Communities of/for Interest: Revisiting the Role of Migrants’ Online Groups

This article offers a critical examination of the role played by migrants’ online communities. With much of scholarly analysis focusing on the new ways in which online groups enable migrants to connect, interact or socialise together in digital space, little attention has been paid to how these groups are actually formed, by whom and with what motivations. Drawing on qualitative interviews with moderators of online groups created by EU migrants living in Wales, UK, our findings reveal the diverse and sometimes ambivalent roles played by these groups, acting not only as networks of support for migrants (‘communities of interest’) but also driven by commercial motives. To capture the impact of this commercialisation and the complexity in the field, we introduce the notion of ‘communities for interest’. The article thus offers new empirical and conceptual contributions that advance our understanding of migrants’ online communities beyond the much-discussed online/offline and virtual/real dichotomies.


Introduction
The rise of social media and the radical transformation of internet use in the last two decades has been described as marking a new shift in the field of migration; 'we are no longer in the age of the uprooted migrant, but have instead entered the age of the "connected migrant" . . . [when] migrant communities are experiencing new forms of connectedness through acting in and occupying digital territories' (Schrooten, 2012(Schrooten, : 1798. Among these 'new forms of connectedness' are migrants' online communities and networks, which have emerged as a major area of study in recent years. Here scholars have drawn attention to the significant role played by these networks in facilitating migrants' communication and connection with their 'home' countries, enabling migrants to form diasporic communities in a transnational space and maintain their cultural and national identities across borders (Ignacio, 2005;Keles, 2016;Mitra, 2005;Oosterbaan, 2010;Swaby, 2013). They have highlighted the novel and more efficient ways in which migrants' identities and belonging are articulated and reproduced in the virtual space. Along with providing an infrastructure for new forms of diaspora, the internet has also been underlined for its empowering effect especially for those migrants who feel isolated and marginalised, for example, by providing 'cybernetic safe spaces' (Mitra, 2005) for users to express their voices and share their concerns openly and freely (Swaby, 2013), or articulating and challenging traditional power relations that prevail in their 'homeland' (Brinkerhoff, 2012). In doing so, the internet 'offers the opportunity of creating a cyber-community where the in-group discourse provides the glue that holds the members together in the virtual space, even though their geographic locations might be distant from each other' (Mitra, 2005: 386).
At the same time, research has also shown that online communities enable migrants to connect and 'glue' together not only in the virtual space but also in places and localities in which they have come to live (Nancheva, 2021;Oosterbaan, 2010;Sabido, 2017). These studies have demonstrated how online communities facilitate migrant settlement and emplacement in the country of destination, thus providing an important avenue of integration and empowerment especially for more marginalised migrant individuals (Kok and Rogers, 2017;Nancheva, 2021). Scholars have also highlighted the importance of social capital generated by migrants' online groups, pointing to different ways in which these groups function as informal support networks for their members (Bucholtz and Sūna, 2019;Chen and Choi, 2011;Oh, 2016;Oosterbaan, 2010;Sabido 2017;Nancheva, 2021). Inter alia, this includes group users sharing information and experiences about their daily lives, receiving practical help and emotional support on a range of issues, meeting new people, socialising among themselves and so on. Researchers have also shown how the circulation of information and knowledge in migration networks can facilitate further mobility and migration (Dekker and Engbersen, 2014;Dekker et al., 2016;Keles, 2016).
A noticeable feature of the migration literature on online communities has been its development as a largely separate field, often in isolation from critical debates within the fields of migration studies (Wimmer and Glick Schiller, 2002) and digital sociology (Evans, 2013). With the focus of scholarly analysis being often on a single ethnic or national group, migrants' online communities are generally noted for their significance in enabling migrants to 'connect', 'group', 'glue' or 'socialise' together in new ways, be it transnationally or locally. Such analysis tends to be underpinned by an 'ethnic lens' (Wimmer, 2007) that perpetuates the idea of migrants as homogenous and distinct ethnic or national groups. As a result, little attention has been paid to how migrants' online groups are actually formed, by whom and with what motivations. As our research reveals, there are a variety of reasons as to why migrants create online networks, which include not only to connect with fellow nationals be it in the 'host' society or their countries of origin but also for economic motives. The commercialisation of online participation, a key topic of interest within the field of digital sociology, has been an area that has particularly seen little engagement within the migration literature. Yet as our study shows, such commercialisation has also penetrated the field of migrants' online communities. To capture and assess this impact, we introduce the notion of 'communities for interest' to highlight new types of migrants' online communities that are created for commercial purposes, a notion that contrasts with Wellman/Castells' concept of 'communities of interest'; that is, the idea of people forming a group around their ethnicity and nationality.
