2.1. Sharing economy

Sharing economy (SE) platforms have diversified considerably in recent years, attracting growing attention from academics, policymakers, and practitioners [11]. The concept has been described using various terms, including the platform economy, the sharing economy, on-demand platforms, the access economy, and the peer-to-peer economy [12], reflecting the expansion and complexity of the phenomenon [13]. The SE can be defined as the coordination of exchanges among individuals to acquire and distribute resources in return for compensation, which may be financial or nonfinancial [14]. This business model generally operates through online platforms that allow individuals to bypass traditional commercial channels and share underutilized resources, thereby reducing transaction costs [15].

The literature generally recognizes that the sharing economy generates economic, social, and environmental impacts that can contribute to sustainable development [16].

In terms of economic impacts, the SE fosters entrepreneurship, job creation and growth, with profit being the main motivation for participation [17,18]. Furthermore, cost reduction plays a key role in participation [19].

From a social perspective, the sharing economy can foster relationships within communities [20]. It may also facilitate access to services and opportunities for exchange while creating social bonds among participants [21]. The rapid development of the internet, mobile devices, and artificial intelligence has played an important role in the growth of sharing economy platforms [22].

However, digitalization has not led to equal participation across population groups. Access to and effective use of these platforms depend not only on technological infrastructure but also on users’ digital skills and resources. This disparity reflects the broader issue of digital inequality, which is discussed in the following section.

2.2. Digital divide and skills

The digital divide refers to disparities in the ability of individuals and groups to access and effectively use digital technologies [23]. These disparities can reinforce socioeconomic inequalities and limit employment opportunities worldwide. Their effects may be particularly significant for young people, who may face exclusion from the labor market and limited access to the educational opportunities needed to participate in an increasingly digitalized economy [24].

Van Dijk [25] distinguishes three cumulative levels of the digital divide. The first-level digital divide (1995–2003) concerns physical access to hardware and internet connectivity. The second-level digital divide (2004–present) focuses on differences in digital skills and patterns of use. This distinction, introduced by Hargittai [26], moves beyond binary inequalities in access to consider differences in what people do online. The third-level digital divide, identified from 2012 onward, concerns differences in the extent to which internet use translates into positive offline outcomes [27]. In technologically advanced countries, fewer than 10% of the population lack basic internet access [25].

In this context, digital competencies are considered a key component of the second-level digital divide and can be defined as the skills required to use digital technologies effectively in everyday life [28]. Related terms, such as digital skills and digital literacy, are often used interchangeably to describe similar capabilities [29].

The European Commission defines digital competence as the safe, critical, and responsible use of and engagement with digital technologies for learning, work, and participation in society. It encompasses information and data literacy, communication and collaboration, media literacy, digital content creation, safety—including digital well-being and cybersecurity—an understanding of intellectual property, problem-solving, and critical thinking [30].

As digital technologies become increasingly central to economic and social participation, differences in digital skills can shape people’s ability to benefit from the opportunities they offer. These differences may reinforce existing inequalities, particularly in the context of SE platforms, where participation depends not only on access but also on the skills needed to use these platforms effectively.

2.3. Digital divide and skills in the sharing economy

Despite the growing body of research on digital inequalities and the sharing economy, the intersection between these two fields remains relatively underexplored. Existing studies have focused primarily on transportation and accommodation platforms, using methods such as interviews, case studies, index-based analyses, and national surveys. Although a limited number of studies have examined digital skills, most of the literature focuses on the effects of digital literacy on workers and companies that provide these services.

Early evidence showed how basic access barriers can limit participation in the sharing economy. Dillahunt et al. [31] documented cases in which individuals using basic mobile phones without internet access were unable to use Uber, illustrating how the digital divide can translate into exclusion from sharing economy services. Quirós et al. [32] found that advanced digital skills and regular e-commerce use were associated with a greater likelihood of using Uber, while Erazo et al. [33] found that gender and income were associated with different patterns of platform use in the accommodation sector.

These findings suggest that digital access, skills, and socioeconomic conditions are associated with how individuals participate in specific sharing economy platforms. However, a fuller understanding of these inequalities also requires consideration of the broader operational and service-related conditions that shape shared mobility systems, particularly car-sharing services.

