1 Background and Context

In this chapter, we present an overview of the Collaboration and Coordination in Networks of Work (2C-NOW) research project. The aim of 2C-NOW is to investigate and theorise the sociotechnical change and transformations to work processes and work practices arising from the appropriation and use of collaboration technologies to support distributed work in organisations. Funded in Phase 1 of the DFG Priority Programme “SPP2267 Digitalisation of Working Worlds” (Henke et al. 2018), 2C-NOW contributes to the overarching programme framework at the micro- and meso-levels of activity; examining transformations to collaborative work processes and practices as they arise for individuals, workgroups and organisations. In terms of the framework’s motion dynamics, the focus is primarily on: permeating, understanding the ways work practices are being shaped, and are shaping the use of collaboration technologies to coordinate everyday work; and making available, investigating the ways collaboration platforms are assembled, are adopted and diffuse across organisations. The study of how people use technologies to collaborate and coordinate work has a long history in the research fields of computer-supported cooperative work (CSCW) and workplace studies (Rogers 1993; Heath et al. 2000; Luff et al. 2000; Schmidt 2011; Schmidt and Bannon 2013). A strength of this research lies in its acknowledgement of the situated nature of human activity. However, such studies tend to ‘privilege particular forms of cooperative work’, restricted to particular contexts and timeframes (Monteiro et al. 2013, pp. 575–576). Whilst a micro-level focus on single-site, small group interactions provides rich insights into the situated, contingent and contextual nature of digital work, it provides limited insights into the wider, meso-level systemic changes that are transforming work practices and the digital workplace. This requires trans-situated research designs that take a broader perspective and accommodate ‘non-local constraints’ and ‘extended temporal scales’ (Monteiro and Rolland 2012; Monteiro et al. 2013). To achieve this, and to address the micro-, meso-, and temporal aspects of transformations to collaborative work, our research programme follows a multi-method, multi-level research design. We combine longitudinal company case studies, surveys and analyses of work activity in operational collaboration systems. Thus, data is gathered from both organisations and their employees and from the systems they are actively using in their work. Organisational data is collected through interviews, surveys and through regular interactive research workshops with the companies participating in the IndustryConnect Initiative, a university-industry research programme involving 41 medium-to-large organisations from the DACH region (Williams and Schubert 2017). These organisations are mostly manufacturing and service organisations from a range of industry sectors (e. g., automotive, chemical, insurance, engineering), who have committed to working with the university research team to share data and experiences. To date, in-depth case studies have been developed with a subset of 13 key IndustryConnect organisations, representing a range of company sizes and industries. Activity data and content data are captured from the databases and file systems of collaboration systems that are in active use in organisations. Such system-level data contains the sequences and outputs of user activities and can be used to identify and analyse emerging work practices.

In this chapter, we focus primarily on the latter, that is, tracing collaborative work through analysis of the activity and content data generated in enterprise collaboration systems. In the following section, we present an overview of the emerging collaboration technology landscape and present the findings of an empirical study to examine the portfolios of collaboration software being implemented in user organisations and the formation of enterprise collaboration platforms (ECP). We describe the generic platform configuration and explain how this platform provides the possibility for organisational workgroups to form digital group workspaces to support different types of collaborative work. Based on these findings, we identify several methodological and analytical challenges relating to the scale and scope, spatiality and temporality of these complex, distributed and heterogeneous technological infrastructures as well as challenges regarding data access. In Sect. 3, we address these challenges and present our work to develop novel methods and tools to i) enable the capture and harmonisation of trace data from these large information infrastructures, and ii) use the trace data to analyse and understand transformations to collaborative work and work practices. The theoretical foundations for these methods are presented along with key illustrative examples of their application to examine transformations to collaborative work processes. The chapter concludes with a summary of progress to date, an outlook on the wider potential of digital trace analytics and our future research imperatives.

