Thriving in turbulent environments through adaptive forms of organizing

Purpose – Thispaperexploreshowadaptiveorganizations,companiescapableofcontinuouslyadaptingtheir organizational model, dynamically solve the universal problems of organizing. Design/methodology/approach – The authors applied grounded theory to data acquired from six interpretative case studies, collected in two rounds of interviews (15 in total), then completing and validating the study ’ s evidence through triangulation with several secondary data sources. Findings – In adaptive organizations, polyarchies and intrapreneurial employees are essential to shape the division of labour, leading to high levels of autonomy and empowering individuals and teams, while reducing bureaucracyandhierarchy.Intermsoftheintegrationofeffort,digitalsolutionsarepreferredtosocialproofin theprovisionofinformation,whiletheauthorsnotethatincentivesarealwaysgearedtowardsdevelopingstronghigher-orderdynamiccapabilities. Research limitations/implications – This paper has some limitations that could be addressed in future research,includinglongitudinalstudiestoanalysethelinkbetweentheuniversalproblemsoforganizinganda company ’ s dynamic capabilities. Practical implications – Adaptive organizations go beyond tech firms in responding to the universal problems of organizing work by making specific use of digital technologies. Originality/value – Thepaperstudieshowcompaniesshouldorganizethemselves so thattheycontinuously adapt to an ever-changing competitive environment.


Introduction
Today's business landscape is increasingly split between traditional firms, with their hierarchic structure and centralized authority (Lee and Edmondson, 2017), and companies that, by exploiting digital technologies productively and adopting precise organizational models (Lanzolla et al., 2020;Alaimo, 2022), seem to have the persistent capacity to adapt and disrupt their reference market, producing impacts that are disproportionally large compared to their peers (Birkinshaw et al., 2016). Traditional businesses have become aware that they need to organize themselves differently, but are beset with uncertainty and failure when trying to make the necessary changes (Iansiti and Nadella, 2022). Digitally savvy companies seem able to organize themselves in unconventional ways (Reeves et al., 2015) and tend to achieve considerable growth in performance (Belitski et al., 2022). Practitioners (Ismail et al., 2014) call these firms exponential organizations and suggest actionable models for "going exponential" (Palao et al., 2019). Behind all these models is the underlying concept of organizational adaptation (Sarta et al., 2021), which seems kindled by making effective use of digital solutions (Bailey et al., 2022;Singh and Del Giudice, 2019;Ransbotham et al., 2017).
In other words, companies that are apparently capable of continuously adapting their organizational model to the increasingly turbulent environments in which they operate (Burton et al., 2020) see a more striking improvement to their performance than their competitors (Belitski et al., 2022). By adopting a strategic organizational design perspective (Kretschmer and Khashabi, 2020) and, more specifically, the microstructural approach proposed by Puranam (2018), we claim that these improvements are achieved by ensuring ownership of specific dynamic capabilities (Teece, 2007). Companies so placed have the means to answer the universal problems of organizing which every organization must tackle (Puranam et al., 2014), that of dividing labour among the various organizational agents and integrating their effort. Digital technologies play a pivotal role in these settings (Beyes et al., 2022), especially because of the benefits of having a global overview of the task architecture and the enabling of virtual collaboration (Puranam et al., 2014).
Currently, there is still scant knowledge on how adaptive organizations organize themselves dynamically (Sarta et al., 2021). The aim of our paper is to start filling this gap through a multiple case study on six adaptive organizations.
Our findings show that adaptive organizations respond in a peculiar way to the universal problem of organizing. We emphasize the role of polyarchies and intrapreneurial employees in the division of labour, as well as that of digital technologies and an incentive system geared towards seizing, transforming and making decisions about new opportunities sensed on the market, both of which are critical in the integration of effort from all sides. In addition, we discovered that, if it emerges that polyarchies are the underlying construct of adaptive organizing, social proofs are not relevant, differently from what had been argued by Felin and Powell (2016), to the integration of information and ideas. Digital solutions are preferred, especially when considering that they can promote greater transparency. Ultimately, we reinforce the link between organizational design and dynamic capabilities (Teece, 2018), providing practical insights into how organizations should structure themselves so that they are in a better position to develop the higher-order dynamic capabilities needed to continue to adapt their form of organizing.

Theoretical background
Our paper lies at the intersection between two streams of research, the microstructural approach to organizational design (Puranam, 2018) and adaptive organizations (Sarta et al., 2021), which were investigated from the perspective of their dynamic capabilities (Felin and Powell, 2016). In the remainder of this section, we will analyse each theoretical stream and point out some of the theoretical gaps that are still open and which we intend to fill with our research.

A microstructural approach to organizational design
To investigate how adaptive organizations organize themselves dynamically, we first leveraged on organization design theory and, more specifically, on the microstructural approach proposed by Puranam (2018). According to this approach, organizations are complex structures, and it could be difficult to scrutinize how they work simply by applying a unitary entity approach (Kretschmer and Khashabi, 2020;Puranam, 2018). An organization is a multi-agent system that operates within defined boundaries to achieve a certain purpose, and the actions of its constituent agents are expected to contribute to this objective (Puranam et al., 2014). In addition, the microstructural approach narrows the focus from the complexity of organizations to a few universal problems, i.e. the division of labour and the integration of effort, and to a few building blocks, i.e. the microstructures.
According to Puranam et al. (2014), the division of labour is the breakdown of an organization's overall objective into tasks and sub-tasks (task division), which should be allocated to individual agents (task allocation). As shown in Table 1, task division can be achieved through workflow diagrams, business process maps and also through self-selection based on people's individual skills and motivation. Task allocation involves assigning the tasks and sub-tasks identified to an individual or a group of agents. Workflow diagrams or other tools can be used and different mechanisms applied in order to assign clusters of similar tasks to the same agents (specialization), minimize interdependence between individuals, increase the diversity of tasks and assign responsibility for tangible outputs.
The integration of effort requires instead solving both cooperation and coordination problems (Gulati et al., 2005), hence motivating individuals (cooperation) and ensuring that the agents involved have all the information they need (coordination). The latter is also defined as the provision of information, and the former as the provision of rewards. The provision of information simply means that organizations should give their agents the information they require to perform their allotted tasks. Moreover, as tasks are often interconnected and must be coordinated, the agents also need to know what the other agents are doing. Documents, plans, grouping and face-to-face meetings are all traditional forms of information provision. The provision of rewards is connected to the fact that every agent has interests that may or may not correspond to the organization's goals and each organization needs to find solutions to induce people to accomplish their allocated tasks. The traditional form of rewarding is through monetary compensation, but can also target people's intrinsic motivations and aspirations.
According to Puranam et al. (2014), these issues are universal, while the relative solutions may vary. The microstructural approach to organization design places these universal problems on the centre stage of research and proposes focussing on new packages of solutions enabled by digital technology (Puranam, 2018;Singh and Del Giudice, 2019). Last but not least, we consider novelty in the forms of organizing but not necessarily in organizational forms (Hsu and Hannan, 2005;Puranam, 2018). Among the main differences between the two concepts, a point worth mentioning is that, in assessing novelty in organizational forms, the goals are not necessarily constant, while, for forms of organizing, any evaluation would be impossible without keeping the goals constant (Hsu and Hannan, 2005).