Turoń [34] identifies 151 quantitative and qualitative criteria affecting car-sharing services during their implementation and operation and classifies them into six groups: economic and technical, transportation, social, environmental, organizational, and other factors. These criteria cover organizational conditions, the characteristics of current and potential users, service costs, transportation infrastructure, and the demographic and social structure of cities [34]. Although this framework provides a comprehensive overview of the conditions affecting car-sharing systems, it focuses mainly on their implementation and operation. Digital skills are not explicitly considered as a separate factor or category. The criterion most closely related to digital inequality is the “accessibility of the IT system for the user,” which is classified as an organizational factor. This limitation highlights the need to complement system-level analyses with a user-level perspective that considers digital skills and inequalities in platform access and use.

At the individual level, other studies have examined the factors associated with the adoption of shared mobility services. In the context of on-demand ride services, Alemi et al. [35] conducted an empirical study using the California Millennials Dataset, collected through an online survey of California residents in fall 2015. The sample included millennials aged 18–34 and members of Generation X aged 35–50 from six regions of the state, including urban, suburban, and rural areas. The findings showed that services such as Uber and Lyft were more likely to be adopted by older millennials—that is, individuals aged 25–34—and by those with at least a bachelor’s degree. Adoption was also associated with previous experience using taxi and car-sharing services, frequent smartphone use for transportation-related decisions, and more positive attitudes toward technology. These findings suggest that greater familiarity with digital technologies in everyday life was associated with a higher likelihood of adopting these services [35].

More recent research has shown that barriers to shared mobility are perceived differently across population groups, including both users and nonusers. Chahine et al. [36] conducted an online survey of adult residents of Indianapolis to examine perceptions of ride-hailing, bike-sharing, and shared e-scooter services. Their analysis considered two perceived benefits—improved connectivity with public transit and reduced parking demand—and two perceived barriers: service cost and technology-related difficulties. The largest groups reported predominantly neutral or indifferent views, suggesting that a substantial share of the population had not fully embraced shared mobility services. In contrast, a smaller “barrier-conscious” group indicated that lower costs and fewer technology-related obstacles could encourage greater use. Across most of the identified groups, safety and reliability were also considered important factors when choosing a mode of transportation [36]. These differences in the perceived benefits and barriers of shared mobility are consistent with broader patterns of participation in the sharing economy.

At a broader level, Lutz et al. [9] found that demographic factors, including age, gender, education, income, and household size, as well as psychological factors such as trust, digital skills, and the frequency of online searches, are important predictors of participation in SE platforms. Previous research has also identified significant differences across population groups in the types and range of activities performed online [37]. In particular, socioeconomically advantaged individuals tend to use the internet more frequently for information-intensive and capital-enhancing activities, whereas less advantaged users are more likely to use it for recreational purposes, such as entertainment [38].

These patterns are also reflected in participation in the sharing economy. Because such participation is mediated by digital technologies and often requires users to search for and evaluate information, it is positively associated with online search skills and behaviors [9]. More recent research by Lutz et al. [39], also identified differences by age and place of residence, with younger urban residents perceiving short-term rental platforms as more beneficial than older and rural residents.

Eichhorn et al. [10] examine digital inequality in a commercial setting, focusing on participation in SE platforms. They analyze how structural social conditions shape different stages of platform access and use. Their findings indicate that van Dijk’s resource-based model of internet use [40] can be applied to explain online participation in the sharing economy.

2.4. Research gap

Although empirical research has established a link between digital inequalities and platform participation, important gaps remain. Much of the existing literature has relied primarily on quantitative analyses, providing limited insight into users’ subjective experiences and perceptions. As a result, the roles of digital skills, internet access, and platform usability have not been examined in sufficient depth from a user-centered perspective.

This study addresses these gaps by exploring how digital skills and access conditions shape individual participation in the sharing economy and how individuals perceive and experience inequalities in platform-based environments. To guide the empirical analysis, the study draws on a set of research propositions derived from the literature reviewed:

• First, limited digital skills and restricted internet access are perceived as significant barriers to effective participation in sharing economy platforms.
• Second, economically disadvantaged individuals face greater barriers to accessing and benefiting from sharing economy services because of limited digital resources.
• Third, digital skills, internet use, and the frequency of online searches are considered important conditions for effective participation in the sharing economy.
• Finally, users identify specific usability and accessibility challenges in existing sharing economy platforms that may hinder broader and more inclusive participation.