2 The Collaboration Technology Landscape

Enterprise collaboration platforms (ECP) are complex, large-scale information infrastructures (de Reuver et al. 2017) that are typically implemented by large organisations to provide the technology infrastructure to support employee collaboration and the coordination of digital work (Leonardi et al. 2013; Williams and Schubert 2018). They are increasingly being integrated with other work systems such as enterprise resource planning (ERP), workflow management and document management systems (Gewehr et al. 2017) to provide a unified space for digital work (Leonardi et al. 2013). Spanning multiple global regions, business divisions and workgroups, ECPs are intricately interwoven sociotechnical ecosystems ‘created and cultivated on top of digital infrastructure’ (Constantinides et al. 2018, p.381) forming complex assemblages of actors, artefacts, practices and processes that interact with each other to create new ways of working and new spaces for work to take place. They are also malleable and open to a process of interpretive flexibility (Doherty et al. 2006), as evidenced during the COVID-19 pandemic, when the users and uses of collaboration technologies increased substantially as organisations adjusted to meet work from home mandates by providing employees with systems and technologies to support remote working (Kamouri and Lister 2020). In the aftermath of the COVID-19 pandemic, many organisations implemented a heterogeneous array of new remote and hybrid working arrangements (Gratton 2021), bringing new levels of complexity to the study of transformations to distributed and collaborative work. However, despite significant and growing interest in the use of collaboration technologies to support and coordinate remote and hybrid work, limited research attention has been directed towards providing in-depth studies and empirical analyses of the mechanisms, strategies and actions that are transforming collaborative work practices in complex, large-scale enterprise collaboration systems. Such empirically based studies are essential if we are to understand and theorise sociotechnical change and the digitalisation of working worlds as they relate to the actual work practices and working lives of individuals and workgroups, the primary focus of 2C-NOW.

2.1 Software Portfolios and Collaboration Platforms

In order to understand the transformations to work processes and work practices arising from the appropriation and use of new technologies to support distributed and collaborative work in organisations, it is first necessary to examine the complex collaboration technology landscapes in use in organisations and identify the collaboration software being implemented and the types of collaborative work they support. Thus, we begin at the organisational level with an overview of the findings of an empirical study to examine the complex collaboration technology landscape in user organisations (Schubert and Williams 2022a). Investigating the infrastructure for all areas of collaborative work is a complex task and requires a classification scheme that is consistent with existing analytical frameworks and is relevant for analysing the functionality bundles of contemporary commercial collaboration software. For this, we purposefully reviewed and combined existing analytical frameworks (Ellis et al. 1991; Bafoutsou and Mentzas 2002; Riemer 2007; Williams and Schubert 2011; Schubert and Glitsch 2016) with an in-depth functional analysis of existing software products. The resulting classification schema, Areas of Collaborative Work (ArCoW), was used to structure an online questionnaire to identify and classify the tools being used in organisations (Schubert and Williams 2022a). The survey (and follow-up workshops) was conducted with 23 user companies (members of the IndustryConnect initiative), who collectively represent a total of almost one million employees. Our goal was to characterise and understand the diversity of types of collaboration software that user companies have implemented in practice (Schubert and Williams 2022b). In the questionnaire the respondents identified all the collaboration software products that are currently implemented in their organisation. The survey findings identify that organisations are far from providing a unified space for digital work; there is currently no single integrated enterprise collaboration system that provides comprehensive functionality and support for all forms of collaborative work activity. Instead, organisations find themselves forced to combine many different software products to meet all their collaborative work requirements, resulting in heterogeneous portfolios of tools (from different vendors) with overlapping (redundant) functionality (Fig. 1).

Fig. 1
figure 1

Range of functionality contained in a typical collaboration software portfolio

Full size image

These portfolios of collaboration software include lightweight tools for specific tasks such as file sharing (e. g. in network directories) or simple message exchange (e. g. in chat tools) to more complex enterprise collaboration systems (ECS) that combine multiple functional components (e. g. workspaces with activity stream, forum and Wiki) in one system.