The article is based on interviews with moderators of various Facebook groups and online communities created and used by EU migrants living in Wales. By exploring the different motivations behind the creation of these groups, our analysis sheds new light on the diverse and complex role played by migrants' online communities. The findings offer an empirical and theoretical contribution to the existing literature that extends our understanding of migrants' online communities beyond the much-discussed online/offline and virtual/real dichotomies.

Conceptualising Online Communities
Since the term 'virtual community' was introduced in the 1990s by Rheingold (1993), the past two decades have seen a great deal of scholarly debate revolving around two key conceptual dichotomies: 'virtual/real' and 'online/offline'. Initially, researchers questioned the idea whether virtual communities should be considered 'real' and thus merited to be studied using the same analytical tools and categories as that of 'actual' communities (Wilson and Peterson, 2002). As Buckman noted, scholars spent a great deal of energy 'trying to work out which online communities are really communities' (Buckman cited in Preece and Maloney-Krichmar, 2005: n.p.). This was also reflected in different concepts used in the field such as 'computer-based groups', 'digital networks ', 'brigdespaces' (Adams and Ghose, 2003) or more loose terms such as 'online social relationships' (Fernback, 2007a: 49), categories that foregrounded those aspects and characteristics that set online groups apart from 'real' or 'actual' communities, for example, lacking face-to-face communication, spatial reference or boundaries, commitment, longevity and so on. Others saw virtual communities as a new type of communities; while there was a recognition that these new communities are disembodied, deterritorialised and thus may lack the 'spirit' or 'commitment' of actual communities, they still played a role in creating a 'sense of community'; that is, subjective feelings of empathy, support and belonging among people (Castells, 2003;Wellman, 2001). Within the migration literature, for example, scholars have been using Anderson's notion of 'imagined community' to analyse the features of online groups and their role in generating a collective sense of belonging (Andersson, 2019;Bucholtz and Sūna, 2019). Here empirical research has highlighted the complex ways in which attachment and belonging operate within (migrants') online communities, for example, that members of online groups may share the feeling of belonging to a community even though they have different degrees of attachment to it. Other studies have shown that in some cases a sense of community among migrants' online groups is based less on national identity and more on users' shared interests, political background or regional affiliations (Sabido, 2017). Moreover, researchers have also pointed to the important role that online communities play as support and solidarity networks, showing that, regardless of whether online communities resemble or not 'traditional' communities, their significance in providing practical support for its users cannot be underestimated (Nancheva, 2021).
Just like with the 'virtual' vs. 'real' distinction, the online/offline dichotomy has also shown to be problematic when it comes to virtual communities. Empirical studies carried out in the late 1990s and early 2000s had already demonstrated that in practice online activities and identities corresponded with offline experiences and everyday life (Boase and Wellman, 2006;Hampton and Wellman, 2003;Kavanaugh et al., 2005;Wellman, 1997;Wellman and Gulia, 1999). The blurring of online/offline distinctions has also been documented within migration scholarship, with research showing how migrants' online groups and sites are often connected with offline community organisations and networks (Kok and Rogers, 2017;Van Den Bos and Nell, 2006). The focus has thus turned away from studying computer-mediated social interactions as online-only phenomena towards exploring the way in which online and offline activities intertwine.