3.1. Data collection

Focus groups were conducted to capture a range of perspectives on the research topic and to understand the issues from the participants’ perspectives. The group setting facilitates the collection of diverse views within a single session [41]. Fern [42] found that a single focus group can identify approximately 70% of the themes that could be obtained from an equivalent number of in-depth interviews, highlighting the efficiency of this method for initial conceptual exploration. The study did not seek statistical representativeness; rather, it aimed to achieve theoretical richness and analytical sufficiency.

Given the exploratory nature of the study, we adopted a qualitative approach to gain an in-depth understanding of individuals’ perceptions of participation in sharing economy platforms. Focus group discussions were conducted using a semi-structured guide organized around four themes: the role of digital skills; access to and barriers within the sharing economy; the social and economic implications of digital barriers; and proposals for improving platform accessibility and inclusivity. The study was guided by an interpretivist perspective, focusing on participants’ subjective experiences and the meanings they attributed to their interactions with sharing economy platforms.

Consequently, two focus groups were conducted, each with seven participants. The same semi-structured discussion guide, consisting of open-ended questions, was used in both sessions (see S1 File) to encourage participants to express their views. The corresponding author reviewed and coded the two transcripts sequentially using the Gioia methodology, beginning with FG1 and then proceeding to FG2. The coding process was documented in an Excel matrix that linked each excerpt to its focus group, participant, first-order concept, second-order theme, and aggregate dimension. The other two co-authors then reviewed the coding. Before combining the data for reporting, the three authors compared the coding structures derived from the two groups. The analysis of FG2 yielded the same five aggregate dimensions identified in FG1 and did not produce any additional higher-order dimensions. Although FG2 provided further examples, nuances, and second-order themes, these were incorporated into the existing dimensional structure. Drawing on Hennink et al.’s [43] distinction between code saturation and meaning saturation, and considering the stability of the five aggregate dimensions across both groups, the three authors jointly concluded that saturation had been reached at the level of the aggregate dimensional structure. After reviewing FG2, they therefore decided that a third focus group was not necessary. The data were considered sufficient to address the exploratory research question at this analytical level, without claiming that meaning saturation had been achieved.

The focus groups were led by a facilitator who guided the discussions, while another researcher served as an observer, monitoring group dynamics and taking notes. All participants provided written informed consent before the sessions began. With appropriate measures in place to protect confidentiality, both sessions were audio-recorded, yielding a total of 45 pages of transcribed material. The Ethics Committee for R&D + I Activities at the Universidad Politécnica de Madrid determined that the study did not require Institutional Review Board review or approval because no personal data were used in the research.

3.2. Selection of participants

Fourteen adults participated in two focus groups conducted in July 2024. Participants were recruited between June 3 and July 14, 2024, using a combination of purposive and convenience sampling. The corresponding author contacted potential participants by email and WhatsApp through professional and institutional networks. Before providing consent, invitees received general information about the study topic and purpose, the voluntary nature of their involvement, and the recording of the sessions. The information provided was intentionally broad to avoid influencing their responses during the focus groups.

Eligible participants were between 25 and 64 years of age and resided in Madrid, Spain, at the time of data collection. The lower age limit was established because the study focused on adults who were currently or potentially active in economic and social contexts and who had varying levels of exposure to sharing economy platforms. The upper age limit reflected the initial scope of the recruitment strategy rather than an assumption that adults over 64 were unaffected by digital barriers. Participants had either current or previous experience using at least one sharing economy platform or had personally encountered barriers that prevented them from using such platforms.

Recruitment sought variation in levels of digital proficiency, ranging from novice to advanced users, as well as diversity in participants’ occupational and educational backgrounds; participants included individuals with experience in business administration, technology education, digital skills training, and academia. Despite this strategy, the final group included a comparatively high proportion of participants with higher education and professional experience. The findings should therefore not be interpreted as representative of platform users or nonusers in Madrid or Spain. Rather, the study was designed to provide exploratory qualitative insights into how usability, access to technology, digital skills, trust, and other usage limitations may shape participation in sharing economy platforms.