Taken together, these portfolios of tools form the “enterprise collaboration platform” (ECP), providing registered users of an organisation with a wide range of tools to support their collaborative work (Schubert and Williams 2022b). The analysis enabled us to conceptualise the generic structure of current enterprise collaboration platforms in user organisations. Figure 2 shows the generic form of an ECP in which the applications are grouped by their functionality into basic groupware tools (e-mail, notepad, network directories), content-oriented applications (web content management system (WCMS), Intranet, file share, surveys), near-synchronous support (video conferencing, chat, visual boards) and a core ECS with multiple components for joint work (e. g. workspaces that can contain chat, blog, forum, Wiki, etc.).

Fig. 2
figure 2

Enterprise collaboration platform (ECP)

Full size image

The responses from the surveyed organisations confirm that there is currently no single pre-existing ECP design, and that user organisations are building their ECPs from many different software products either purposefully or in an ad hoc manner. However, the platform provides the necessary infrastructure on which to build digital workspaces, where our attention now turns.

2.2 Digital Workspaces and Work Processes

A large percentage of the digitally-supported collaborative tasks in companies are carried out in digital workspaces, the digital environments where organisational units and project teams work together. Digital workspaces are created by selecting and assembling the required functional components provided by the collaboration platform (Fig. 3).

Fig. 3
figure 3

ECP provides the technology to form workspaces

Full size image

The creation of a new workspace starts when a workgroup is being formed. As discussed above, the functionality offered by current ECPs is often broad and decisions need to be made about the choice of the software components to support the different types of collaborative work. Some organisations recommend the use of certain tools for specific work types. However, the process of forming a digital workspace is frequently the outcome of negotiation among group members. In some cases, for example, where the group includes external members (e. g. customers), the specification of the collaboration software products (and therefore the shaping of the workspace) may be a requirement of a project contract. The design of each new workspace is dependent on the specific context of use, the nature of the workgroup involved, the type of work being undertaken and the affordances of the available technologies (Gerbl and Williams 2023).

Collaborative work processes are sequences of tasks, some of which are synchronous, requiring employees to work together at the same time (e. g. using a video conferencing tool), whereas other tasks are performed asynchronously where work is conducted sequentially, with one employee working independently on a task then handing over the work products to another employee when their part is done. Thus, an individual employee might be a member of multiple workgroups and digital workspaces and move between them over a working day (Fig. 4).

Fig. 4
figure 4

Sequence of work tasks alternating between synchronous and asynchronous activity

Full size image

2.3 Methodological and Analytical Challenges

The complexity and scale of enterprise collaboration platforms and the diversity of types of workgroups and forms of digital workspaces raises a number of challenges for research investigating the transformation of collaborative work processes. In the following, we outline three specific areas of impact that are particularly significant for the analysis of trace data captured from active collaboration systems.

Scale and scope. As discussed above, ECPs are large-scale information infrastructures supporting the work of hundreds (often thousands) of employees who are widely dispersed (often globally) across the different departments and divisions of an organisation. Further, the scope of analysis is broad given the complexity of the technology portfolios in use, the diversity of platform configurations and the variety of workgroups being assembled. The scale and scope of collaboration platforms and group workspaces requires the development of multi-system analytics methods capable of capturing and displaying data from multiple systems and tracing work processes as they transition between different work types and applications.

Spatiality and temporality. The growth (particularly following the COVID-19 pandemic) in distributed and hybrid workgroups brings greater attention to the locations where workgroup members are conducting their work. The transitions between work locations and types of work increase as employees move between, for example, home and office, and between different digital spaces. Our interest is in understanding how work is coordinated across workgroups and how this evolves over time, and this requires methods that enable us to move out of a single site of activity so we can gather data across multiple spatial and temporal frames. The distributed nature of the work under examination calls for research methods that are capable of tracing activity across time and space to make the actions and interactions between people and work artefacts visible and interpretable, revealing the rhythms and flows of everyday work in hybrid and distributed workgroups.