Adaptive organizations arising through organizational dynamic capabilities
Organizational adaptation is "intentional decision making undertaken by organizational members, leading to observable actions that aim to reduce the distance between an

Division of labour
Integration of effort Thriving in turbulent environments organization and its economic and institutional environments" (Sarta et al., 2021). More specifically, the definition defines adaptation as intentional, relational, conditioned and convergent. Adaptation is intentional because it is rooted in the organizational members' awareness of their environment, resulting in their choice to react to, anticipate or ignore changes in that environment. It is relational because organizations and environments influence one another. It is conditioned since environmental characteristics also depend on, and evolve with, the actions of other organizations. Lastly, adaptation is convergent, in that organizations seeking to adapt are attempting to move closer to a set of environmental characteristics.
Organizational adaptation is equivocal. On the one hand, the concept is ubiquitous in management research (Aldrich and Ruef, 2006) and acts as the glue binding together the central issues of organizational change, performance and survival (Vivas L opez, 2005;Pina Cunha and Vieira Cunha, 2006;Sarta et al., 2021). On the other hand, it lurks around in various guises (e.g. "fit," "alignment," and "congruence"), studied in multiple theoretical streamse.g. behavioural (Li et al., 2012), resource-based (Walter et al., 2016) and institutional (Meyer et al., 2009)and at different levels of analysis (Gavetti and Levinthal, 2000).
The fact that multiple organizations respond to one another suggests that adaptation is partly dependent on the actions of peer organizations and should not be assumed as independent (Pfeffer and Salancik, 1978). Whether or not adaptation is pursued by an organization is related to its strategic intention and ability to change (Lewin and Volberda, 1999). However, the streams most closely linked to organizational theory (Lawrence and Lorsch, 1967) interpret adaptation differently, often with an emphasis on adaptation as an outcome (Meyer and Rowan, 1977), with a corresponding focus on what organizations can modify internally to achieve these outcomes under particular constraints (Hannah and Freeman, 1977). More specifically, several authors (e.g. Donaldson, 1987) claim that some organizations seem capable of adapting their structures "to gain and regain fit" (Cardinal et al., 2011;E Cunha et al., 2022).
In an increasingly shifting and turbulent competitive landscape (Millar et al., 2018), organizations must be more adaptable than ever (Eggers and Park, 2018). In fact, the functional and operational routines that drive competitive success in stable conditions -"baseline" capabilities such as supply chain management and access to distribution channels become rapidly obsolescent (Teece, 2007). In these settings, it is essential for organizations to develop dynamic capabilities (Teece, 2007(Teece, , 2018Mohaghegh and Gr€ oßler, 2022), and through them continuously work on their baseline capabilitiesanticipating, shaping and adapting to novel business opportunities (Teece, 2007(Teece, , 2018. Many scholars (Burns and Stalker, 1961;McNamara et al., 2003;Felin and Powell, 2016) have pointed out that the old tools of organizational designhierarchy, chains of command, functional areas, formal reporting and long-term planningare by and large no longer effective to cope with volatile, uncertain, complex and ambiguous business settings. New forms of organizing, defined by strong decentralization, agile processes and individual empowerment, are required and seem typical of adaptive organizations (Sarta et al., 2021;Chen et al., 2020;Majchrzak et al., 2018). Felin and Powell (2016) have called for organizations to be re-designed, to make them more adaptable and capable of developing and nurturing dynamic capabilities. According to the authors, if an organization fails in its internal structural differentiation, it will fall into the traps of insularity or folly, finding it impossible to respond appropriately to rapidly changing environments. At the same time, if an organization differentiates but does not integrate with the outer environment, it will be vulnerable to organizational chaos. An organization that finds the balance between the two mechanisms is likely to create and develop strong higherorder dynamic capabilities, quickly adapting to the context and likely growing more than its competitors (Pezeshkan et al., 2016). Felin and Powell (2016) proposed two levers for an adaptive organizational model, polyarchy and social proof. Polyarchy is "an organizational practice, which consists in having autonomous individuals, who are empowered to make significant choices about the nature and scope of their own work" (Sah and Stiglitz, 1986). This concept is closely linked to the mechanisms of differentiation (Knudsen and Levinthal, 2007;Csaszar and Eggers, 2013) and also connects to Puranam's first problem of organizing.
Exploring this topic means taking both an individual and a company perspective. The first viewpoint refers to an individual's and/or subunit's empowerment, giving them full autonomy of judgement, decision making and execution. In return, as they are the people closest to the market and, therefore, with the best information, experience and expertise within the company, they are those asked to suss out and grasp potential new opportunities (Stinchcombe, 1990). The second viewpoint refers to the company's organizational culture. Implementing this employee-empowered organization requires a culture that encourages independent, creative and non-standardized thinking (Evans and Salaiz, 2019).
There is also another, meso-level, perspective (Salvato and Vassolo, 2017), which brings out the second concept of social proof, defined as "any mechanism of social influence that tends to produce coordinated behaviour among individuals" (Cialdini, 1984). Felin and Powell (2016) suggest that this system can handle the problems/deficiencies brought about by polyarchy, such as cost inefficiencies and duplication of effort (Knudsen and Levinthal, 2007) and coordination with colleagues and accountability to the company as a whole (Felin and Powell, 2016). These problems are connected to what Puranam describes as "integration of effort".
To conclude, adaptive organizations are those that intentionally decide to adopt dynamic organizational forms which enable them to adapt continuously to an everchanging competitive environment. In other words, while their actions may inevitably also depend on external factors, these are companies whose internal decisions and structures are always geared towards adaptation and over time will regain their fit with the environment.