A key conceptual development that has become influential (and continues to be so) in the field, including within migration studies outlined above, is the 'network thesis', which advances an understanding of online groups through the lens of social ties and relationships. Among the key proponents of this thesis are network analyst Barry Wellman and sociologist Manuel Castells. In a series of empirical studies in the late 1990s and early 2000s, Wellman showed how, with the rise of internet use among residents in various North American neighbourhoods, networks began to play an increasingly crucial role in the way in which local community interacted and organised itself, rather than the locality where these residents lived and worked. In the context of these changes, he suggested the community be (re)defined as 'networks of interpersonal ties that provide sociability, support, information, a sense of belonging and social identity' (Wellman, 2001: 228). Castells (2003: 126) also shared Wellman's definition and suggested that 'virtual communities' should be considered 'networked communities', arguing that the network has become the 'central form of organizing interaction' in contemporary societies and has displaced previous forms of community formations. For Castells, the result of the 'rise of networked society' is the emergence of new forms of community and sociability such as 'communities of interest' or 'specialised communities' in which people build their networks, organise/interact around a specific theme or interest (e.g. migrants belonging to a particular ethnic group, or patients suffering a particular illness). Like Wellman, Castells also emphasised the importance and novelty of interpersonal ties in the internet age and downplayed the significance of locality, arguing that placed-based sociability was important in past community forms and social interactions but no longer played a significant role.
Castells' rather dramatic diagnosis of contemporary society and his celebratory picture of the internet's civic potential, however, have been widely criticised. First, the idea that place-based community would diminish with the rise of networked society has not necessarily materialised. Empirical studies have demonstrated that the internet has not inevitably led to the detachment of people from physical places but rather enabled them to reconnect with their localities in new ways (Evans, 2013;Postill, 2008). As shown above in the review of current migration literature and as will be demonstrated through our findings, this also applies to the case of migrants, with online participation also facilitating their integration and connection to places in which they have come to live. Second, the civil society potential of online participation for transforming communities and democratic spaces in a new and profoundly positive way has been shown to be greatly undermined by market forces and the 'power of consumer-driven capitalism' (Evans, 2013: 82), which are increasingly shaping the online world. Research has shown that online communities are not only created by individuals for participatory purposesthat is, to organise and interact around a specific theme or interest ('communities of interest') -but are also set up by private companies and organisations to connect with as well as 'exploit' their customer base, through, for example, the gathering and use of personal data (Fernback, 2007b;Hackathorn, 1999). As Fernback (2007b: 316) notes, this has led to 'community' being commodified, 'as a form of wealth that can be cultivated, mined, or sold like other capital commodities'. One implication of such commercialisation has been the emergence of new forms of surveillance and control, a major profitable industry that is another negative consequence of internet use and online participation that has been overlooked by the 'network thesis' (Bauman and Lyon, 2013).
The issue of commercialisation of online communities has led to debates around 'real' vs. 'counterfeit' communities to resurface again in recent years. As Evans (2013: 88) notes, it is difficult to take these commercially driven communities seriously when they 'hardly speak to our essential humanity and connectedness'. We agree that the rise of the internet has brought about new changes that require critical reflection on the idea of 'community' and what it means in the contemporary world. At the same time, we also think that distinctions between 'virtual' and 'real' can be unfruitful when it comes to understanding these changes. As many scholars have noted, such dichotomy is often underpinned by a romantic notion of 'community' as a socially and spatially bounded entity, an idea that itself has long been contested within the social sciences (Castells, 2003;Postill, 2008;Wilson and Petterson, 2002). Rather than trying to engage in definitional issues, we find it more productive to adopt an approach that 'accept[s] community as a concept with fuzzy boundaries that is perhaps more appropriately defined by its membership', an approach that 'may encourage us to concentrate on more substantive issues such as how communities are created, evolve or cease to exist online' (Preece and Maloney-Krichmar, 2005: n.p.). In the following discussion we follow this approach in order to shed light on the evolving nature of migrants' online communities, including the emergence of 'communities for interest' as new kinds of online groups.

Methodology
This article is based on qualitative research conducted with EU migrants living in Wales, which took place between February 2016 and October 2017. In common with other parts of the UK, EU migration to Wales increased considerably following the accession of eight countries from central and eastern Europe in May 2004. In particular, according to estimates from the Annual Population Survey collected by the Office for National Statistics, the estimated number of people born in EU countries living in Wales rose from 40,000 in 2004 to 93,000 in 2017-2018. The highest concentrations of EU migrants have located in the cities of southern Wales, with an estimated 21,000 living in Cardiff and a combined total of 15,000 in Newport and Swansea in 2017-2018. In addition, there are some other parts of Wales with notable concentrations of EU nationals including Wrexham, Flintshire and Carmarthenshire, especially in Llanelli. In terms of nationalities, the most populous group is from Poland; Polish nationals accounted for just under a third of the around 80,000 EU nationals registering for settled status in Wales up to March 2020.