Data access. A related challenge is that of data access. ECP and digital workspaces are formed in closed platforms within an organisation; obtaining data access permissions and handling the legal and organisational requirements for data privacy and data protection can be challenging for conducting research using system data. In 2C-NOW, we collect activity-level data from our own platform UniConnect, a large-scale enterprise collaboration system hosted for academic institutions and their research partners with currently more than 3000 users. The platform is hosted on SPARCI, a DFG-sponsored large research infrastructure (Großgerät) (INST 366/7-1 FUGG). UniConnect contains over 12 years of data and provides a unique testbed for developing and refining trace analytics methods and tools to analyse collaborative processes and work practices. System data collection requires special attention to research ethics and methods for compliant data collection, data storage, pseudonymisation, anonymisation and confidentiality. All IndustryConnect member organisations have signed a memorandum of understanding agreeing to their participation in the long-term research programme and currently, for the analysis of system data, we have access to samples of organisational data provided by two IndustryConnect member organisations.

3 Analysing the Digital Traces of Collaborative Work

To address the challenges outlined above, the first phase of the 2C-NOW project has focused on the development of tools and methods to analyse the digital traces users leave when they interact with computer systems. In the following, we describe the fundamentals of digital traces in enterprise collaboration systems and introduce two approaches for investigating them: i) through analysis of activities of users and ii) through the mediating role of digital documents and artefacts. This work draws on, and contributes to, theoretical and methodological progress in the emerging fields of computational ethnography (Beaulieu 2017; Abramson et al. 2018), trace analytics (Geiger and Ribes 2011) and process mining (van der Aalst 2016).

3.1 Events as Traces of User Activities

The analysis of event logs from ECP poses some significant challenges.

Process. The nature of work that is supported by collaboration software is significantly different from the work carried out in what are referred to as process-aware information systems (such as ERP or CRM systems), which support clearly structured, recurring business processes. The collaborative work is more flexible and less well-structured, in that the tasks may be carried out in unpredictable, changing sequences.

Formats. In addition, the content and event logs of the different software systems store data (events and content) in different, proprietary formats. To date, there are no tools available that can convert the proprietary formats into a uniform format that contains enough information that it could be used for trace analytics. This is the reason why we needed to develop new and specialised methods for pre-processing and analysis. Data preparation requires an intricate process during which data is collected, enriched, flattened, converted, abstracted and stored in digital data spaces (Just, Schubert, et al. 2023). The analysis of these data spaces, again, requires special methods based on a profound knowledge of the functionality available in the software and its ability to support use cases and collaboration scenarios (Schubert 2023).

Trace analytics approach. The activities performed by users in information systems are recorded as events stored in event logs. In enterprise collaboration systems, many of the events relate to the manipulation of content items. According to database theory, there is a limited set of manipulations that a user can perform on content; these are the basic (atomic) actions of create, read, update and delete (CRUD). The CRUD actions together with the content type being worked on (e. g. a Wiki page) provide a good starting point for investigating what users are doing. These sequences of activities, e. g. when a blog post is created by one user and then read and edited by another (blog.post.create, blog.post.read, blog.post.edit) can be used to identify the sequences of actions that occur as users work together on specific artefacts (e. g. a blog post).

Figure 5 and 6 show two (pseudonymised) real-world examples from the log of HCL Connections (CNX), one of the leading integrated enterprise collaboration systems in use by organisations. Fig. 5 is an example of events in a forum and shows a discussion between three people on various topics over a period of two months.

Fig. 5
figure 5

Extract of a discussion between three people over a period of two months (pseudonymised)

Full size image
Fig. 6
figure 6

Activities of five group members working on the same Wiki page over 5 months (pseudonymised)

Full size image

Figure 6 displays the traces of joint work on pages in a Wiki and shows the activities of five group members working on the same Wiki page over a period of 5 months.