Participants attended one of the two sessions based on their availability on the scheduled dates. Both groups included individuals with varying levels of digital experience and diverse professional backgrounds, and the data from both sessions were analyzed jointly. Table 1 summarizes the participants’ main characteristics and the criteria supporting their inclusion in the study.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Characteristics of the participants.

https://doi.org/10.1371/journal.pone.0357784.t001

3.3. Data analysis

The discussions were transcribed and carefully reviewed to correct transcription and spelling errors and remove filler words. Participants’ identities were anonymized by replacing their names with unique research codes in the format FGx-Pi, where x indicates the focus group number and i identifies the participant. The data were then analyzed using the Gioia methodology, complemented by natural language processing (NLP) techniques as an exploratory analytical tool. This combination supported a structured interpretation of participants’ perspectives while providing additional exploratory insights into recurring patterns in the textual data.

3.3.1. Gioia Methodology.

The primary data analysis was conducted using the Gioia methodology, which provides a structured approach to inductive qualitative research [44]. This methodology involves organizing the data into first-order concepts, second-order themes, and aggregate dimensions. Coding is a central part of the analytical process and involves systematically reviewing and interpreting field notes, transcripts, or other qualitative materials while preserving the relationships among the different parts of the data [45].

Both focus group transcripts were first read in full to develop an overall understanding of the discussions. Coding was then conducted sequentially, beginning with FG1 and followed by FG2. Relevant passages were identified as units of meaning and assigned first-order labels that remained close to the participants’ own words and the meanings they conveyed, in accordance with the Gioia methodology [44]. This process yielded 330 first-order concepts. The coding was documented in an Excel matrix that recorded the focus group, participant identifier, original quotation, first-order concept, second-order theme, and aggregate dimension associated with each excerpt. This procedure maintained clear traceability between the original data and each level of the final analytical structure.

In the next phase of the analysis, the first-order concepts were compared and grouped into 64 second-order themes. This process involved examining similarities, differences, and recurring patterns across the first-order concepts to organize them into a more manageable set of categories. The emerging themes were then evaluated according to their ability to describe and explain the phenomena reflected in participants’ accounts [44]. To support this process, pivot tables were created for each emerging dimension. These tables facilitated the systematic organization and comparison of related concepts, the identification of overlaps and redundancies, and the assessment of coherence within each grouping. They were not used to automate the coding process but rather as an analytical support tool that allowed the researchers to examine conceptually similar labels together and refine the boundaries and names of the second-order themes. When necessary, overlapping labels were merged, and coded excerpts were reassigned to the theme that most accurately represented their meaning.

Finally, the second-order themes were grouped into five aggregate dimensions, providing a structured interpretation of the data. These dimensions were usability, technology, skills, trust, and other usage limitations. The resulting analytical structure was used to develop a data structure diagram that visually represents the relationships among the first-order concepts, second-order themes, and aggregate dimensions.

The initial coding was conducted by the corresponding author. The preliminary coding and emerging data structure were then reviewed by the other two co-authors. This collaborative review covered the grouping of first-order concepts, the formulation and naming of second-order themes, and the definition of the aggregate dimensions. Differences in the interpretation, naming, or classification of concepts and themes were discussed jointly and resolved by consensus. The analytical structure was refined through this iterative process until all three authors agreed that it adequately reflected the participants’ accounts and was conceptually coherent across the different levels of analysis.

The initial coding was conducted by the corresponding author. The preliminary coding and emerging data structure were then reviewed by the other two co-authors. This collaborative review examined the grouping of first-order concepts, the formulation and naming of second-order themes, and the definition of the aggregate dimensions. Differences in the interpretation, naming, or classification of concepts and themes were discussed jointly and resolved by consensus. The analytical structure was refined through this iterative process until all three authors agreed that it adequately reflected the participants’ accounts and was conceptually coherent across the different levels of analysis.