Our study explored various ways in which EU migrants engaged in civil society organisations and groups in different parts of Wales, forming part of a wider research programme on civil society. We adopted a broader conceptualisation of civil society to include a diverse range of organisations and initiatives in the sample; that is, small and large, new and established, online and offline groups. We conducted qualitative interviews with 42 respondents involved in or representing 25 organisations or groups in a range of localities across Wales. Following a mapping exercise of various groups and organisations identified through online searches, interviewees were recruited through the contact details publicised. Some respondents were also identified through snowballing, or from contact made at various community events attended. The vast majority of the participants were migrants, although in the course of our research we also interviewed and had conversations with several non-migrant key individuals who represented migrant organisations or were involved in migration policy making in Wales. Of those 25 organisations, 17 were either online groups or existing organisations that had both online and offline presence; they are the focus of this article.
Although most respondents were migrants from central and eastern Europe, the research sample also included Portuguese nationals, whose inclusion was informed by fieldwork; we found organisations that involved or catered for both central and east European and Portuguese-speaking migrants. Among the respondents representing the 17 organisations associated with online groups, whose experiences are discussed in this article, 11 came from Poland, while others were from Slovakia, Czechia, Hungary, Latvia, Lithuania and Portugal. The resulting research sample thus consisted of a diverse group in terms of ethnicity and nationality as well as regarding age and gender; respondents included 15 women and 11 men aged between 22 and 70. The interviews were conducted in English; they were open ended, allowing participants to talk freely and openly about the organisations/groups/initiatives that they were involved in. Participants were asked questions, inter alia, about the aim and purpose of these groups, their activities and membership, their use of social media, as well as questions around personal motivation and experiences of participating in these groups and organisations. Interviews were recorded with participants' consent and transcribed verbatim; data were entered into NVivo and analysed through coding and repeated reading and re-reading of the data, drawing out relevant themes and issues and their interconnections.

Communities of Interest
Our research found a large presence of Facebook groups administered by EU migrants in Wales, a trend that has been noted with regard to EU migrants living in other parts of the UK (Nancheva, 2021). Typically, these groups operated in migrants' native languages and varied with regard to their size, activity and accessibility. Some were more active than others and membership size ranged from those with a few hundred users to groups that had more than a thousand members. There were those that operated more 'loosely', as 'open' groups (i.e. needing only a click or 'like' for people to join them), while others were 'closed' and prospective members required an 'approval' from the moderator(s) to be part of the group.
Typically, the names of these groups included references to the nationality/ethnicity of migrants as well as the cities in which these migrants lived in Wales, for example, Polacy w Swansea ('Poles in Swansea'), Polacy w Aberystwyth ('Poles in Aberystwyth'), Newport-i/Cardiff-i Magyarok ('Hungarians in Newport/Cardiff'), Slovaci a Cesi v Cardiffe ('Czechs and Slovaks in Cardiff'), Comunidade da Lingua Portuguesa de Wrexham ('Portuguese-Speaking Community in Wrexham') and so on. 1 This nationality/ place combination resonates to some extent with other studies of migrants' online groups, for example, Oosterbaan's (2010) study of Brazilian migrants' online diasporas, which included group names such as 'Brazilians in Barcelona', 'Brazilians in Amsterdam' or groups such as 'Bulgarians in London' and 'Bulgarians in Liverpool' mentioned in Nancheva's (2021) research on Bulgarian migrants in the UK. We found, however, that in our study some of the groups were formed around language, rather than nationality. This includes Comunidade da Lingua Portuguesa de Wrexham, which connected all Portuguese-speaking migrants in Wrexham. The moderator of the group was a Portuguese national from a southern African country and confirmed in the interview that other members were of African descent who came from former Portuguese colonies, as well as migrants coming from Portugal. Similarly, groups such as Slovaci a Cesi v Cardiffe and Cesi a Slovaci v Newportu ('Czechs and Slovaks in Newport') were also formed across categories of ethnicity, nationality and culture, connecting Czech and Slovak speakers including Roma migrants.