The academic field of Process Mining (PM) provides methods and tools for the analysis of event logs (van der Aalst 2016). For PM to work successfully, event logs need to be formatted using a standardised format (e. g. XES) so that they can be processed by PM tools. As discussed above, this is a challenge for the analysis of collaborative work since collaboration processes frequently span multiple software products, which record user activity in their own proprietary log format. The data from different systems must be enriched, transformed and aggregated before it can be used by PM tools. To address this issue, we developed a novel method for the harmonisation and aggregation of log files from collaboration systems (Just, Schubert, et al. 2023). In this method, events are described as ‘user actions on documents’ (Just and Schubert 2023). The concept is formalised in the Collaborative Actions on Documents Ontology (ColActDOnt) (Just and Schubert 2022). ColActDOnt specifies the concepts and properties of collaboration events. A major part of the work in the first phase of 2C-NOW focused on the aggregation and harmonisation of event logs from multiple collaboration systems. As explained previously, the digital collaboration infrastructure in companies is not contained in one integrated system but consists of a portfolio of software products developed by different software vendors (e. g. Microsoft Skype, HCL Connections (CNX), Atlassian Jira). Every system comes with its distinct way of logging events. Since the collaborative work of employees extends across many software products, it is necessary to perform cross-system process mining. For example, two colleagues might initiate a chat in Skype, create a Wiki page containing meeting minutes in CNX, @mention a third colleague to loop him in and then plan the discussed tasks in Atlassian Jira (Fig. 7). The actions of these three users all relate to the same “collaboration process” and therefore must be tracked across these three different systems.

Fig. 7
figure 7

Visualisation of the sequence of events in a simplified cross-system process example

Full size image

A reliable method for trace ethnography hinges on the combination of knowing what a person is really doing (in front of the screen) and the analysis of the event logs created by the system (behind the screen). The ELI (Event Log Interpretation) analysis, which is part of 2C-NOW, explicitly addresses the challenge of combining real-world observation with digital logs. Four researchers independently interpreted and (manually) assigned codes for “work types” to events captured in the log files of three months of their own work activity (Schubert, Williams, et al. 2024). At the end of each cycle of coding, the identified codes were discussed and agreed code definitions were added to a joint code book. The code book was then used to create a software programme that automatically assigns codes to future events. Figure 8 shows an example of the type of analysis that is now possible using the codes listed in the legend on the right side. The bars show a comparison of the frequency of the codes found in two workspaces that are used for the same use case, that is, the coordination of joint work in a research group.

Fig. 8
figure 8

Two workspaces revealing different work styles between two groups with a similar purpose

Full size image

As can be seen, the identified codes differ between the two groups, reflecting the different work styles and work routines in place in the two research groups. Whilst the EIM workspace is focused on the collaborative work on documents (enriching information), the BAS workspace is used mostly for the coordination of tasks (administering tasks).

3.2 Social Documents as Traces of Collaborative Work

In addition to examining the events, actions and collaborative work types, it is also of interest to investigate the digital artefacts that are being worked on. The structure of the content in ECS can be described as social documents (Williams et al. 2020). Social documents are the content (work products) that is created through user activity and enhanced through interactions. They are initiated by the create action of the user who creates the intellectual entity, that is, the starting point (core) of the document. Once created, the intellectual entity can be enriched by further content items by any author that has access to the document. This way, the items of a social document can be read (R) and changed (U) and additional content elements can be created (C) and deleted (D) by multiple authors. Social documents are compound documents and can contain multiple different content types (e. g. a forum contains posts with responses and tags). This characteristic makes the social document an ideal study object for examining the joint interactions of people around specific content.