Research methods
In order to achieve the purposes of this study, we used a qualitative research methodology based on a multiple case study (Yin, 2014), particularly suited to analysing a complex phenomenon through a holistic approach (Flynn et al., 1990), to "emphasize the rich and real context in which the phenomena occur [. . .] and provide a bridge between rich qualitative evidence and traditional deductive research" (Eisenhardt and Graebner, 2007).
Our primary aim was to understand how adaptive organizations respond dynamically to the universal problems in organizing (Puranam et al., 2014).
The six case studies proposed were used to examine how organizations tackle these problems, with an emphasis on the resulting development of idiosyncratic higher-order dynamic capabilities and the crucial role of digital technologies. To achieve data triangulation (Stake, 2013;Yin, 2017), we also made extensive use of secondary sources, such as online documentation and archival documents.
Of course, an exploratory case study does not lead to statistically generalizable results (Yin, 2003), but we believe that our empirical evidence can inform future theoretical and empirical studies on the creation and/or transformation of organizational models towards more adaptive configurations where digital technologies are extensively exploited.

Case selection
The six cases were selected according to theoretical and purposeful sampling principles, in order to guide our research and address the research problem effectively (Eisenhardt, 1989;Eisenhardt and Graebner, 2007). Moreover, by being very careful in how we selected our cases, we could be relatively sure of avoiding organizations that could bring nothing to our research, and we were able to define the boundaries of validity for the generated theory accurately. Additionally, it is easier to select cases more likely to provide the needed evidence if this sampling method is used (Eisenhardt, 1989).
Several reasons led us to concentrate on our six companies. First, they are all recognized examples of organizations that leverage on digital technologies not simply to conduct their business, but also to organize themselves. Second, the chosen companies all have unconventional forms of organizing and a powerful corporate purpose such that they can inspire their employees, customers and community to pursue a given transformation or goal. These companies are commonly known in the practitioner world as being "people-centric", "customer-centred" and "community-oriented". Third, we selected the cases for reasons of data access, particularly the secondary sources. The literature on adaptive forms of organizing is particularly convoluted, and a growing number of communities of practitioners, consultants and innovators are offering relevant content that could be exploited beneficially. In line with the principle of engaged scholarship (Van de Ven, 2007), we chose the six cases on the basis of the authors' contacts and/or identified secondary sources, to help us prepare more thoroughly for the interviews, dig deeper into their organizational characteristics and achieve a more effective data triangulation.
We designed our case selection process by looking for the same set of circumstances in companies from different industries, sizes, countries and ages, to add transversality to the study and avoid the risk of coming up with context specific findings. Moreover, we covered the spectrum from extreme cases of organizational autonomy (out of all cases, C and G are the most famous) to cases where organizational innovation is more modest. This research design allowed us to respect the maximum variation sampling principle, whereby extreme cases are included to obtain variations in the dimensions of interest (Patton, 2005;Agostini et al., 2015).
Case A is the first and leading Italian crowd testing platform to test apps, websites, chatbots and any other digital product through crowd power. This company places its community of professional testers at its centre, helping its customer companies to take smarter decisions. It has grown significantly over the years and doubled its revenue in 2021.
Case B is universally recognized as one of the world's most innovative companies, listed on Fortune 500 and with an average growth in revenue for the past three years of 21%. To date, its revenue is about $260 million and its market value is about $1.5 billion.
Case C is a world leader in home appliances and consumer electronics products and services. The company began its transformation in 1984 and owes its growth to a management model that has led the company to organize itself into more than 4,000 teams to serve idiosyncratic and fragmented local markets.
Case D originated from a strategic partnership between a bank (19% of equity) and an IT incumbent (81% of equity), to encourage digital transformation and innovation in banking. The company brings together employees from both companies with the purpose of innovating. This is an unconventional solution, compared, for example, to creating an "Innovation" business unit in a bank or employing a consultancy firm to manage its information systems. In 2020, the company was fully incorporated by the IT incumbent for $2.1 million.
Case E is seen as one of the UK's fastest growing tech companies and most innovative organizations. This company has implemented a unique organizational structure and culture that promotes autonomous decision-making, self-sustainability, curiosity and exploration. Moreover, it uses a product it also sells to automate and optimize the process of work allocation dynamically. In 2021, the company was acquired by a global marketing, communication and advertising agency listed in the New York Stock Exchange.
Case F is famous for having adopted a holacracy and self-organizing model. The company has no organizational structure and responsibilities are managed through a series of nested "circles". Employees can move from one circle to another, choosing what challenge they wish to tackle. The main company circle, the only circle not nested within another, has 18 sub-circles, while each circle contains an average number of 1.8 sub-circles. The company was acquired by a tech giant in 2009 for $1.2 billion and still retains its idiosyncratic organizational configuration and culture. Table 2 details the general characteristics of the selected cases.