The final data structure was reviewed by tracing the analytical progression from the original participant quotations to the first-order concepts, second-order themes, and aggregate dimensions; examining the relationships among the three analytical levels; and comparing the coding structures of FG1 and FG2 before combining the data from both groups for reporting. FG2 contributed additional examples, nuances, and second-order themes, all of which were incorporated into the five-dimensional structure identified in FG1 and did not require the development of an additional aggregate dimension. Table 2 provides illustrative examples of how participant quotations informed the development of first-order concepts, second-order themes, and aggregate dimensions.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Illustrative examples of the coding structure (from quotation to aggregate dimension).

https://doi.org/10.1371/journal.pone.0357784.t002

4.1. Dimensions of user experience on sharing economy platforms

The analysis of the focus group discussions identified five aggregate dimensions: usability, technology, skills, trust, and other usage limitations. Fig 1 presents the data structure developed following the Gioia methodology, including the first-order concepts, second-order themes, and aggregate dimensions.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 1. Data structure diagram using the Gioia Method.

https://doi.org/10.1371/journal.pone.0357784.g001

4.2. Complementary text analysis using NLP

Building on the qualitative findings presented in the previous section, word-frequency analysis, lemma-frequency analysis, and named entity recognition (NER) were used as complementary exploratory procedures to examine recurring lexical patterns in the focus group transcripts. The five dimensions identified through the Gioia analysis provided the primary interpretive framework, while the NLP analysis offered an additional, transparent overview of the words, lemmas, platform names, and geographic references that appeared repeatedly in the discussions. These procedures were not intended to statistically validate the qualitative findings but rather to make recurring elements of the textual corpus more visible. The results are presented in bar charts showing the most frequent words in the transcripts (Fig 2), the most frequent lemmas obtained through lemmatization (Fig 3), and the most frequent named entities classified as locations (LOC) and organizations (ORG) (Fig 4).

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Most frequent terms in the focus group transcripts.

https://doi.org/10.1371/journal.pone.0357784.g002

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Most frequent lemmas in the focus group transcripts.

https://doi.org/10.1371/journal.pone.0357784.g003

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. Most frequent named entities.

https://doi.org/10.1371/journal.pone.0357784.g004

Fig 2 shows the most frequent words in the focus group transcripts. The most frequent word was “platforms,” with 100 occurrences, followed by “people” [37], “skills” [29], “problem” [29], “Uber” [29], “data” [28], “internet” [27], “digital” [25], “economy” [25], and “card” [22]. The recurrence of these words was consistent with several issues identified in the qualitative analysis. The terms “skills,” “digital,” and “internet” appeared in participants’ accounts of digital competence and access conditions, while “data” was frequently associated with concerns about trust and the disclosure of personal information. References to “card” appeared mainly in discussions of payment requirements and financial barriers, whereas “economy” was commonly used in references to the sharing economy.

The frequent occurrence of words such as “platforms” and “problem” indicates that digital platforms and difficulties associated with their use appeared repeatedly in the discussions. Participants frequently referred to platforms such as Uber, Booking, and Airbnb when describing their experiences, often in connection with usability, access, and the digital divide. Similarly, the recurrence of “skills” was consistent with participants’ accounts, in which digital competence emerged as an important condition for accessing and using these platforms.

Fig 3 presents the most frequent lemmas identified through the NLP analysis. Lemmatization groups inflected word forms under their corresponding base forms. The most frequent lemma was “platform,” with 100 occurrences, followed by “digital” [46], “problem” [42], “has” [42], “see” [40], “person” [40], “skill” [38], “people” [37], “use” [34], and “application” [33]. The recurrence of lemmas related to platforms, individuals, skills, use, and problems indicates that these elements appeared repeatedly in participants’ discussions of platform participation. This lexical pattern was consistent with the qualitative accounts of usability and digital skills barriers. However, the frequency counts alone do not establish relationships among the terms or independently confirm the qualitative dimensions.

Fig 4 presents the most frequent named entities classified as locations (LOC) and organizations (ORG). The location entities included references to Spain, Europe, Madrid, Brazil, and Peru. These places served as contextual reference points when participants described or compared experiences related to connectivity, platform availability, payment practices, travel, and access conditions. Their occurrence helps situate the experiences discussed in the focus groups but should not be interpreted as evidence of broader geographic patterns or the global expansion of sharing economy platforms.

Regarding organizations (ORG), the most frequently mentioned names included Booking, Airbnb, Cabify, Uber, and BlaBlaCar. Participants referred to these platforms as concrete examples when discussing registration, navigation, payments, trust, accessibility, and customer support. Their recurrence helps link the qualitative findings to the specific platforms discussed in the focus groups, without implying market leadership, broader brand recognition, or popularity beyond the study corpus.