The interviews with the moderators of these online groups revealed a range of roles played by these communities. As communities of interest, connecting migrants together was one of the key reasons. The moderator of Slovaci a Cesi v Cardiffe, for example, explained how the group was initially created as a loose network, a social 'get-together' for Czech and Slovak migrants living in and around Cardiff: As the group of Slovaks and Czechs in this area increased there was a sort of a need to create something which will actually be able to organise perhaps um, not necessarily meetings but you know just something like a disco party, something where we can meet and be you know slightly get more get togethers. (Moderator, Slovaci a Cesi v Cardiffe, 1393 members) 2 Among the social get-togethers organised by the group were frequent trips to the Brecon Beacons National Park, a popular, mountainous destination in southern Wales, around 50 km from Cardiff. For the creator of Polacy w Swansea UK, an online group of Polish migrants in Swansea, the aim was not only to connect with the migrants living locally but also to help them, an aim that was driven by the moderator's own experiences: 'I faced the same challenges as Polish people who arrived in Swansea who could not speak English well. My motivation [to create Polacy w Swansea UK] was to help them.' The moderator described the different ways in which this online community helped its users: They ask about the cost of living and where they can buy Polish products like food. They also ask about where they can rent houses and which areas have good schools. They also ask about where they can attend Polish church services. I also share events for Polish people from different websites. They ask about where they can find work. They ask about public transport and where they can attend English lessons. (Moderator, Polacy w Swansea UK, 1504 members) The role of Polacy w Swansea UK as a 'network of support' (Bucholtz and Sūna, 2019;Chen and Choi, 2011;Nancheva, 2021;Oh, 2016Oh, : 2225Oosterbaan, 2010;Sabido, 2017) is evident here; it operated as a self-help community allowing migrant users to post and respond to questions, asking for information, receiving advice for a range of concerns and issues as well as socialising and sharing experiences online. As well as setting up and moderating the groups, moderators also played their own part in facilitating these discussions. For example, when sharing or posting news or links in English they would often provide a brief translation of these posts into migrants' native language, thus making the information more widely available including to those individuals with limited English language skills.
What the above quotation also makes clear is the local orientation of these groups, with migrants asking for services, programmes, activities related to their local area. This includes not only information about migrant-related activities such as Polish church services in the case above but also about job opportunities, school places, health services, English language classes and so on. In this sense, the role played by these online groups was not only about bringing migrants together in the virtual space but also facilitating their connections and incorporation to the localities and places in Wales. This finding chimes with Kok and Rogers' (2017) quantitative study of Somali migrants based in several western countries. Analysing the content of a large number of websites and Facebook networks used by these migrants, their research found that 'the Somali diaspora appear[ed] far more concerned with social integration activities' (Kok and Rogers, 2017: 41) rather than diasporic engagement, and as such their online participation encouraged rather than hindered their social integration in the 'host' countries. The integration aspect is particularly significant here given the lack of relevant information that exist in Wales for central and east European/EU migrants, despite, as noted earlier, their relatively large statistical presence in the country; in 2006 the Welsh Government produced a Welcome to Wales pack for migrant workers, 3 which included general information about the country, but this pack was by now out of date and it did not contain more local and context-specific information that could have shaped decisions relating where specifically to live in order to access schools, cultural services and so on. Such groups allowed updated, tailored and local information to be conveyed.
While some Facebook groups were set up by individuals out of their desire to connect with and help their fellow migrants living locally, among the groups studied, seven were set up by local organisations and often served as an extension to their services. Having a Facebook group/online presence enabled these organisations to connect with their migrant users in a more efficient way and without incurring any costs, as these online groups could be set up for free. As the moderator of the Comunidade da Lingua Portuguesa de Wrexham group noted, it was much easier to inform users about events happening locally through Facebook rather than contacting 'people one by one because you always forget someone'. In a similar way, the moderator of Cesi a Slovaci v Newportu described how the idea of creating a Facebook group was triggered by a local need: In this case, the moderator worked as a support worker for a local charity in Newport offering support and advice to a range of marginalised groups in the area, including migrants coming from Czechia and Slovakia. Following conversations with her migrants' service users, she decided to open the group as a way of helping them with getting information about what was happening locally. We can see clearly here how the opening of Facebook groups for these organisations to connect with her clients obfuscates the online/offline distinctions, as activities 'on the ground' are interlinked with those online (see Van Den Bos and Nell, 2006). At the same time, these online groups served not only as informational boards but also opened up new ways of offering service to their users. For example, the above moderator of Cesi a Slovaci v Newportu often used Messenger, the Facebook app, to deal with her users' needs and solve their problems online, without them having to receive the service in person: Most of them are my clients as well, yes. Another great thing which we get from it is the Messenger because they directly message me with their stuff. So sometimes I don't have to make appointments, for example, 'I'm pregnant. I need to go on maternity leave. I need a letter from my boss. I do not know how to write it. Could you help me?' I have a template. I can send it to her on Messenger. She doesn't have to come to see me. So I deal with the client issues through Messenger as well which is very helpful. (Moderator, Cesi a Slovaci v Newportu) As the quotation indicates, providing such a confidential service via social media was possible because the users were known to the moderator. Interestingly, as noted above the group was initially set up to provide information to migrants about their local area but then later evolved to offer advocacy and advice services. The example clearly shows how formal and informal lines are blurred here, with a (formal) organisation offering extended services through social media. It sheds light on new boundaries that aid our understanding of migrants' online communities beyond the online/offline dichotomies.

