffee SIAL?

The general objective is to analyze how these processes can be enhanced through the collective activation of territorial resources in order to design market strategies aimed at expanding the possibility space and constructing a coffee SIAL. The specific objectives are: (i) to characterize the coffee shop segment in eight cities through data mining, identifying value logics and quality demands; (ii) to document interaction practices between SPOs and market actors, identifying second-order learning; and (iii) to evaluate how these learnings enable SPOs to capture greater value, transcending raw commodity marketing logics and developing advantages based on emergent differentiation and coevolutionary adaptability.

2 Methodology

The study’s methodology is grounded in a sequential mixed-methods approach that combines quantitative data mining strategies with qualitative participant observation and collaborative workshop approaches. This design responds to the complex nature of the research object, which requires both the identification of aggregate patterns in the specialty coffee shop market and the capture of situated learning and innovation processes at the level of Small-Scale Producer Organizations (SPOs). The adoption of a sequential mixed-methods design allows initial quantitative findings to guide case selection for the qualitative phase, establishing a dialogue between both strategies that enriches the understanding of the phenomenon under study (Creswell, 2009).

The study was structured in three complementary phases, each designed to address one of the specific research objectives. The first phase characterized the coffee shop segment in eight Mexican cities through digital review data mining. The second phase documented and analyzed deliberate interaction practices between SPO agents and market actors through coffee shop visits and cupping sessions. The third phase consisted of a final validation and co-construction workshop for market strategies with participating producers.

2.1 Operationalization of key concepts

A distinctive aspect of the methodology was the construction of an operationalization matrix that linked theoretical concepts from complex systems and the SIAL approach with observable empirical indicators (Table 1). For each concept, a precise operationalization was defined, specifying empirical indicators, recording techniques, and measurement criteria. This operationalization enabled the translation of theoretical constructs into concrete research procedures, addressing the recommendation by Boucher and Reyes-González (2016) regarding the need to operationalize SIAL concepts for their empirical study in Latin American contexts.

TABLE 1. Operationalization of concepts addressed in the research.

Source: Own elaboration based on Kauffman (1995), Smith (2012), Bohórquez and Espinosa (2015), Muchnik et al. (2011), Correa et al. (2006), Boucher and Fraire (2012), and Fournier and Muchnik (2012).

The matrix included central concepts from the SIAL approach—specific territorial resources, territorial embeddedness, collective activation of territorial resources, cooperation, and agents—as well as concepts from complex adaptive systems theory—emergence of innovations, self-organization, coevolution, and possibility space—in addition to the cross-cutting concepts of intermediate projects and second-order learning, which articulate both theoretical frameworks.

2.2 Phase one: analysis of digital repositories

The first phase was based on the analysis of 12,253 reviews posted on the TripAdvisor platform, corresponding to 148 coffee shops located in eight cities across the country (Table 2). The selection of these cities was guided by criteria of economic dynamism, relevance as specialty coffee consumption hubs, and geographic diversity, covering regions in the north, center, west, and south of the country. The choice of TripAdvisor as a data source is justified by its broad geographic coverage, the availability of structured information on establishments, and the possibility of accessing a considerable volume of reviews spontaneously generated by consumers, thereby capturing perceptions unmediated by the researcher.

TABLE 2. Number of coffee shops and reviews analyzed in the data mining.

Source: Own elaboration.

For the systematic classification of comments, a deterministic rule-based lexical classifier was implemented that evaluated the presence of predefined keywords in the reviews. The classifier construction process included: (i) the definition of a dictionary of terms associated with each SIAL and complex systems concept, based on the specialized literature; (ii) the specification of classification rules that considered term co-occurrence and context; and (iii) validation of the classifier using a test set of 500 reviews manually labeled by two researchers. The classifier validation showed acceptable performance, with precision of 84%, recall of 81%, and accuracy of 83%.

The analysis of comments was oriented toward identifying the presence of SIAL and complex systems concepts and their relationship with customer ratings across dimensions such as food, customer service, beverages, and establishment atmosphere. To this end, three types of statistical analyses were performed: (i) Spearman correlation analysis to determine the clustering of SIAL and complex systems variables; (ii) frequency analysis to identify the most common aspects in the reviews; and (iii) analysis of variance (ANOVA) to assess the association between the presence of SIAL characteristics and customer satisfaction. It is important to note that correlation and ANOVA analyses identify statistical associations, not causal relationships. Customer satisfaction may be influenced by multiple factors not considered in the analysis, such as establishment location, price, or service quality, which could be correlated both with the presence of embeddedness narratives and with customer satisfaction.

2.3 Phase two: interactions and emergence of innovations

The second phase implemented a qualitative strategy based on participant observation during coffee shop visits and cupping sessions. A total of 39 coffee shops were selected across three strategic cities (Oaxaca de Juárez, Monterrey, and Guadalajara), prioritizing establishments that in the previous phase showed mentions of cooperation or superior quality; the selection was therefore intentional. This selection criterion, while appropriate for identifying revelatory cases of the phenomena of interest (Flyvbjerg, 2006), implies that the qualitative findings are not generalizable to the entire universe of coffee shops, but rather document processes occurring in a particularly receptive subset of the market.