The recurrence of these company names identifies the platforms that were most prominent in this particular corpus. More broadly, the LOC and ORG classifications provide a descriptive overview of the organizations and geographic contexts referenced by participants, without supporting broader conclusions beyond the focus group discussions.

Overall, the NLP analysis provided a complementary descriptive layer to the qualitative findings by making recurring words, lemmas, platform names, and geographic references more visible. These patterns highlighted recurring language associated with usability, technology, skills, trust, and other conditions of use across the transcripts. Nevertheless, the interpretation of the findings remained grounded in the Gioia analysis, participants’ quotations, and the contexts in which their statements were made.

5.1. Usability

Participants described several basic platform functions as intuitive and user-friendly, including requesting a ride, booking accommodation, and filtering results. Within this sample, these accounts suggest that many platforms provide relatively accessible entry points for routine tasks. However, as platforms expand into multiservice ecosystems that combine transportation, marketplaces, and financial services, their use may become more demanding, requiring users to adapt to new features and increasingly complex interfaces. This evolving environment is consistent with van Dijk’s [25] characterization of the second-level digital divide as dynamic: narrowing an initial skills gap does not necessarily ensure continued ease of use because technologies, and the demands they place on users, continue to evolve.

Social influence also emerged as an important factor shaping decisions to adopt or avoid a platform. Recommendations from peers and the positive or negative experiences of others appeared to influence participants’ willingness to try a particular service. These accounts suggest that usability is not assessed solely through direct interaction with an interface but may also be shaped by the experiences and opinions circulating within users’ social networks. However, this small qualitative sample does not allow the relative influence of peer recommendations, advertising, and interface design to be determined. This finding should therefore be considered exploratory.

5.2. Technology

Participants generally perceived that barriers associated with the first-level digital divide, particularly limited internet access, have become less restrictive in urban areas of Spain because of the wider availability of public Wi-Fi and more affordable mobile data plans. However, these perceptions should be interpreted cautiously because all participants were residents of Madrid, and the study did not directly include rural or peri-urban users. The discussions nevertheless suggested that improvements in urban connectivity may coexist with persistent territorial inequalities, particularly in areas where network coverage and service availability remain limited. In such contexts, sharing economy platforms may be difficult or impossible to access rather than merely less attractive to potential users. This interpretation is consistent with Lutz et al. [39], who identified differences between urban and rural areas in the perceived benefits of the sharing economy and noted that limited service availability may itself constitute a barrier in some rural contexts.

Participants also referred to individuals who may be unable to afford smartphones or mobile data plans. These accounts suggest that economic constraints can create barriers at the level of basic access, before users encounter difficulties related to digital skills, platform usability, or trust. However, individuals experiencing the most severe forms of material and digital exclusion were not represented in the sample. This interpretation is consistent with Dillahunt et al. [31], who showed how limited access to digital devices and internet connectivity can result in exclusion from platform-based services.

5.3. Skills

Participants described age-related differences in how individuals approached and navigated digital platforms. Younger users were generally perceived as more familiar with digital technologies and more likely to find platform use intuitive, whereas older users were described as potentially requiring more time, support, or assistance. However, adults aged 65 or older were not represented in the sample. These observations should therefore be interpreted primarily as participants’ perceptions and accounts involving relatives or acquaintances rather than as direct evidence from older platform users. This pattern is consistent with Lutz et al. [9], who identified digital skills as an important predictor of participation in the sharing economy, and with Hidalgo et al. [24], who reported that age and educational attainment were key predictors of digital skills gaps in Spain.

The discussions also highlighted self-directed learning through experimentation, trial and error, and gradual adaptation as an important way in which participants developed digital competence. Rather than showing that informal learning is more effective than formal digital literacy interventions, the findings suggest that experiential learning played a prominent role in this particular sample, which included a relatively high proportion of participants with higher educational attainment and professional experience. For individuals with lower educational attainment, limited digital access, or less confidence in using technology, however, self-directed learning may be more difficult and may not provide adequate support. At the same time, the ongoing evolution of SE platforms may make previously acquired skills less sufficient over time, creating a continuous need for adaptation that may place greater demands on users with less time, confidence, or access to assistance.