Data collection
The data collection process ran for about two years (2020-2022) and involved multiple sources (see Table 3). Two main rounds of semi-structured interviews were held between May and September 2020 and between August and October 2022. The multiple sources of evidence allowed us to limit potential biases and gather stronger insights (Eisenhardt, 1989;Eisenhardt and Graebner, 2007), as well as exploit the synergistic effect of data triangulation (Eisenhardt, 1989;Stake, 2013;Yin, 2017

Thriving in turbulent environments
All the interviews were based on a common protocol, but many initial questions were later changed, or new ones added, to extract the most from each interview. The questions were continuously reviewed and tweaked, on the basis of company material read in advance when available, and new concepts discovered through earlier interviews. This cyclical process allowed us to move forward progressively following the informants' lead, learning from past interviews and adjusting the research protocol. Our starting point was to explore their form of organizing (decentralized, unconventional and flexible). From the interviews, it emerged that there was the need for, and a sort of thinking conductive to, being dynamic, adaptive and responding quickly to changein other words, what we refer to as dynamic capabilities. We, therefore, added questions along these lines to increase the "'representativeness and consistency" of our evidence (Corbin and Strauss, 1990). First, we began the interviews by asking our informants to give a brief description of their roles and the company's organizational design, structure, set of values and culture. We then moved on to open questions, letting the interviewees "speak first"thus unearthing more authentic and reliable information (Flick, 2009). We only asked detailed questions after the interviewee had broached the subject independently. We started by asking general questions about research problems and variables we had come across during the literature review, but we tried to place the interviewees' perspective over the theoretical perspective as far as possible, to generate new findings and apply a non-biased viewpoint.
Each interview lasted at least an hour, was conducted using online tools (Microsoft Teams, Google Meet, Zoom) and was recorded and transcribed verbatim. Two authors independently cross-checked the findings and shared their initial ideas (Bourgeois and Eisenhardt, 1988). The third author reviewed the observations critically. This approach meant that we could maintain a high-level perspective (Gioia et al., 2013).
Potential information bias was managed by ensuring anonymity (Eisenhardt, 1989) and including informants with different perspectives and roles (Ozcan and Eisenhardt, 2009). The interviews were lastly completed with secondary data for three main reasons: (1) we could prepare for the interviews more thoroughly; (2) we could extrapolate additional data for the empirical analysis; (3) we could triangulate this additional data with data from the interviews to ensure the validity of the research. We drew on various reports and renowned academic and practitioner communities to find more information and gain different perspectives on the six cases. For instance, we garnered information from the wide-reaching interview on Platform Design Toolkit [1] for Case C, the "Exponential View" podcast [2] for Case E and articles on "Corporate Rebels" community [3] for Case F. Table 3 lists the number of interviews, the roles of the key informants and the secondary sources we consulted to further our understanding of the cases.
As we wanted to interview people with visibility on the topics under analysis, who are usually in Human Resources (HR) and business operations, we started out by asking: "Who deals with these matters strategically and operationally in your company?". This question often produced suggestions about people in HR, or more appropriately, the People and Culture area, because of the strategic role they play in these organizations. This area/role sits at the intersection of vision, strategy, culture and operations, and people potentially also have experience in "organizational design" and in adopting "new ways of working". As HR people work with both managers and employees on a daily basis, they are particularly useful when trying to find the organizational and team structure most likely to bring about high company performance and employee engagement. The only "profile exception" was in Case A, where there is currently no HR manager as the Chief Executive Officer (CEO) and the Chief Technology Officer (CTO) oversee personnel matters. Without a strategic HR function (except for talent acquisition) in this IT company, it is the CTO who takes the final decision on talents. The CTO is also the "Chief Happiness Officer", and, alongside the CEO, looks after the well-being and engagement of employees and the community.

Data analysis
To analyse the rich body of data collected, we adopted a grounded theory approach (Glaser and Strauss, 1967;Strauss and Corbin, 1998). According to the tips set out for multiple case study theory building, within-and cross-case analyses were performed (Eisenhardt, 1989;Eisenhardt and Graebner, 2007). Each author first analysed the primary data individually and the data was triangulated with secondary sources (Jick, 1979). Then, adopting an inductive approach (Saldaña, 2013) and the method proposed by Gioia et al. (2013), we coded the various interviews. The systematic approach suggested by Gioia et al. (2013) allowed us to review the transcripts line by line, and develop new concepts while keeping "qualitative rigour" in conducting and presenting our findings.
In the first order analysis, the most promising concepts were extracted and the interviewees' words were left unchanged, to keep a faithful record of the original terms. At the beginning, we seemed to have mustered a large and rather unwieldy stash of quotes, but it is important not to lose anything at this stage, and be sure that all the promising themes are considered (Gioia, 2004). After that, a second order analysis (cross-case) was carried out by replicating the logic across the cases and grouping sentences with similar meanings. Each group was labelled, and the original terms retained were possible. At this point, we started behaving as knowledgeable agents, making connections between the interviewees' level and the theoretical level, and understanding the theoretical implications of the topics that were emerging from the interviews (Gioia et al., 2013). As suggested by Gioia et al. (2013), whenever the researchers did not all completely agree, we revised the analysis until we reached consensus. We then defined the aggregate dimensions to bring the number of categories down to a more manageable number. During this process, particular attention was paid both to the concepts strictly related to the previously read literature and to the new and unexpected topics that cropped up during the interviews.
Once the cross-case analysis was running, we moved between the case data, the emerging concepts and dimensions and the academic literature, in order to refine the emerging construct definitions, abstraction levels, construct measures and theoretical relationships (Gilbert, 2005). For clarity, the process of data analysis and the evolution of conceptual categories (Suddaby, 2006) are presented in Figure 1.

Findings
Our research has revealed that there is a particular way of implementing adaptive forms of organizing. These findings are set out in the upcoming four sections, where we run through the ways adaptive organizations address the universal problems of organizing identified by Puranam et al. (2014), i.e. task division, task allocation, provision of information and provision of rewards.