A group of ten individuals from four organic fair-trade SPOs (Campesinos Ecológicos de la Sierra Madre de Chiapas, Café Gourmet Sierra Azul, Unión Majomut, and Café Capitán) participated in the visits and cupping sessions. Participants included managers, roasters, and sales staff. The selection of these organizations was guided by criteria of active certification (fair trade and organic), specialty coffee production (SCA score >80), interest in entering the domestic market, and availability to participate throughout the entire research cycle. The inclusion of four SPOs with diverse characteristics allowed for the capture of a variety of perspectives and experiences, strengthening the richness of the qualitative analysis.

During coffee shop visits, conducted between 2024 and 2025, participants utilized a semi-structured observation guide that included sections on offerings, atmosphere, service practices, and linkages to the production chain. Each visit lasted 45–60 min, followed by a 30-min collective reflection session, and was recorded in both individual and collective field diaries. Participant observation enabled researchers to document not only observable practices but also producers’ interpretations and reflections on their experience.

Between 2024 and 2026, four cupping sessions were organized (two in Guadalajara, one in Monterrey, and one in Oaxaca) under an open format that fostered direct interaction among producers, baristas, and coffee shop owners. In these events, SPOs prepared coffee from their own organizations using different brewing methods, allowing market actors to directly evaluate product quality and provide feedback to producers. Each cupping session was documented through audio recording, field notes, photographic records, and producer reflection logs. These spaces were conceived as intermediate projects (Smith, 2012) that enable translation between the fair-trade niche and the conventional market regime, facilitating the collective activation of specific territorial resources through face-to-face interaction between production- and consumption-side agents.

2.4 Integrated analysis and final workshop

The information generated in the two phases was analyzed in an integrated manner. For the quantitative analysis, Spearman correlation analysis was conducted to identify the degree of association among key variables, and chi-square tests were performed to assess independence between categorical variables. The use of Spearman correlation, a non-parametric rank-based statistic, was appropriate given that the variables did not meet the normality assumptions required for parametric correlations.

Concurrently, a qualitative thematic content analysis was carried out following the six phases proposed by Braun and Clarke (2006): familiarization with the data, generation of initial codes, searching for themes, reviewing themes, defining themes, and producing the report. To strengthen the reliability of the qualitative analysis, the coding process was conducted independently by two researchers, who subsequently compared and reached consensus on the emerging codes and themes, following the recommendations of Lincoln and Guba (1982) to ensure credibility and confirmability of qualitative findings. The inter-coder agreement index reached 86%, indicating a satisfactory level of reliability.

In 2026, the results were presented and discussed in a 2-day workshop with SPO participants. The objectives of this space were to systematize the documented second-order learnings, validate the researchers’ interpretations, identify configurations of market strategies oriented toward expanding the possibility space, and explore the potential formation of a basket of goods and services that articulates coffee with other products and services from the territory. The workshop was structured in four moments: presentation of results, learning mapping, feasibility analysis of strategic configurations, and co-design of a roadmap. This co-construction space not only allowed for the validation of findings with participants but also generated practical knowledge for SPOs regarding viable market strategies.

2.5 Ethical considerations and limitations

The project was conducted under collaborative research principles: participating producers gave written informed consent, confidentiality of sensitive information was guaranteed, and the workshop results are jointly owned by researchers and SPOs. The collaborative design responds to the recommendation in the literature on participatory action research in agri-food systems (Boucher and Reyes-González, 2016), which emphasizes the importance of involving local actors in the production of knowledge relevant to them.

The methodology acknowledges explicit limitations that affect the scope of the findings. First, the selection of coffee shops prioritized establishments with high visibility on TripAdvisor, which introduces a bias toward coffee shops that have already achieved a certain positioning in the digital market, leaving out less visible establishments that might exhibit different dynamics. Second, the TripAdvisor platform may not be representative of the universe of coffee consumers in Mexico, as its users tend to be a specific subset with particular characteristics in terms of age, educational level, and technology access that may not reflect the diversity of coffee consumption in the country. Third, the lexical classifier has a margin of error (84% precision) that may have affected the classification of some comments. Fourth, the qualitative sample size (ten participants from four SPOs) is small and does not permit generalizing findings on learning and self-organization processes to other SPOs with different characteristics. Finally, the bounded timeframe of the fieldwork (2024–2026) does not allow for capturing long-term coevolutionary processes, which by their nature require extended temporal horizons to observe structural changes in actors’ practices and narratives. These limitations are considered in the interpretation of results and are explicitly acknowledged in the discussion, where their implications for the validity and scope of the conclusions are addressed.