Communities for Interest
While the reasons behind the formation of the online groups discussed above varied, what they all have in common is that they were created for altruistic motives. Whether to provide information or services for migrants, or enable them to get together and socialise, or encourage connection to Wales, these groups were set up with the aim of helping others, thus performing a key civil society function. Not all the groups that we studied, however, had such a straightforward aim. Among those were Polacy w Swansea ('Poles in Swansea') and Polacy w Cardiff ('Poles in Cardiff'), which were created for economic reasons. The administrator who ran both groups explained how the main motivation to set up these two online communities in Wales (and many more others elsewhere) was to help his personal business: By creating a large number of online networks of Polish migrants living in the UK and elsewhere in Europe, the moderator was able to generate income in various ways. This included building online traffic and directing a significant number of 'unique users' to his website Trybuna Emigranta -an advertisement portal and community website. He also offered services to migrant businesses by advertising for them in the groups, often by pinning a post advertising their businesses or placing it as a cover photo. Unlike the moderators of the groups introduced in the previous sections, the administrator of Polacy w Swansea and Polacy w Cardiff did not live in Wales and had no connection himself with the localities of Swansea and Cardiff. These two cities, along with many others in Britain and Europe, were identified as major locations with a significant Polish migrant population, and thus ideal for generating an online crowd for his website. The moderator described how he would 'keep the groups local' by adding new members continuously to increase the size of the groups: The way I manage the groups is that I'm trying to keep it local, local people in local groups. If group is called Polacy w Londynie ('Poles in London') I try to keep it local, anything that is connected to London stays, what's not is being deleted. Same apply to Polacy w Swansea [. . .] I had a situation when people were strongly opposing to people who were not living in certain city to be removed but that wasn't a case when I said -what if some people were trying to find out things about certain cities. So there are local people and not local but the main focus is in Swansea and a particular city. (Moderator, Polacy w Swansea) As indicated here, the 'localness' of these groups and therefore who is 'worthy' of membership are clearly contested issues; in this case, people who were not living in a particular area but wishing to have information about the city were included. What the quotation also illustrates is the technical skills and knowledge required to administer and 'manage' more than 100 online migrant communities, for example, in terms of determining the location of potential members, setting up rules for users' postings and participation, or dealing with spam and abusing posts and so on.
At first glance, it would be easy to dismiss the online communities created by this moderator as 'counterfeit communities' (Freie, 1998). Freie (1998: 4) refers to the latter as those communities 'created and marketed in order to maximize profit and advantage for those in positions to do so'. Both Polacy w Swansea and Polacy w Cardiff were 'invented' by a moderator with no local links to the cities in which these groups were attached to and whose motivation for creating these groups was not because of the 'desire for connectedness'. Yet although these groups were created for private gain, this was not their only role; as the moderator noted, the groups were also helpful for migrants living locally, operating as networks of social support: 'As far as I know they [the users] are asking each other about church masses, religious celebrations, they're looking for jobs, flats, doctors you name it -everything . . . they socialise in the group.' In this sense, we argue here that while Polacy w Swansea and Polacy w Cardiff were set up as 'communities for interest' -that is, created and managed for commercial purposes -they also acted as 'communities of interest' (Castells, 2003) offering help and support to their users.