Using the representation of social documents defined in the Social Document Ontology (SocDOnt) (Williams et al. 2020), concepts from graph theory were used to develop the “Content Dashboard” application, a method and tool for the graphical visualisation of social documents. Using data extracted from an operational enterprise collaboration system (UniConnect), we used the Content Dashboard to identify and display different types of social documents and define their characteristic structure (Mosen et al. 2020). Figure 9 shows the outputs of this work to visualise cross-level aspects of social documents and documentary practices. The left side a) represents the structural view showing the hierarchy of concepts from the platform level down to the single item in a social document. The right side b) shows the sequential view of how users work together around a single document (a forum post) on the UniConnect platform.

Fig. 9
figure 9

a) Hierarchy of concepts, b) traces of collaborative activity around a document (forum post)

Full size image

Social documents consist of digital items that are created as people engage in joint work. There are many different types of social documents in collaboration systems (e. g. files, Wiki pages, forum topics or blog posts). They are created ‘with the express intention of being interactive and collaborative’ (Hausmann and Williams 2016, p. 48) and they evolve over time. For example, when an employee creates a forum post containing ideas for the improvement of a work process, this content is then extended by others who attach comments to the original content, add recommendations and tags and share it with other colleagues. These “attached” elements become important components of the original post and show how discussion and activity evolves around the original topic. By examining social documents as traces of collaborative activity, we were able to gain insights into how employees are collaborating with each other to organise and coordinate work within the enterprise collaboration platform (Williams et al. 2020). These constellations of work around social documents take very different forms according to the type of work involved (Schubert et al. 2020).

Fig. 10
figure 10

Visualisation of content (blog, files, forum and Wiki) in two different workspaces on UniConnect

Full size image

Interpretation of structures. Different types of content allow interpretation of the kind of work that is occurring in workspaces. Figure 10 shows a visualisation of the content in two workspaces on the UniConnect platform. The functionality and selected content types (blog, files, forum and Wiki) are identical in the two workspaces. The graphical analysis, however, shows notable differences in the structure of the content in the two spaces.

Fig. 11
figure 11

Visualisation of content structure

Full size image

The explanation for this lies in the different purposes of the two spaces. The left side of the figure shows the workspace of a project group, which is mostly used for the preparation of workshops and the publication of the workshop minutes and project findings. The image on the right side shows the workspace for an organisational unit, which is mostly used for storing files and coordinating a yearly schedule of events. The social document containing the schedule is seen in the large round element in the lower right, representing the Wiki page that is constantly changed when the schedule needs to be updated. Every edit generates a new version of the page which leads to a graph structure that is constantly growing. Such large social documents are frequently a signal for a coordination mechanism.

Deriving interpretation from structure. The occurrence of certain content types makes it possible to interpret what people are doing in the workspace. Figure 11 shows an example of our research to visualise the constellations of work around different types of social documents and collaboration scenarios. It shows the content types in the workspace of the project group in separate images. As can be seen, the collaborative work patterns that result from people (a) “posting news/informing” is distinctively different from (b) “sharing files”, (c) “discussing topics” or (d) “preparing and documenting workshops”.

These visualisations tracing digital work can be further analysed to reveal typical collaborative work practices and coordination mechanisms and to observe how these are shaped and evolve over time.

The structure of social documents can also be used to analyse how work is transformed over time. Figure 12 shows an observation of activity in a workspace that is created and used for the same purpose every year.

Fig. 12
figure 12

Number of social documents created for Q&A (Forum) and course material (Wiki+Files)

Full size image

The workspace is used to coordinate a university course including communication between participants, provision of learning materials as well as a forum for questions and answers (Q&A). The number of students (150), their backgrounds (enrolled study programmes) and the course content were the same over the 8-year timeframe from 2015 to 2022 (years on the x-axis). The biggest transformation occurred during the two COVID-19 years when the campus was closed and all teaching was offered online. In the first year (2020) the amount of material (# of documents shown on the y-axis) increased because the lectures were now provided in videos instead of on-campus. In addition to videos, the material for each week was summarised in a Wiki page, and a link to it was added to the course schedule. At the same time, the discussions in the forum increased; from 133 documents (questions and responses) in 2019 to 196 posts in 2020 (first online year) and even 228 in 2021 (second online year). In 2022 this number went back to 171 when the need to ask questions in the forum decreased due to the possibility to ask them in class. The number of social documents containing the lecture material, however, only increased for the first online semester and then stayed the same. The reason for this is that the digital teaching material produced during Corona times remains useful and is thus still provided to supplement on-campus teaching.