3 Results

The findings of this research offer robust empirical evidence on the dynamics of SIALs in the context of the TripAdvisor-visible coffee shop segment in Mexico, while simultaneously revealing the intricate mechanisms of emergence, self-organization, and coevolution underlying SPOs’ entry into the domestic market. The convergence of quantitative and qualitative analyses allows us to move beyond description and delve into understanding how value is constructed, new market logics are learned, and the strategic capabilities of territorial actors are reconfigured. The results are presented in two major sections: quantitative findings on the one hand, and qualitative findings on the other.

3.1 Quantitative results: characterization of the coffee shop segment in eight mexican cities

3.1.1 Structural relationships among SIAL variables

Spearman correlation analysis (Table 3) reveals patterns that allow for the identification of the most relevant elements for the development of competitive advantages by SPOs. The strongest correlation observed was between cooperation and territorial embeddedness (rho = 0.354), followed by emergence of innovations and collective resource activation (rho = 0.243), and collective resource activation and territorial embeddedness (rho = 0.209), indicating an association between cooperation and the linkage of the product to its territory of origin, as well as the innovation involved in the concerted mobilization of resources. The lowest correlations in the study were found for the cooperation variable, particularly in its interaction with the emergence of innovations variable (rho = 0.014).

TABLE 3. Overall correlation analysis.

Source: Own elaboration.

3.1.2 Territorial variations of the SIAL by city

The city-level analysis reveals substantial differences in the patterns of association among variables, allowing for the identification of distinct levels of SIAL consolidation (Table 4).

• Oaxaca: The most consolidated SIAL

Oaxaca stands out for presenting the highest correlation between collective resource activation and emergence of innovations (0.608), as well as the highest correlation between collective resource activation and territorial embeddedness (0.402). This indicates that in Oaxaca, the concerted mobilization of territorial resources is strongly associated both with the linkage of coffee to its origin and with the emergence of innovations. For SPOs, this represents a paradigmatic case of how collective activation can enhance differentiation and innovation in the marketplace. The correlation between agents and emergence of innovations (0.270) is also the highest among the cities, suggesting that in Oaxaca, agents play a central role in innovation processes.

• b. Tijuana: A SIAL under construction through demand

Tijuana presents the second-highest correlation between collective resource activation and emergence of innovations (0.379) and a notable correlation between agents and emergence of innovations (0.251). This is significant because Tijuana is not a city with a coffee-producing tradition. The high association between collective activation and innovation suggests that, even in urban contexts without local production, specialized demand and cooperation with producing regions can construct origin narratives and generate innovations.

• c. Guadalajara and monterrey: SIALs in development

These cities present moderate correlations across all pairs, without standing out in any specific pair. The correlation between collective resource activation and emergence of innovations is 0.167 in Guadalajara and 0.127 in Monterrey, indicating that innovation is associated with collective activation, but less intensely than in Oaxaca. The correlation between agents and emergence of innovations is 0.213 in Guadalajara and 0.224 in Monterrey, suggesting that agents are a relevant factor for innovation in these cities.

• d. Puebla and Morelia: Particular patterns

Puebla presents one of the highest correlations between collective resource activation and emergence of innovations (0.276), but the lowest correlation between collective resource activation and territorial embeddedness (0.104). This suggests that in Puebla, innovation emerges from collective activation, but without strong territorial embeddedness. Morelia, in turn, presents a particularly high correlation between territorial embeddedness and territorial resources (0.242), indicating a solid base of territorial resources that are linked to origin.

• e. Querétaro and Guanajuato: Incipient SIALs

Querétaro and Guanajuato exhibit the most homogeneous and moderate patterns, with no pair standing out particularly. In Querétaro, the correlation between collective resource activation and emergence of innovations (0.203) and between agents and emergence of innovations (0.206) is consistent with overall values. The negative correlation between cooperation and emergence of innovations (−0.004) in Querétaro and between cooperation and territorial resources (−0.009) in Guanajuato suggests that cooperation is not associated with innovation or territorial resources in these cities, indicating systems that have not yet been configured.

TABLE 4. Correlations by city (selection of relevant pairs).

Source: Own elaboration.

CRA, collective resource activation; TE, territorial embeddedness; COP, cooperation; TR, territorial resources; AGE, agents; EI, emergence of innovations.

3.1.3 Frequency of mentions of SIAL concepts in reviews

The frequency analysis of mentions of each variable in the reviews allowed for the identification of patterns in SIAL construction (Table 5). The percentage of positive mentions relative to the total sample per city was calculated. Territorial embeddedness was the variable with the highest presence across all cities, with Oaxaca (61.0%) and Morelia (59.8%) standing out. Emergence of innovations was highest in Oaxaca (39.5%) and Guadalajara (34.8%), while territorial resources presented the lowest frequency overall, with Puebla (14.4%) as the lowest value. Cooperation showed relatively high values, especially in Guadalajara (43.9%), Querétaro (43.4%), and Tijuana (42.2%).

TABLE 5. Frequency of SIAL variables by city.

Source: Own elaboration.