Younger participants also suggested that frequent use or apparent ease of use did not necessarily imply a deeper understanding of how platforms operated. Some accounts indicated that users could complete routine tasks without fully understanding how personal data, payments, permissions, or security settings were managed. This distinction suggests that operational familiarity and critical digital competence should not be treated as equivalent. Even users who navigate platforms with ease may remain exposed to risks that they do not fully recognize or assess.

5.4. Trust

The security of personal data, the risk of payment fraud, and the perceived vulnerability of some groups, particularly older adults, were recurring concerns in the discussions. Participants described how their fear of fraud had increased over time as they became more aware of negative experiences involving friends, relatives, or others within their social networks. In several cases, this concern appeared to be shaped not only by direct experience but also by accounts shared by people around them. These findings suggest that trust in SE platforms may be influenced by socially shared perceptions of risk as well as by personal experience.

The analysis also identified several precautionary strategies that participants used to manage these concerns, including using cards with low credit limits for online payments, avoiding public Wi-Fi networks when conducting financial transactions, and limiting their use to familiar or well-established platforms. Within this sample, these practices suggest that mistrust did not necessarily lead participants to abandon platforms altogether. Instead, some adapted their behavior in ways that allowed them to continue using these services while attempting to reduce the perceived risks.

Finally, participants expressed particular concern about older adults, whom they perceived as potentially more vulnerable to scams and fraudulent practices. They associated this perceived vulnerability with lower levels of digital literacy and greater difficulty distinguishing legitimate requests from suspicious ones. However, these observations were based primarily on participants’ accounts involving parents, relatives, or acquaintances, as adults aged 65 or older were not represented in the sample. The findings should therefore be interpreted as participants’ perceptions of age-related risk rather than as direct evidence from older platform users.

5.5. Other usage limitations

Financial exclusion associated with reliance on credit cards emerged as an additional concern in the discussions. Participants noted that some platforms require or strongly favor card-based payments, which may create barriers for individuals who do not have a credit card or have limited access to formal financial services.

Participants also raised concerns about the accessibility of SE platforms for people with visual or hearing impairments. The spontaneous emergence of this issue suggests that accessibility was important to some participants, although it does not demonstrate that such barriers are widespread across platforms or user groups. Moreover, people with visual or hearing impairments were not specifically recruited for the study. These observations should therefore be interpreted as participants’ perceptions of potential design limitations rather than as firsthand evidence from affected users.

Finally, the quality of customer support, particularly the reliance on automated chatbots, was described as another source of difficulty. Participants suggested that users who encounter complex problems may also be those who most need clear and responsive assistance. Within this sample, automated support systems were often perceived as frustrating or ineffective when they failed to resolve specific issues or provide access to a human representative. In such situations, inadequate support may intensify uncertainty and mistrust, especially when users are already dealing with payment, security, or access-related concerns.

5.6. Interconnection of dimensions

An important finding of this study is that the five identified dimensions do not appear to operate independently. Instead, the focus group discussions suggest that difficulties in one area may intensify barriers in others. For example, participants indicated that limited digital skills may make it more difficult to assess platform security, recognize potential risks, or manage privacy settings, thereby increasing trust-related concerns.

Participants also perceived that limited connectivity in rural areas could reduce opportunities for frequent platform use and, consequently, for developing digital familiarity through continued practice. However, rural residents were not represented in the sample, so this relationship should be interpreted as an exploratory proposition rather than as direct evidence from rural users. Similarly, participants noted that the absence of a credit card may restrict access to SE services regardless of an individual’s digital skills or willingness to use them.

The discussions further suggest that barriers may become more difficult to overcome when multiple forms of disadvantage occur simultaneously. This may be particularly relevant for groups such as older adults living in rural areas, individuals with low incomes and lower educational attainment, and people with disabilities. However, the study did not directly examine the experiences of all these groups or the combined effects of these disadvantages.

These findings are consistent with the structural model of digital inequality proposed by Eichhorn et al. [10], particularly in suggesting that interactions among different dimensions may contribute to cumulative disadvantages in platform use. A related argument appears in Lutz et al. [9], who indicate that the demographic and psychological factors associated with participation may operate in interconnected ways.

6.1. Conclusions

This study examined how individuals perceive and experience digital inequalities when participating in the sharing economy. Based on data from two focus groups conducted with 14 adults in Madrid in July 2024, the Gioia analysis identified five interrelated aggregate dimensions: usability, technology, digital skills, trust, and other usage limitations. Together, these dimensions suggest that participation depends not only on internet access and digital competence but also on platform design, accessibility, payment options, customer support, and trust in data and transaction security.