Task division through polyarchy
From the interviews, it emerged that the underlying reason for creating adaptive forms of organizing comes down to the desire to adapt quickly to the market, thereby individual employees and teams are given the responsibility of identifying new tasks that need to be accomplished and also the authority to make these tasks part of the organizational structure (Martela, 2019).
All six cases have elements that can be traced back to polyarchy, as suggested by Felin and Powell (2016). Our results, however, go deeper into how polyarchy is implemented. All the organizations we analysed are based on four aspects: (1) giving people independence; (2) giving basic organizational units full autonomy; (3) setting up temporary ad hoc teams; and (4) giving people time slots to spend on other projects.
The first aspect was particularly on show in Case E and Case F, both of whom created an organizational system built around roles and circles. Case E is organized around independent projects where people have high autonomy to decide who will be working at any time, while employees in Case F are completely free to create and engage in new roles in the organization: People in the team can pick and choose the work they want to do. We don't ask people to take on a specific role, everyone can move around and develop their career. (People & Culture Operations Manager, Case E) Anyone can propose roles or circles. This is special, as it usually comes from the top in a traditional organization. It isn't just about stopping people getting stuck in one job, it's also a way for people to do different things [. . .] We wanted to become a self-organized company, one where people can respond to hunches or opportunities by self-organizing around the work. (Insights Team, Case F) The second aspect was employed wholeheartedly in Case C, which achieved interesting results by giving full autonomy to its basic organizational units, "micro-enterprises", which are seen as small entrepreneurial companies within the organization. The underlying idea is that decisions must be taken by the people in the micro-enterprise ("MEntrepreneurs"), who are closest to the market and thus most abreast of their own needs and those of their customers, as well as of the opportunities that could be leveraged. In the words of the HR Director for Europe and Russia at the company: MEntrepreneurs are people expected to take their own decisions. They should be able to know who they need, recruit the right people, settle them in, and give them the right pay so that they give customers what they want. It's like having your own company. What we want is that only people close to the market can and should make these decisions. (Case C) One interesting aspect that emerged in the second round of interviews is that Case F is edging towards Case C's situation. Case F's Insights Team in fact noted: Holacracy gave us a lot of words that were helpful when we were explaining what it means to be selforganized and self-managed, and the concepts of purpose and accountability [. . .]. However, when we built our constructs of market-based dynamics, the triangle of accountability was the really important step. Our first guideline helps teams work as micro-enterprises, and it connects to our second guideline, so each team has its own financial statements, can be a customer/supplier/investor of other teams within our company, and so on. (Case F) The third aspect was used above all in Case A, where instead of organization-wide polyarchy, they use something similar on an ad hoc basis. When there is the need, or someone senses a potential business opportunity up for grabs, they set up a team and define exactly what its scope is and what it has to do. The CEO and CCO noted that: Interesting solutions often emerge and are put into practice. We set up experimental teams as and when, they are based on a concrete concept raised by someone who is part of the team. They are optional. (Case A) Connected to this aspect, the corporate community plays a primary role: We have a layer of community experts who help us with management operations and collaborate with us on a daily basis. We give them training and monitor them through digital systems and they help us stay up to date with what is happening on the market. (Case A) Polyarchy works in a similar way in Case B, which primarily "embodied" the fourth aspect, that of allowing employees to spend time on innovative projects outside their usual tasks. Employees have a certain latitude to decide what projects they want to work on in their allotted time slots, agreeing them with their managers. Polyarchy comes into play in these instances, allowing employees to tackle the business ideas they find most promising. In the words of the Market HR Cluster Lead for Greece, Ireland, Italy, Portugal and Spain: In general, we allow our employees to use some of their time on developing projects that are not necessarily related to their core job but could equally be useful. These projects are decided jointly with the employee's manager, as is the time they can spend on it. (Case B) This time spent on innovative projects is treated as real working time with a real role in the company, and people can apply and be assessed for the position. The Digital Sovereignty Leader gave us a better understanding of how it works: There is an internal file of all open jobs and positions worldwide, including those in innovative projects (where people can spend 5, 10, 15 or 20% of their time). People can propose projects, and this often leads to them being put into contact and working with various international communities. Proactivity is encouraged, and the process is codified, structured and monitored via analytics. (Case B) Comprehensively, these results suggest that: P1. Task division in adaptive organizations is mainly achieved through polyarchy, giving decision-making powers to highly autonomous agents who tend to organize themselves dynamically into cross-functional and/or temporary teams which define the task to be executed independently.

Task allocation through intrapreneurial agents
Our results confirmed that an adaptable organization requires a decentralization of decisionmaking powers (Zhao et al., 2022) and a reduction in bureaucracy and hierarchy (Lee and Edmondson, 2017) in exchange for increasingly accountable employees whose job is to sense, seize and transform market opportunities, within a process of continuous adaptation, innovation and growth for their company (Felin and Powell, 2016;Graetz and Smith, 2008).Our interviews revealed that intrapreneurial individuals are crucial to an adaptive organization, as is a recruitment process that brings people's relative skills and mindset to the fore. Both Case A and Case D, for instance, ask their employees to be proactive and propose new ideas, and these are then evaluated by their managers in view of being implemented:

Thriving in turbulent environments
We continually encourage people at all levels to bring us new ideas. I don't expect strategy to come from below, but ideas do. (CEO, Case A) Each of us can be innovative and must be innovative [. . .] Each of us is the bearer of innovation so there is no scope, there is no space, there are no allocated hours. It is always a good time for innovation. (CHRO, Case D) To find people who fit this "highly proactive" and dynamic way of working, all six companies search for people who are confident to test new ideas and go outside their comfort zone. These traits count more than technical skills alone, which by themselves are not enough to ensure the company's value creation in the long term: When hiring, we give more weight to someone's attitudinal, entrepreneurial, human and nontechnical part. (CEO, Case A) The ideal candidate must have qualities of responsibility and leadership. (CHRO, Case D) Case E, differently from Cases A and D, has adopted a simple process to get the best out of intrapreneurial individuals and rapidly transform the opportunities seized in the market. As Regarding the recruitment process, it is key to find people who can easily get used to this unconventional and highly empowered and empowering way of working. The head of HR at Case F pointed out that: People have to be in line with our corporate culture, they have to be proactive if they are to change and question the current status. (Case F) The People and Culture Operations Manager instead noted a risk that should be taken into consideration during the process: This system is probably a bit challenging for someone who is shy, with less experience and so less confidence to go out and explore and find stuff. (Case E) The concept of an autonomous, proactive and intrapreneurial employee also has a leading place in Case F. The Insight Team stated that: Every employee is a kind of "sensor" within the organization. Everyone who has a hunch or wants to propose something new can do so. We then run a proposal past a set of questions and, if it safe and the answer is yes, we have to see if it works for the organization. (Case F) To be adventurous, creative, open-minded and risk-taking is what Case C also asks of its employees. This is spelled out in their form of organizing, which is grounded in microenterprises, the visible form of the employees' aspirations. The MEntrepreneurs decide themselves which opportunities to capture and which objectives to pursue. In the words of the HR Director for Europe and Russia: Our employees are energetic entrepreneurs. I recruit people who can tell me what to do and can take the right decisions. [. . .] Our employees have the right entrepreneurial mindset. (Case C) Case C also indicates that forms of organizing to develop dynamic capabilities should be geared towards reducing bureaucracy and hierarchy. This company is seen as a collective of numerous independent micro-enterprises that collaborate as if in a network, with little hierarchy and bureaucracy, despite its vast size. The HR Director for Europe and Russia at Case C noted that: We have an organizational model that mimics the architecture of the internet: "small pieces, loosely joined". The web is incredibly diverse and yet still coherent. While it has spawned countless innovations, it's held together by common technical standards that makes cyberspace navigable and allows sites to swap resources like data. (Case C) Moreover, the director also gave us her perspective on bureaucracy: Why is there so much bureaucracy in companies? Because of power! Because people want to control other people. People are much safer in an environment full of rules where everything is clear. So, for me, it's really a question of power games and fear of change. (Case C) On this point, Case A also put forward an interesting perspective, in that it manages its community leaders as a real asset in its business: We have a functional structure, but decisions are not made in a hierarchical way. We set up a crossfunctional team of three or four people to solve a problem or seize an opportunity. These teams work in an agile way, they organize themselves and there is no team leader. Innovation often comes from these teams [. . .]. The manager acts as a servant leader. (CTO, Case A) There is no hierarchy, it's all based on experience. We are very one-to-one at the level of interaction, and it's all managed through various collaborative tools. There is verticality in the sense of experience, that is, whoever is doing a job has to know how to do it and can be consulted directly, without people having to go through all the different managers, and this avoids bottlenecks. (CCO, Case A)