3.1.4 Association between SIAL concepts and customer satisfaction

Table 6 presents the p-values resulting from the comparison between the active SIAL group (=1) and the inactive SIAL group (=0) for each variable and each dimension of customer satisfaction, using analysis of variance (ANOVA). Territorial Embeddedness shows large effects (η²p between 0.17 and 0.22) across all dimensions. Territorial Resources also presents large effects on Food (η²p = 0.14) and Service (η²p = 0.16), and a small-to-medium effect on Beverages (η²p = 0.028). This confirms that the mobilization of territorial resources is an important driver of value, though somewhat less consistent than embeddedness.

TABLE 6. General ANOVA results with effect size (η²p).

η²p ≥ 0.14 is considered a large effect (highlighted in bold).

Significance legend: *p < 0.05, p < 0.01, *p < 0.001. Effect size is reported as partial eta-squared (η²p). According to Cohen’s (1988) criteria, values of 0.01, 0.06, and 0.14 are considered small, medium, and large effects, respectively. Bold values indicate a large effect (η²p ≥ 0.14).

Collective Activation, Cooperation, Agents, and Innovations present small effects (η²p < 0.02) across all dimensions. Although some p-values are significant, the effect size indicates that their practical contribution to customer satisfaction is very limited compared to territorial embeddedness.

Given that territorial embeddedness was the variable with the strongest and most consistent effect, Table 7 presents the city-level breakdown with respective effect sizes. Morelia and Oaxaca show small effects (η²p ≈ 0.03–0.04) for beverages and atmosphere, suggesting that in these cities territorial embeddedness has less impact on those specific dimensions, although it remains strong for food and service.

TABLE 7. ANOVA by city–effect size for territorial embeddedness (η²p).

Source: Own elaboration.

η²p ≥ 0.14 (highlighted in bold) indicates a large effect. Effect size is reported as partial eta-squared (η²p). According to Cohen’s (1988) criteria, values of 0.01, 0.06, and 0.14 are considered small, medium, and large effects, respectively. Bold values indicate a large effect (η²p ≥ 0.14).

Territorial embeddedness is not only statistically significant but also has a large and consistent effect size across all cities and dimensions. This reinforces the conclusion that integrating territory into the value proposition is a practically relevant success factor, rather than a statistical artifact due to large sample size.

Territorial resources also show large effects on Food and Service, suggesting that mobilizing territorial assets (history, landscape, know-how) is an equally powerful strategy.

The other SIAL variables, although conceptually important, have a limited practical impact on customer satisfaction according to these data. Their effect is small and, in many cases, not significant once controlled for effect size.

3.2 Qualitative results: interaction, learning, and self-organization processes

3.2.1 Intermediate projects as spaces for interaction

The visits to 39 coffee shops in three strategic cities (Oaxaca de Juárez, Monterrey, and Guadalajara) and the four cupping sessions organized with participants from four organic fair-trade Small-Scale Producer Organizations (SPOs) constituted the spaces for direct interaction between production- and consumption-side agents.

Observations documented that the primary territorial resource activated is coffee origin, although with differences in the depth of embeddedness. Coffee shops such as Alchemy Coffee Lab, Café Palreal, and Expendio Labranza value the specific origin of their beans (Veracruz, Oaxaca, Chiapas, Nayarit), linking it to quality. However, narratives tend to be generic (referring to the state) or linked to a particular farm, without a story connecting the consumer to the history, landscape, and know-how of the producing community. In contrast, Gato Chato Café de Barrio and Gato Café stand out for offering Oaxacan coffee with extensive knowledge and detailed explanation, indicating a deeper activation of the territorial resource.

Collective activation showed incipient but promising sprouts. At Como Si Fuera Domingo, the barista mentioned that their roaster is “a friend” and that they form a “community,” highlighting a local ecosystem of cooperation between roaster and coffee shop. At Café Palreal, it was mentioned that “one person is a partner in this coffee shop and is the roaster,” demonstrating vertical and cooperative integration. Observations documented a high receptivity among coffee shops to receiving information from SPOs, with establishments such as Lavendula Bake House, Cafetería Victoria, and The Coffee Legacy showing openness to receiving contacts and samples. Nevertheless, the activation observed was predominantly unidirectional or bilateral at the time of the visit, with no evidence of a collective and sustained strategy involving multiple actors in a common territorial narrative.

3.2.2 Emergence of innovations at four interconnected levels

The innovations observed during visits and cupping sessions operated at four interconnected levels:

In coffee valuation: New quality criteria emerged that transcend the traditional focus on roasting or price. Coffee shops classify coffee by “European quality,” “specialty,” “microlot,” “natural process,” and use scores (84.3 points, 83–87 points). This sophistication in valuation criteria represents a fundamental innovation that is creating a segment where SPO coffee, with its territorial narrative and quality attributes, can find a differentiated space.

In service practices: Innovation was evidenced in preparation through brewing methods such as V60, Chemex, Aeropress, and Japanese siphon, as well as in the barista’s specialized knowledge. Alchemy Coffee Lab exemplifies this trend, offering a wide range of methods and “highly developed” knowledge.