Events and social documents are both suitable instruments to study the interactions among people in their joint work. Events allow us to follow the sequences of activity of users in the system. Often, users do not work only on one single social document at a time but alternate between different ones. A user might, for example, create a Wiki page and then decide to upload a file before continuing with further edits to the Wiki page. The event log shows the occurrence of actions over time. The growing social document graph, on the other hand, shows the interactions of users around the same content.

4 Summary and Future Research Agenda

In this chapter, we have laid out the foundations for a new form of trace analytics that is capable of tracing work activities and work products in large, distributed information infrastructures such as enterprise collaboration platforms. We outline the conceptual and practical challenges of analysing digital traces of user activities, which include that: i) user activities (and thus digital traces) are spread across different software systems, the supporting software products have proprietary and differing log formats, and there is no common standard for logging collaboration events; ii) joint work processes are (in general) not linear, meaning employees conduct their work across multiple work products and systems in flexible and often spontaneous ways; iii) event logs are noisy; they contain many events that are irrelevant and need to be filtered out and the analysis of work processes requires the synthesis of additional attributes (e. g. document id, content type, title) not available in the native event logs. These must be extracted by querying other database tables; iv) the scale and scope of collaboration platforms and their associated log data is very extensive, often involving thousands of users and millions of events and significant computing power and data storage to enable their analysis; v) longitudinal research designs and data with a spatiotemporal view are required to understand sociotechnical change and to visualise these changes through the tracing of work practices over different locations and timeframes.

There are no standard tools for collaboration analytics, and it was therefore necessary to develop conceptual frameworks for all the involved concepts, beginning with the enterprise collaboration platform, which is assembled from company-specific portfolios of collaboration software tools. These platforms provide the possibility for workgroups to form workspaces, places for joint work where work products (social documents) are created and stored. In addition, there is also no standard framework for the description of digitally-supported work processes that can be used for the analysis. We used inductive (explorative) research methods to develop such a framework. All this work was achieved through empirical studies, surveys, organisational cases and in-depth analysis of existing tool portfolios.

The strength of the developed methods is that they can be applied to examine work in all kinds of platforms. We are currently investigating enterprise platforms, but it would also be possible to examine gig economy platforms, as they have the same components of functional, activity and content elements, which can be analysed with trace analytics.

More work is required in the areas of data pre-processing and the creation of rich harmonised data spaces for collaboration events from different collaboration systems (e. g. HCL Connections, Microsoft 365, Atlassian Confluence/Jira, Hyland Alfresco). However, the foundational concepts, methods and tools now exist and the interesting work of investigating and interpreting transformations to work processes and practices can now begin. Work is already underway to interpret the threading of work between different systems, to examine the emergence of work routines and work handover points and to trace the emergence and shaping of collaborative work processes. Cross-system data stores, containing rich data descriptions that are linked to harmonised event logs, provide a valuable data source for investigations to, for example, investigate social networks (how are people connected?), interaction patterns (who is working with whom?), document types (what work products are being created, by whom?), time (when are people working?) and space (where are people working?).

Overall, the field of trace analytics provides a vast potential for workplace studies and sociological studies of work. The methods developed in 2C-NOW now enable us to combine a micro-level focus on single-site, workgroup interactions to examine the situated, contingent and contextual nature of digital work, as well as to examine the wider, meso-level systemic changes that are transforming work practices across systems and organisations.