Respondents from Cases E and F expressed similar views:
Being somewhere where you have lots of autonomy is quite refreshing and rather enjoyable, without the shackles of hierarchy, which are kind of political. We don't have traditional job titles in the organization, it cuts out the politics. [. . .] Being adaptive is key in AI, and here also. It's how to make the organization as adaptable as possible. (People & Culture Operations Manager, Case E) We didn't want to be defined by a title but by the work we do. (Insights Team, Case F) Overall, the interviews showed that by implementing polyarchy as a way of working, it is possible to reduce bureaucracy and hierarchy in organizations, seen as the main culprits of a "political" climate, disengagement of intrapreneurial individuals and lack of the ability to respond in a fast, agile and continuous way to the market. Employees can more or less decide which roles to take on (Martela, 2019) and which teams to interact with to exploit market opportunities. These results suggest that: P2. Task allocation in adaptive organizations is based on there being intrapreneurial, proactive agents and an organizational design that minimizes bureaucracy and hierarchy. These aspects allow individuals and teams to be autonomous in deciding or proposing sub-tasks. Felin and Powell (2016) suggest that businesses such as adaptive organizations that are constructed largely on polyarchy should use social proof (i.e. build consensus for an idea) to integrate individual effort and, thus, handle their coordination, alignment and duplication issues. Puranam et al. (2014) cover these aspects in the provision of information problem, which means that the company must ensure that each organizational agent has the information needed to perform his or her tasks. Our interviews with people in adaptive organizations, especially the most advanced, revealed that digital technologies play a key role in this matter. The Insights Team in Case F, for instance, stated that:

Provision of information through digital technologies
In our organization, we are not building consensus, because all too often, when people come up with an idea, they're so busy trying to find room for everyone's questions and feelings that the original idea gets lost and the person doesn't get what he or she needs from that proposal. So, the proposer can go it alone to decide whether to change the idea or not. (Case F) While the People and Culture Operations Manager at Case E told us: We have very strict processes and roles, but they are organized so as to allow "swarm behaviour". No consensus is needed, but we just make sure to select the best idea from the best people in that context. How? We have a digital, Darwin repository to match ideas with teams, which are set up dynamically around the ideas. (Case E) In order to put "the right person in the right place", managers at Case E explained that: We are building a software platform to analyse our in-house skills more intuitively, see what we need and incorporate it with network analysis. In other words, who is working with whom and how people network inside the organization, since these aspects influence people's skills and their direction. [. . .] We're building a technological infrastructure to monitor which projects people are working on and how they are working, extracting info from all that. (Case E) Similarly, Case F has a digital and centralized platform that enables the transmission of information in real-time and avoids the duplication of ideas within corporate settings. The Insights Team argued that: Everything is transparent using a digital tool [. . .] If someone wants to propose something, one thing we always ask ourselves is whether it's something that's already been done by someone else in the organization. People are expected to check this out thoroughly and, if it's not been done before, they can take their proposal to the next step and organize a team around it. (Case F) Similarly, Case C wanted to minimize the risk of duplicating initiatives and ideas from different micro-enterprises. Here they applied market proof and not social proof, using digital technology systems to inform the other business units in the organization about the various initiatives: In our organization, we publish an online newsletter every fortnight, reporting on the performance of our micro-enterprises and initiatives. Every successful micro-entrepreneur is expected to give a rundown of what he or she has done. (HR Director for Europe and Russia, Case C) Our leaders, recognizing they may not be well placed to judge the merits of a new idea, often require a start-up team to obtain outside funding from one of the company's venture capital partners before agreeing to contribute internal resources. (HR Director of Europe and Russia, Case C) Social proof is not applied in Case D either, and instead they built a digital platform, similar to Case E (which avoids duplication of projects, rather than ideas). The platform shows how many hours each employee spends on a given project. Employees are also evaluated and then remunerated on the basis of their ideas and the innovative solutions they come up with. The reason why they decided not to use social proof is because, under that arrangement, the ideas belong to the crowd, not to the single employee and individual assessments would not be feasible. As noted by the CHRO in Case D: There is no social mechanism to filter ideas because innovativeness is an element that we evaluate within performance. We ask for each of us to be proactive in this. Ideas are always welcome and always stimulated, so we expect each of us to come up with an innovative idea. (CHRO, Case D) Moreover: We use a tool that indicates how many hours each employee spends on a given project, because, among other things, they can also work on different projects. The objective is to get the most qualified people for the ongoing projects. (Case D) There is also a second benefit of adopting digital technology instead of social proof to solve the problem of providing information effectively throughout an adaptive organization, as highlighted by the Market HR Cluster Lead for Greece, Ireland, Italy, Portugal and Spain in Case B: Full transparency is a strong part of our culture. We use real-time dashboards and a calendar of meetings to make sure people have access to the information they need. The main difference is that most companies are afraid of sharing confidential informationwe take the risk and embrace the big impact this has on people. (Market HR Cluster Lead for Greece, Ireland, Italy, Portugal and Spain, Case B) Moreover: We have lots of internal platforms where you can find information of all kinds, post information automatically on LinkedIn, suggest people to recruit, and so forth. (Digital Sovereignty Leader Italy, Case B) Our empirical data underline the thought digital process through which adaptive organizations contrast the duplication of ideas (Cases C and G) or projects (Cases D and E) that may result from giving independence to individuals and basic organizational units.
Overall, our findings are summarized as follows: P3. Provision of information in adaptive organizations is achieved mainly through digital solutions that integrate information and ideas across the organization without the extensive use of social proof, providing greater transparency while avoiding the duplication of information.