In customer interaction: The figure of the “barista-educator” emerged, who explains the origin, roasting, and tasting notes, transforming the consumption experience from a commercial transaction into an educational and emotional encounter.

In SPO narratives: A fundamental learning was documented, expressed in the reflection of one participant: “What type of coffee, roast, and what quality or qualities could my organization sell in Guadalajara?” This questioning reveals that SPOs are recognizing that they cannot offer a single homogeneous product but must adapt to different market niches.

3.2.3 Self-organization and reconfiguration of roles

The observed agent system is composed of a diversity of actors with distinct logics and knowledge: (1) SPOs and producers, who possess knowledge of the territory, production, and the bean, but have limited understanding of the urban market and specialty consumer expectations; (2) roasters, who act as a key “bridge” between both worlds; (3) baristas and coffee shop owners, who interact directly with the consumer and possess knowledge of service, preparation, and market trends; and (4) consumers, who range along a spectrum from traditional to specialty.

Self-organization manifested in the agents’ capacity to create new connections and roles informally. During cupping sessions, producers and SPO representatives assumed a role as “educators” of origin, while baristas became “evaluators” and “critics” of the coffee, generating an exchange of knowledge that was not centrally planned. The openness of coffee shops to receiving cards and samples constituted the germ of informal agreements that could become structured in the future. Thomas Café's willingness “not to change suppliers” also demonstrated self-organization from resistance, where the system stabilizes around relationships of trust and local convenience.

3.2.4 Coevolution and expansion of the possibility space

Coevolution manifested as a two-way process. SPOs are adapting their strategies based on the visit and cupping experience. A member of one SPO acknowledged that they must segment their offering and adapt to different market niches, constituting a fundamental transformation in their perception of the domestic market, shifting from a homogeneous block to a spectrum of opportunities. Concurrently, specialty coffee shops and baristas coevolve with their customers, educating them about new qualities and consumption practices.

The possibility space for SPOs expanded significantly. Prior to the exercise, the domestic market could be perceived as a homogeneous and inaccessible block. Following the interactions, SPOs visualize a spectrum of strategic options: (1) the specialty market, where microlots, processed coffees, and scored coffees can be offered; (2) the intermediate market, where consistently quality coffee with a good story can be offered; (3) the traditional market, where they could compete by offering superior coffee at a competitive price; and (4) new roles, such as the possibility for SPOs to become roasters or establish alliances with local roasters.

3.2.5 Second-order learning

The most important learning derived from the exercise was that “not all coffee shops are the same.” The segmentation of the market according to barista knowledge, decor, customer type, and willingness to pay for origin and quality constitutes second-order learning that implies that the sales strategy must be differentiated and adapted to each niche. SPOs learned that their main asset—the story and origin—is highly valued in a segment of the market, but must be communicated effectively and translated into the language and expectations of the urban specialty consumer.

An SPO member’s observation regarding the need for a “strategic plan” prior to visits and to speak with owners, not only with employees, constitutes another second-order learning that transforms the understanding of the marketing process. Additionally, it was learned that specialty coffee shops do not merely sell coffee, but experiences and narratives; therefore, SPOs must develop their capacity to tell stories and generate Supplementary Material (photos, videos, maps, testimonials) that facilitate this narrative.

For the ecosystem as a whole, it was learned that cooperation and network building are fundamental to overcoming initial distrust and opening market channels previously perceived as inaccessible. The openness shown by coffee shops to receiving information from SPOs suggests that there is an unsatisfied demand for more direct and authentic connections to origin.

3.3 Implications for SIAL construction and value appropriation

The ensemble of results converges on a central finding: SPOs’ entry into the domestic market, through TripAdvisor-visible coffee shops, can yield competitive advantages if their participation is associated with territorial SIAL construction. The quantitative analysis indicates that: (1) territorial embeddedness and specific resources constitute the foundation of the SIAL, with the strongest correlation across all cities; (2) innovation is associated with collective activation; and (3) cooperation and agents constitute the social fabric of the SIAL. The qualitative analysis reveals that coffee shop visits and cupping sessions act as intermediate projects that facilitate the collective activation of specific territorial resources through face-to-face interaction.

The strategy associated with better outcomes for SPOs is an integrated system where territorial embeddedness, cooperation, and agent visibility reinforce one another. Humanizing the value chain—giving face and name to those who participate in the process—emerges as an effective strategy for connecting emotionally with the consumer and building a credible and attractive product story.

3.4 Note on the scope of the results

The quantitative findings presented are based on the analysis of TripAdvisor reviews and therefore characterize the segment of coffee shops with high visibility on this platform, not the entire universe of coffee shops in the cities studied. The qualitative results derive from an intentional selection of coffee shops that showed mentions of cooperation or superior quality in the quantitative phase, implying that the findings are not generalizable to all establishments. The associations reported through Spearman correlations and ANOVA indicate statistical relationships, not causal relationships.