As communities for interest, Polacy w Swansea and Polacy w Cardiff clearly show that commercialisation has also penetrated the online world of migrants' communities. Yet there is a difference between these groups and 'commercialised communities' (cf. Evans, 2013;Fernback, 2007b) created and run by large companies to connect with their customer base and promote consumerism. As noted earlier, for example, Polacy w Swansea and Polacy w Cardiff were created to help with the promotion of Trybuna Emigranta, which is an advertisement portal and a community website set up 'for all Poles looking for work, services, goods and other adverts abroad'. Another such website often shared in Slovaci a Cesi v Cardiffe was niesomdoma.eu ('iamnothome.eu'), which focused on 'the most important events, laws and changes that affect the lives of Slovaks living abroad'. The group Velso Lietuviai ('Lithuanians in Wales') used a similar website in Lithuanian called nemokamai.uk ('forfree.uk'). These migrant-run websites can be seen here as part of the 'migration industry' that often plays an important role regarding migrants' incorporation to the 'host' society (Garapich, 2008). This is further evident in the case of Majka Llanybydder Aberystwyth, a Facebook page created to advertise a local Polish shop as well as offering information and help to migrants living in the local area. The moderator explained the important role of the site as a hyper-local news outlet for migrants: Yeah, I advertise [on the Facebook page]. Pass the messages. If someone is looking for the job or someone, like local people want to rent their house, accommodation I could advertise like that. If I was looking for staff to the shop I'd put it on the Facebook. Everything I want to pass the message to the people I put on the Facebook really, not only my stuff, you know, promotion . . . I don't keep only for me, for the shop purpose. I keep for everything. How I can help. Say if the Police come to me and say, we are looking for the bag, blue bag it was stolen from the shop, I would put that on my Facebook, and also someone stealing also the bag I would put that.
You know, everything that will help the people or get the information from . . . (Moderator, Majka Llanybydder Aberystwyth) It is clear here from the description offered by the moderator that even a small Facebook site like Majka Llanybydder Aberystwyth can play a significant role as an information hub for both migrants and local population. Its importance becomes particularly apparent when considering the offline context in which this was situated; the Polish shop that the site was linked to was located in a rural area in west Wales, an area that has attracted Polish and other (European) migrants owing to employment opportunities offered by factories located in the vicinity (Jones and Lever, 2014). Given the remoteness of the location and in the absence of services offering information and support, the shop had become a key focal point for migrants and local stakeholders, as the quotation shows; indeed, the moderator noted that the shop was often referred to as the 'Polish Centre' by the local population, indicating its significance as both a commercial unit and a social hub.
The dual roles played by these online groups and sites are clearly evident here both in the case of Majka Llanybydder Aberystwyth as well as in groups like Polacy w Swansea and Polacy w Cardiff, which as noted earlier acted simultaneously as communities of and for interest. It is important to note, however, that this duality is not always fixed. The case of Newport-i/Cardiff-i Magyarok ('Hungarians in Newport/Cardiff') is illustrative of how this can change. The online group aimed at Hungarian migrants living in Newport and Cardiff, was initially created as a community for interest, promoting a delivery company: The group was originally created in 2011 . . . At that time we have tried to create a group to get all people together in Newport/Cardiff area. We have had an international delivery service company set up and has been dealing with delivery between UK and Hungary. If anyone wanted to send a parcel between the two countries was able to do at a better price that you would do in the Post office, DPD and other parcel delivery companies. So the group was aimed for advertisement purposes. (Moderator, Newport-i/Cardiff-i Magyarok, 1054 members) Although far from operating on the scale of the groups administered by the moderator of Polacy w Swansea and Polacy w Cardiff, the objectives were similar here too; the moderator saw the potential market of promoting their delivery service, which may have been of interest to group members. Yet a year later after the formation of the group, the international delivery business ceased to exist and Newport-i/Cardiff-i Magyarok was no longer used as an advertisement for this business but as a community of interest: At the end of 2012, our personal circumstances changed and we have stopped this business but still left the group running but changed the aim to have free information flow for people living the area and help each other with any questions, interests.
The case of Newport-i/Cardiff-i Magyarok demonstrates how the creation of migrants' online groups can be underpinned by diverse motivations and that the roles that these groups play can change over time. It provides another example of the unfruitfulness of trying to pigeonhole online communities into neat 'either/or' categories, as 'real' or 'counterfeit'/'commercialised', as in practice groups associated with or motivated by commercial enterprise (communities for interest) can transform into important networks of support (communities of interest).