The findings suggest that digital exclusion in SE may be cumulative. Individuals may have internet access and some experience with digital technologies yet still face difficulties related to complex interfaces, limited payment options, inadequate customer support, limited digital skills, or concerns about privacy and fraud. The study therefore offers a user-centered framework that links individual digital resources to platform-level conditions and broader socioeconomic constraints.

The Gioia methodology provided the primary interpretive framework for understanding participants’ experiences. NLP was used as a complementary exploratory tool to identify recurring words, platform names, and geographic references in the transcripts. It was not intended to statistically validate the qualitative findings but rather to enhance transparency in the descriptive examination of the corpus.

Because the sample was small, geographically limited to Madrid, and included a relatively high proportion of participants with higher educational attainment or professional experience, the findings should be interpreted as exploratory rather than representative. They may not fully capture the experiences of older adults, rural residents, individuals with lower educational attainment, or those with very limited access to digital technologies. Nevertheless, the findings suggest that more inclusive participation may require attention not only to digital skills and infrastructure but also to simpler interfaces, greater accessibility, flexible payment options, trustworthy data practices, and access to human support.

6.2. Implications

The findings have practical implications for platform designers, policymakers, and educators. Although the identified barriers are interconnected, each stakeholder group can address different aspects of digital inequality in participation in the sharing economy.

For platform designers, the findings highlight the need to simplify interfaces and reduce the number of steps required to complete essential tasks. Navigation, registration, identity verification, booking, payment, and cancellation processes should use clear language, visible instructions, and consistent icons. Platforms should also incorporate accessibility features for users with visual, hearing, motor, or cognitive impairments. Greater flexibility in payment methods could reduce barriers for individuals who do not have access to credit cards or who are reluctant to provide financial information online. In addition, access to human customer support should be maintained alongside automated assistance, particularly when users encounter account, payment, security, or service-related problems.

For policymakers, the findings suggest that inclusion in SE requires more than access to a digital device or an internet connection. Public policies should combine investment in affordable and reliable digital infrastructure with initiatives to strengthen digital skills, particularly among older adults, individuals with lower educational attainment, people living in areas with limited connectivity, and other groups at risk of digital exclusion. Policymakers could also encourage platforms to adopt inclusive design principles, accessibility features, transparent data practices, alternative payment methods, and multiple channels for customer support. Where appropriate, regulatory standards could be introduced to promote compliance with these practices. Such measures may help prevent platform-based services from reinforcing existing digital and socioeconomic inequalities.

For educators, training should focus on the practical skills needed to use SE platforms safely and independently. Programs could include guided exercises on creating accounts, navigating platforms, comparing services, making and canceling bookings, completing digital payments, managing passwords, protecting personal data, preventing fraud, and accessing customer support channels. Training should be tailored to participants’ prior experience and level of confidence rather than assuming a uniform level of digital competence. Libraries, adult education centers, universities, local associations, and community organizations may be particularly well positioned to provide personalized support and practice-based learning in trusted settings.

6.3. Limitations and avenues of further research

This study has several limitations that should be acknowledged. First, its geographic focus on Madrid, Spain, together with the relatively small sample size, limits the generalizability of the findings. Although the qualitative approach provides in-depth insight into participants’ experiences and perceptions, the findings may not capture the cultural, social, and economic variations present in other contexts.

In addition, the sample was limited to adults aged 25–64 and therefore did not include the direct perspectives of older adults. Insights related to this group were based on participants’ accounts involving relatives, which limited the extent to which older adults’ experiences could be represented in the analysis. The sample also included a relatively high proportion of participants who had completed or were pursuing higher education. Consequently, the experiences and challenges of individuals with lower educational attainment were not directly captured.

Future research could broaden the scope of this study by comparing different countries or regions to examine how the identified dimensions vary across cultural, social, and economic contexts. Longitudinal studies could also explore how users’ perceptions and digital skills evolve over time. In addition, quantitative research could complement the qualitative findings and assess the extent to which the observed patterns apply to larger and more diverse populations. Social media sentiment analysis could provide a further exploratory perspective on how different population groups discuss and perceive the use of SE platforms.