4 Discussion of results

The discussion presented below is organized around the three specific objectives of the research, in order to systematically evaluate how the findings contribute to answering the research questions and to understanding the processes of competitive advantage construction for SPOs in the domestic coffee market.

4.1 Characterization of the coffee shop segment: territorial embeddedness as a structuring axis

The first specific objective sought to characterize the coffee shop segment in eight cities through data mining, identifying value logics and quality demands. The quantitative results reveal that territorial embeddedness is the variable with the highest presence in the reviews (up to 61.0% in Oaxaca) and the one that shows the strongest association with customer satisfaction (p < 0.001 across all dimensions). This finding, although statistically robust, must be interpreted with precision: the Spearman correlation indicates an association, not a causal relationship. It is possible that coffee shops that already possess greater resources and investment capacity are also those that can develop more sophisticated embeddedness narratives, without embeddedness necessarily being the factor that generates satisfaction. As Boucher and Reyes-González (2016) note, the activation of territorial resources is a complex process where causality is multidirectional and depends on the history and specific context of each territory.

Nevertheless, the consistency of the pattern across the eight cities analyzed suggests that territorial embeddedness is associated with value construction in the segment of coffee shops with high visibility on TripAdvisor. The observed territorial variations relate to what Fournier and Muchnik (2012) term the basis of typicality and embeddedness: the conscious mobilization of place-based assets—know-how, landscape, cultural identity—is associated with higher consumer valuation. Cities in Group 1 (Oaxaca, Monterrey, Guanajuato) present particularly high correlations between collective activation and territorial embeddedness (0.780), indicating that a denser network of agents mobilizing the territory may be associated with more intense embeddedness.

The presence of cities without direct producing traditions—such as Monterrey and Tijuana—in groups with different levels of SIAL consolidation suggests that the construction of localized agri-food systems does not depend exclusively on geographic proximity to producing regions. In these urban contexts, the origin narrative and cooperation with producing regions may be related to the substitution of local tradition, underscoring the importance of organized and institutional proximity (Gilly and Torre, 2000) over mere geographic proximity. This finding supports the thesis of Torres and Muchnik (2012) regarding the active role of the market in SIAL construction, a dimension that traditional analyses of localized agri-food systems tend to underestimate. Recent research on SIALs in Latin American countries such as Colombia, Peru, and Guatemala (Boucher and Reyes-González, 2016; Samper, 2016) has documented similar processes where specialized urban markets become nodes of territorial value construction, even in the absence of local producing traditions.

4.2 Intermediate projects and learning processes: from information to transformation of valuation frameworks

The second specific objective was oriented toward documenting interaction practices between SPOs and market actors, identifying second-order learning. The intermediate projects—coffee shop visits and cupping sessions—emerge as key spaces where the encounter between the fair-trade niche and the conventional market regime occurs, following Smith’s (2012) conceptualization of translation spaces between niches and sociotechnical regimes.

Participant observation documented that these spaces operate as living laboratories where innovations emerge at four interconnected levels: quality criteria, service practices, customer interaction, and SPO narratives. The emergence of new practices and languages—such as classification by SCA scores, brewing methods, or the concept of the “barista-educator”—was not centrally planned but arose from local interactions among agents, constituting an empirical manifestation of the concept of emergence in complex systems (Kauffman, 1995; Bohórquez, 2013). As Lara-Rosano (2025) notes, in complex adaptive systems, innovation is not the result of centralized design but of local interactions that generate unanticipated global patterns.

The documented second-order learning—particularly the segmentation of the market into differentiated niches and the need for a strategic plan prior to visits—represents a fundamental transformation in producers’ valuation frameworks. As Smith (2012) points out, second-order learning implies a change in valuation frameworks, not merely the acquisition of new information. SPOs transitioned from perceiving the domestic market as a homogeneous and inaccessible block to visualizing a spectrum of differentiated niches—specialty, intermediate, traditional—and new strategic roles (producer, roaster, ally). This shift in their possibility space (Kauffman, 1995) is fundamental because it allows them to design differentiated strategies, transcending the logic of raw commodity sales. The literature on complex systems applied to social economy organizations in Latin America (Bohórquez and Espinosa, 2015; Earls, 2013) has documented how second-order learning is a central mechanism for the adaptation and survival of these organizations in changing market environments.

However, it is necessary to recognize that the selection of coffee shops for visits was intentional, prioritizing establishments that showed mentions of cooperation or superior quality in the quantitative phase. This implies that the documented interaction processes occurred in a particularly receptive subset of the market and are not necessarily representative of the dynamics that SPOs would encounter in the entire universe of coffee shops. This selection bias limits the generalizability of the qualitative findings, although it does not invalidate the richness of the documented processes. As Correa et al. (2006) suggest, SIAL research must explicitly recognize the tension between the study of successful cases—which allow for understanding activation mechanisms—and the need for analyses that capture the diversity of trajectories, including those that fail to consolidate territorial systems.