Provision of rewards through incentives that foster dynamic capabilities
A further topic relating to adaptive organizations is the provision of rewards, which means focussing on the set of incentivesmonetary and non-monetaryused in a company to motivate its agents into executing their tasks to the best of their ability (Puranam et al., 2014). These rewards are designed to encourage both the integration of effort dispersed among employees and the implementation of individual intrapreneurial behaviour. In all the adaptive organizations we interviewed, these rewards are used to develop dynamic capabilities in their employees, continuously challenging them to sense, seize and mould the various opportunities on the market, within a highly effective organizational model. Cases A, D and E use a more traditional reward system to drive individual innovation. In Case A, nobody is rewarded in money for a new idea, but people are incentivized to uphold company values, which are strictly related to creativity and on challenging the status quo, being proactive and helping the company to adapt and grow: We have a variable compensation system for shared goals. There is no direct cash incentive for new ideas but there is constant focus on the creative part of one's work, on how to increase effectiveness, efficiency and quality, on how creative processes stop you from making do with existing processes, on the subject of continuous improvement and learning from mistakes, that everything can be questioned. These are company values and people are encouraged to uphold them (including by giving them financial sweeteners). As a result, we encourage continuous creativity / innovation indirectly. (CEO, Case A) Case A offers monetary incentives to keep its employees aligned with its corporate culture and induce people to develop their own dynamic capabilities and ability to sense market opportunities. Cases D and E use more explicit incentives to encourage individual entrepreneurial behaviour and innovation. Individual innovation is a key pillar of the corporate culture in Case D and is strongly encouraged, starting from what people are paid. Individual salaries are based on five elements, one of which is innovativeness: We absolutely expect and stimulate people to generate and propose their ideas. In fact, we evaluate individual performance on five elements and, a year ago, innovativeness became a key element. (CHRO, Case D) The compensation system in Case E strongly encourages employee innovativeness, considering it as decisive to the employee's career and salary. Employee creativity and innovativeness are taken into consideration in four of the eight areas in the payment structure.
We look for four qualities in our people and we want to work on these through our new individual compensation system: (1) be considerate, know how to maintain relationships and be a team player; (2) be curious, go and find out, explore different things; (3) be courageous, do not be afraid of challenging, (4) come up with new ideas, be creative, explore ideas and so on. We want to develop people's careers and salaries in eight different areas. Four of these relate to the four qualities above, which are about you as a person, how you behave. The other four areas involve more technical elements. (People & Culture Operations Manager, Case E) A more extreme way to boost a person's dynamic capabilities comes from Cases F and C. In Case F, the Insights Team explained that: We are trying to get more towards what we call the "50:50 club split model", which is essentially the idea that your circle is a small business, with its revenue and its expenses. Then, at the end of the day, you make some profit. And one of the things that you might want to do with some of that profit is spread it among your people, the circle members who have helped you make that profit, just as in a business. We're extending that model to where, if your circle makes a profit, you can split that profit 50:50 between a kind of bonus or employee compensation and the company's profit pot. (Case F) Case C goes even further in its approach towards giving a cash reward to employees as an inducement to develop their dynamic capabilities and intrapreneurial mindset. In fact, the HR Director for Europe and Russia argued that: The logic is this: a 40% basic salary and all the rest is variable. The main principle is that the basic salary is the exception, the variable part is the rule. [. . .] I could create my micro-enterprise, hiring internal recruiters. And I could be paid on the savings generated. This is very easy to do, as a microenterprise. I could be paid on this. (Case C) Our findings highlight the fact that these companies do not only introduce carefully crafted monetary incentives to foster people's dynamic capabilities and drive the company's corporate culture, in that a whole series of non-monetary incentives also come into play. These non-monetary incentives are strictly connected to each person and are key to employees embracing the company's way of working, and so be truly motivated to sense, seize and shape exciting new market opportunities. Thus, we conclude with the following proposition: P4. The provision of rewards in adaptive organizations is achieved through both monetary and non-monetary incentives geared toward seizing, transforming and making decisions about the opportunities sensed in the market.