Conclusion
This article has examined the different and complex roles played by online groups and virtual communities created and used by EU migrants living in Wales. Whether acting as extended services for local organisations or charities, or bringing together migrants living in a particular location, encouraging migrants' attachment to Wales, or promoting migrants' businesses, the article has shed light on the varied and sometimes ambivalent roles played by these groups. First, as the cases of Comunidade da Lingua Portuguesa de Wrexham and Cesi a Slovaci have demonstrated, these online groups connected not only co-ethnics but also migrants across ethnicities, nationalities and cultures. Second, rather than being transnational or diasporic in their orientation, what our findings have shown is the predominantly local focus of these online communities, facilitating the anchoring of migrants to places and localities in which they lived and more generally to the 'host' society (Kok and Rogers, 2017). The findings thus challenge the notion that migrants' online groups are not grounded in localities and exist in an abstract cyberspace.
Third, by looking at these groups through the perspectives of moderators, our analysis has revealed the blurring of various boundaries. We have seen how the offline/online distinction is problematised in many of the groups that have been discussed throughout the article, be it in the case of those that corresponded to organisations, charities or businesses 'on the ground', or those groups whose activities combined both online and offline interactions. The case of Cesi a Slovaci v Newportu offered a clear example of how formal and informal lines are blurred, with a charity pursuing new ways to offer its services online. The ambivalent role played by groups such as Polacy w Swansea and Newport-i/Cardiff-i Magyarok laid bare the blurring between altruistic and commercial motives. We noted the dual nature of these groups and how they were started to promote websites or businesses while also operating as networks of social support for migrants living in and around the corresponding cities. In this context, we argued that these groups can be seen as both 'communities of interest' and 'communities for interest'. While the motivation for creating some of the online groups was commercially driven, we argued that referring to them as 'commercialised communities' or dismissing them as mere 'counterfeit communities' would overlook the important roles these groups played as networks of support as well as how these roles can change over time. The notion of 'communities for interest' is introduced here to describe these new migrants' online groups and capture the complexity in the field. It shows the changing and evolving nature of communities in a digital age, offering a conceptual contribution to the sociology of online communities that enables us to move beyond the debates around 'real' vs. 'commercialised', or 'genuine' vs. 'counterfeit' communities (Evans, 2013;Freie, 1998).
Whatever the motivations behind the creation of these groups, what they all shared was their significance as self-help communities, allowing migrant users to receive practical information and advice from fellow migrants, to identify jobs and opportunities, to locate services and events and so on. These groups were also valuable in terms of generating social capital and emotional support, enabling a network of migrants to socialise and share experiences online. Yet in terms of policy making, migrants' online groups, and more broadly their online participation, continue to be either overlooked or considered unimportant. Whether this neglect is rooted in the negative view of internet use vis-a-vis community engagement that prevails in popular discourse, or in the commonly held perception that migrants' online participation promotes 'ethnic identity and networking' rather than integration, or in the ongoing preoccupation among policy makers with formal forms of participation and civil society, it is a topic that remains to be explored further. While we do not aim to overstate their importance, nor to understate the impact of commercialisation on the formation of communities in digital space (Evans, 2013;Fernback, 2007b), or to say that online groups should be a substitute for offline services and support for migrants, what our article has demonstrated is that the empirical reality of migrants' online participation is far more nuanced and complex than we are often led to believe, and that online communities can also play an important role as support groups and in facilitating migrants' social integration in the 'host' society. While further research is required from the perspective of users to explore the extent and effectiveness of support and advice provided by these groups, especially those impacted by commercialisation, it is hoped that the findings and insights provided by this article will open up the field for such exploration.

Funding
Rhys Dafydd Jones is a social geographer at Aberystwyth University. His research interests are broadly related to migration and diversity in non-metropolitan regions. He has been involved in the ESRC WISERD Civil Society and WISERD Social Stratification and Civic Repair research centres, and the Horizon2020 IMAJINE project, which looked at European migrants and civil society in Wales, boundary mechanisms and civil society in Czechia, Sweden, Ireland and Wales, and social polarisation and civic participation, and migration to and staying in 'less-popular' regions respectively.

Date submitted April 2021
Date accepted April 2022