4.3 Value appropriation and construction of competitive advantages: the challenge of coevolution

The third specific objective sought to evaluate how the documented learnings enable SPOs to capture greater value, transcending raw commodity marketing logics. The results show that self-organization—manifested in the spontaneous reconfiguration of roles during cupping sessions, where producers assume the role of educators and baristas that of evaluators—constitutes a key mechanism for the generation of innovations and the exchange of knowledge. This process, which operates at what Kauffman (1995) terms the edge of chaos, is the breeding ground for innovation, as it allows for the exploration of new configurations without the rigidity of centralized design. Research on self-organization in small-scale producer organizations in Mexico and Central America (Boucher and Reyes-González, 2016; Coscione, 2018) has documented similar processes, where the informality of initial agreements constitutes the germ of more structured governances that emerge from below.

The observed coevolution is a two-way process: SPOs adapt to the value logics of the market, while coffee shops and baristas, by educating their customers, are coevolving with them, creating a segment where SPO coffee can find a differentiated space. This coevolutionary process allows SIAL agents to exit their inertial attractor (Lara-Rosano, 2025)—the raw commodity logic—and explore new states of higher fitness in the adaptive landscape of the market. The segmentation of the offering and the need for a strategic plan are manifestations of this navigation. The literature on complex systems and social economy (Bohórquez, 2013; Kelly and Allison, 1999) indicates that the capacity to coevolve is a fundamental competitive advantage in high-uncertainty environments, such as the coffee shop market in Mexico.

A finding that merits critical discussion is the contrast between the strength of the association of territorial embeddedness with customer satisfaction and the weakness of the association of cooperation (which did not reach statistical significance in any satisfaction dimension). This result appears to contradict the literature that emphasizes cooperation as the core of the SIAL (Boucher and Reyes-González, 2016; Correa et al., 2006). However, this apparent contradiction can be interpreted in two complementary ways. First, cooperation may be a mechanism that operates behind the scenes: coffee shops that achieve effective territorial embeddedness have likely already established cooperative relationships with producers and roasters, but these relationships are not necessarily visible to the consumer in the review. As Fournier and Muchnik (2012) note, cooperation is a foundational process that sustains the activation of territorial resources, but it is not always directly perceptible to final consumers. Second, the final consumer values the outcome of cooperation (territorial embeddedness and origin narrative) more than the cooperation process itself. This suggests that for SPOs, investing in communicating the story and origin of the coffee—the visible outcome of cooperation—may be more effective for increasing perceived value than communicating the cooperation mechanisms themselves. This interpretation resonates with the notion of relative or specific quality defined by Sylvander and Melet (1992), as cited in Tibério et al. (2001), where the consumer values attributes that they can directly perceive and experience.

4.4 Limitations and scope of the findings

The findings must be interpreted in light of significant methodological limitations. The selection of coffee shops was based on their visibility on TripAdvisor, implying that the results characterize the segment of coffee shops with high visibility on this platform, not the entire universe of coffee shops in the cities studied. TripAdvisor users constitute a specific subset of consumers, with profiles that may not be representative of the total coffee consumers in Mexico. This limitation affects the generalizability of the quantitative results and suggests that future research should incorporate more diverse samples of establishments and consumers.

The consistency and magnitude of the effect of territorial embeddedness (η²p between 0.17 and 0.22 across all dimensions) may be influenced by the fact that TripAdvisor-visible coffee shops have already invested in building these narratives, while less visible ones may not do so or may do so less effectively. In a study with a more diverse sample, the effect size could be smaller.

Additionally, the lexical classifier used for review content analysis, although validated with 84% precision, has a margin of error that may have affected the classification of some comments. The thematic coding process for the qualitative analysis, although it followed the phases proposed by Braun and Clarke (2006), did not include an explicit double-coding process with inter-coder agreement calculation, introducing a degree of subjectivity in the identification of emerging themes. For future research, it is recommended to implement inter-coder validation procedures to strengthen the reliability of qualitative analysis.

The bounded timeframe of the fieldwork (2024–2026) does not allow for capturing long-term coevolutionary processes; the documented learnings correspond to specific moments of interaction and may not be sustainable over time. Finally, the small size of the qualitative sample (10 participants from four SPOs) limits the capacity to generalize findings on learning and self-organization processes to other SPOs with different characteristics. As Correa et al. (2006) note, SIAL research in Latin America faces the challenge of balancing the depth of case studies with the need for findings that can be relevant to public policy and organizational strategies at a larger scale.

The methodological limitations noted have important implications for the interpretation of the findings:

Findings on self-organization and second-order learning are contextual: These processes emerge under specific conditions of market receptivity. Their absence in other contexts does not mean that SPOs cannot generate self-organization, but rather that the process may require particular conditions or more time to develop.

The strategy of humanizing the value chain may not be universally effective: Although it emerges as an effective strategy in the studied segment, its effectiveness could depend on the type of coffee shop and consumer. In less specialized markets, other strategies (such as price or convenience) could be more relevant.

Need for complementary research: The findings suggest the need for studies that include coffee shops with lower visibility on digital platforms and lower receptivity to origin, in order to understand whether the processes of self-organization and territorial embeddedness are generalizable or depend on specific market conditions.