Discussion and conclusion
Previous research (e.g. Puranam et al., 2014) has highlighted that, in turbulent environments, innovative forms of organizing enable companies to perform better, because they know how to use digital technologies (Puranam, 2018) and develop dynamic capabilities (Felin and Powell, 2016). Following these suggestions, we considered it promising to use the concept of organizational adaptation (Sarta et al., 2021) and the microstructural approach to organizational design (Puranam, 2018) to investigate six adaptive organizations, delving into how they organize themselves dynamically in their environments. We believe that our research contributes to existing knowledge, adding insights to the existing literature, whilst also offering a relevant model to practitioners.
5.1 Towards a dynamic solving of the universal problems of organizing Our primary theoretical contribution focuses on providing empirical insights into this topic by considering the universal problems of organizing proposed by Puranam et al. (2014). The intersection of this theory with the work of Felin and Powell (2016) allowed us to highlight the pivotal interplay among dynamic capabilities, organizational adaptation and polyarchy. Dynamic capabilities are often associated with strategy studies, while, from an organizational design perspective, there is still much to explore (Teece, 2018). Organizational adaptation is the ultimate goal in these organizations, and it implies, as suggested in organizational theory, that adaptation is an outcome. In other words, it is what an organization can change internally to achieve these outcomes despite any potential constraints. Polyarchy is also a construct and not a definition and is hailed as a new form of organizing, meaning that it allows for a good degree of generalizability. The solutions taken by the six organizations to respond to the four problems of organizing have led them to set up a polyarchic form of organizing. In a polyarchy, individuals and teams are empowered by being given full autonomy and independence, digital technologies are exploited intensely, and the company has a set of incentives designed to develop strong higher-order dynamic capabilities.
First, the managers interviewed focused on task division. Hence, they mapped the main objective of their organizationi.e. zero distance to customersinto a set of sub-objectives related to increasing agility, speediness and motivation. To accomplish this objective, they all implemented polyarchy each in its own way, with the aim of empowering individuals and teams, allowing them to organize themselves cross-functionally and/or have the time to explore new opportunities (Proposition 1).
Further, moving on to task allocation, the managers stressed that, by introducing a form of organizing that minimizes bureaucracy and hierarchy, it is critical to concentrate on the employees' mindset and capabilities right from the early recruitment stages. Since there are no traditional job titles and the way of working is adaptive, geared towards people's individual innovativeness, employees must be intrapreneurial, managing sub-tasks independently and/or proposing new ones (Proposition 2).
In our cases, the interviewees and our consideration on information provision revealed the pivotal role played by digital technology in minimizing duplication and enhancing transparency, so decreasing the importance of social proof. Social proof is seen as a limitation in some cases, whereas digital technology enables many tasks to be carried out automatically. While we note that our contribution on polyarchy converges with that of Felin and Powell (2016), we have also added a missing feature (the inclusion of the role of digital technology) to their work on designing organizations for dynamic capabilities (Proposition 3).
On the subject of provision of rewards, the interviews revealed that a system of monetary and non-monetary incentives can be also implemented on purpose to advance dynamic capabilities, and not only, as defined by Puranam et al. (2014), "to motivate the agents to cooperate by taking costly actions toward executing the tasks they have been allocated". Furthermore, these incentives help organizations to keep their employees aligned with their corporate culture, goals and purpose, as well as to leverage on individual innovativeness (Proposition 4). Figure 2 summarizes the above considerations.
Using the universal problems of organizing as a basic framework has helped us to simplify and ground a literature scattered among numerous research streams and definitions, clearly highlighting the novelty in the form of organizing found in adaptive organizations. Novelty, in this sense, may be a new set of old solutions or the understanding of complementarities among solutions to problems of organizing and does not have to be a theoretical novelty. As expressed by Puranam et al. (2014), the bar for novelty in an organization is significantly lower than the bar for theoretical novelty. As a consequence, when defining a new form of organizing, one may still find concepts where the literature already provides a solid foundation. It is rare to find something that is totally new in the world.
What is interesting and innovative about this work is that it provides empirical insights into organizations that focus on developing and nurturing strong higher-order dynamic capabilities, responding to the problems of organizing and wisely orchestrating more or less new solutions to gain and regain fit with their environment.
We found that the concept of polyarchy is an underlying construct of adaptive organizing, and social proofs are not relevant, as instead was argued by Felin and Powell (2016), for the integration of information and ideas. Digital solutions are to be preferred for various reasons, including because they promote greater transparency and enable the continuous, real-time monitoring of projects and market signals.
Our work strengthens the link between organizational design and dynamic capabilities, as suggested by Teece (2018) and Schilke et al. (2018), giving centrality to organizational adaptation. We argue that Puranam's framework should incorporate this concept and, consequently, the procedural dimensions of dynamic capabilities (sensing, seizing and transforming). In accordance with these findings, organizations should not only respond to the four problems of organizing, but do so continuously by enquiring into which solutions enable them to do the sensing, seizing and transforming for each problem.
This direction becomes even more relevant when we look at the consequences of the upheaval caused by COVID-19. This study was carried out during those dark days, and it is interesting to note how these organizations, on the one hand, got themselves ready by gaining Theoretical framework familiarity with digital collaborative tools and, on the other hand, used the time advantageously to grow and develop. This point is certainly something that could be explored by other scholars in future work. Other opportunities for further research are within the scope of this study. Among the several options, we wish to highlight that of interviewing other C-levels and managers to gain more insights into organizational and strategic matters or conducting longitudinal studies to test which higher-order dynamic capabilities are being developed over time to respond quickly to the market.

Insights from adaptive organizations
Considering the practical implications of this paper, CEOs and HR managers in traditional companies can use an actionable model to converge progressively towards a more adaptive form of organizing.
First, they can look at the different ways adaptive organizations introduce polyarchy and try out some of their ideas (e.g. temporary cross-functional teams to tackle business opportunities they may have spotted). Functional hierarchy is difficult to eliminate, and in some companies is still essential, but careful attention should be paid in exploiting the network of connections between people that is often at the basis of innovativeness. Innovations are usually crafted by cross-functional teams, so, whatever organizational structure is deployed by the company, it should be flexible enough to allow people from different parts of the organization to collaborate together. Managers could adopt digital solutions to give their employees the right information to make decisions, keep track of the projects, avoid duplications, etc. Moreover, they should consider that, in order to implement an adaptive, polyarchic form of organizing successfully, it is essential to have intrapreneurial employees. It is certainly worthwhile for a company to scout out the most suitable people and change their recruitment process, keeping an eye out for attitudinal and entrepreneurial skills more than or alongside technical ability.
Concluding these insights for managers, before implementing an adaptive form of organizing, they should act on four levers: (1) corporate culture, (2) digital mindset and governance, (3) the recruitment process and (4) reward systems. Corporate culture means establishing an organizational purpose and value systems that encourage employees to pursue corporate goals by fearlessly riding out change, thinking like entrepreneurs. Digital mindset and governance mean leveraging on digital technologies, where scattered employees can collaborate easily in teams, knowing they can rely on a flow of real-time and transparent information and so can engage more in value-added operations, e.g. seize market opportunities. The recruitment process means taking into account attitudes and capabilities which are required in the new system (entrepreneurial skills more important than technical skills). Reward systems means reviewing or designing a system to incentivize employees so as to engage them in "exploring" ventures, while measuring and rewarding people's individual innovative contribution, without penalizing anyone if things go wrong.
3. For more information: corporate-rebels.com