In conclusion, although the findings are robust within the studied segment, their generalization must be made with caution. The research successfully characterizes a particular market niche—coffee shops with high visibility on TripAdvisor and with mentions of cooperation—but the documented processes may not be representative of market dynamics as a whole. This does not invalidate the richness of the findings, but it does require that SPOs and policymakers consider that strategies effective in this niche may need adaptations to function in other market segments.

4.5 Implications for theory and practice

Despite these limitations, the convergence of quantitative and qualitative analyses allows us to affirm that SPOs’ entry into the domestic market, through the segment of TripAdvisor-visible coffee shops, benefits from territorial SIAL construction. SPO competitiveness does not lie solely in bean quality, but in their capacity to articulate a system of actors, resources, narratives, and innovations that anchor coffee to its territory. The fundamental learning that not all coffee shops are the same and that their main asset—the story and origin—is highly valued in a segment of the market constitutes a transformation in producers’ valuation frameworks. Complex adaptive systems theory (Kauffman, 1995; Lara-Rosano, 2025) offers a useful framework for understanding how these transformations emerge from local interactions and consolidate into patterns of higher fitness in the adaptive landscape of the market.

For practice, this implies that SPOs must develop capacities in areas such as effective communication of the coffee’s story, generation of Supplementary Material (photos, videos, maps, testimonials), and design of differentiated strategies for distinct market niches. Self-organization and coevolution, although processes that emerge spontaneously, can be facilitated through the creation of interaction spaces—intermediate projects—that allow for face-to-face encounters between producers and market actors. The fragility of these processes, however, suggests that SPOs require accompaniment and capacity development to manage self-organization and maintain the flexibility necessary to navigate between order and chaos, without falling into the rigidity of mass production or the uncertainty of improvisation. As Boucher and Reyes-González (2016) note, the collective activation of territorial resources is not a one-time event but a process that requires the continuous strengthening of organizational capacities and the construction of trust networks.

In synthesis, the path toward greater value appropriation lies in the consolidation of second-order learning and the construction of strategic alliances with market actors who value authenticity and territoriality, in a process that transforms both producers and the market itself. The research suggests that SPOs’ capacity to successfully articulate these elements will determine their ability to compete in the Mexican coffee shop market, transcending raw commodity sales logics and developing advantages based on emergent differentiation, coevolutionary adaptability, and strategic autonomy. The literature on SIALs in Latin America (Boucher et al., 2012; Correa et al., 2006; Torres and Muchnik, 2012) coincides in pointing out that the sustainability of these systems depends on local actors’ capacity to articulate with dynamic markets without losing their territorial identity—a challenge that coffee SPOs in Mexico are beginning to address through the processes documented in this research.

5 Conclusion

Territorial embeddedness is the most relevant element for SPO competitiveness in the domestic coffee market. A total of 74.82% of reviews contain mentions of origin. Oaxaca represents the most consolidated SIAL (CRA-EI = 0.608), while Tijuana demonstrates that SIALs can be constructed through specialized demand and organized proximity (CRA-EI = 0.379), transcending geographic dependency.

The visits and cupping sessions functioned as living laboratories where innovations and self-organization germinated. Innovation was associated with collective activation (rho = 0.243) and the presence of diverse agents (rho = 0.209). Producers assumed the role of educators and baristas that of evaluators, generating innovations at four levels: quality criteria, service practices, customer interaction, and SPO narratives.

SPOs recognized that the domestic market is not homogeneous, transitioning from perceiving it as an inaccessible block to visualizing differentiated niches (specialty, intermediate, traditional) and new strategic roles (producer, roaster, ally). Cooperation, although not statistically significant (η²p < 0.01), operates as an invisible foundational process that sustains collective activation and territorial embeddedness.

The spontaneous reconfiguration of roles evidences the capacity of agents to create patterns of collaboration without external direction, operating at the “edge of chaos.” Coevolution allows for the exploration of new states of higher fitness in the adaptive landscape of the market, adapting to value logics while shaping a new consumer.

SPO competitiveness lies in articulating an integrated system of actors, resources, narratives, and innovations that anchor coffee to its territory. Humanizing the value chain—giving face and name to producers—is an effective strategy for connecting emotionally with the consumer. The path toward greater value appropriation lies in the consolidation of second-order learning and the construction of strategic alliances with actors who value authenticity and territoriality.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. Generative AI was used to identify the relationships among the three types of information analysis employed in the document, as well as to evaluate the coherence between the components of the manuscript. Generative AI was also used to translate the manuscript into academic United States English.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Supplementary material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fcpxs.2026.1830387/full#supplementary-material

Supplementary material Supplementary material at the source

Footnotes

Edited by: Theano Iliopoulou, Harokopio University of Athens, Greece

Reviewed by: Siti Nurhayati Khairatun, Putra Malaysia University, Malaysia

Carlos Ríos-Llamas, Autonomous University of Baja California, Mexico