Accounting for change: the impact and implications of using the intervention-based research method GenderLAB at a big four accounting firm Free

Researching diversity and inclusion (D&I) is not for the faint of heart. It can, at times, be discouraging to bear witness to and analyze the often severe and negative effects of discrimination and to keep seeking and dissecting the often banal—yet so deep—roots of such discrimination, with the aspiration to further current academic debates. At the same time, D&I research also holds immense potential for change, and when our research does change the status quo and does make a difference, it can feel extremely rewarding. However, such occurrences can be rare, and few in-between. The academic research process is a long and winding path toward creating impact, and it often takes years between data collection and research publication (if the article is published at all).

It has to be this way—or so we are told ^[1]—because we must ensure research excellence. According to the traditional perception of research excellence, researchers are supposed to deliver solid and impactful results that set new standards for academic thinking. Moreover, we should have methods, processes, and practices in place to ensure research excellence and valid, reliable research findings. This entails, amongst others, the ideals of objectivity and impartiality (Basu, 2006; Van Raan, 2005). Similar to how a film crew that is filming a lion chasing an antelope in the savanna does their work without interrupting the animals, we are often told to exert as little influence on the research site as possible, as this would taint and disturb the data collection and thus distort the research process, consequently endangering our findings and, ultimately, the research impact. Our purpose is, therefore, to collect data in the service of finding the truth and solving puzzles, in addition to increasing, improving, and updating research knowledge. Research is supposed to be unbiased, value-free, reliable, and factual (see Resnik, 2001, for an overview of research objectivity); and only then we might consider impact at the research site itself.

However, this “rule of excellence” has recently been receiving increasingly louder criticism, as it not only reproduces inequality (Deem, 2009; Husu, 2004; van den Brink and Benschop, 2012) but also limits the reach of its impact to smaller (and very privileged) audiences. Ashcraft (2017) calls the rule of excellence a product of “the white, North American, English-fluent, heterosexual, masculinist norms of intellectual culture” (p. 37) and directs harsh criticism toward a system that seems closed in on itself, allowing it to be impacted very little beyond the ivory towers of North American—and, to some extent, European and Oceanian—universities. In this vein, recent criticism has also been directed toward how ideas on research excellence and the “publish or perish” rule have created a competitive, stressful, individualist, and unhealthy work environment. As Gill (2009) explains, “You are only as good as your last paper, and that now has a half-life that is shorter than ever” (p. 238).

Criticism of excellence, however, is tricky. In 2012, Butler and Spoelstra pointed out, that the critical management researcher is so profoundly entangled within systems of excellence in the university that it may no longer be possible, strictly speaking, to ‘respond’ to this encroaching commercialization” (p. 892). Ashcraft continued this criticism in 2017, stating, “If you lean critical, you harbor a special ambivalence and irony about participating in this ever-enterprising enterprise. You know better but may not know how to do better (be sardonic while working harder?). Like me, you may long for more” (p. 36).

Yes(!), we reply to Ashcraft. We do—like you—long for more! In this paper, we embark on a journey to formulate our stance on what this “more” may entail. Ashcraft’s “longing for more” consists of developing a critique from affect theory; in this critique, she proposes “an alternative posture of resistance rooted in ‘sense-abilities’ of home, field, and their relation” (Ashcraft, 2017, p. 37). In doing this, Ashcraft is occupied with the home base, the processes of our workplaces, and how they construct inequality and exclusion through the rule of excellence. In this paper, we draw on Ashcraft’s and others’ (e.g. Van den Brink and Benschop, 2012) critique of research excellence, but we direct our attention more specifically to the method of this presumed excellence: the research process at empirical research sites. In our longing for more, we contribute to the critique of who benefits from the conception of research excellence and question the value of impact in these long research processes. However, we take this one step further and emphasize the importance of the local impact that intervention-based methods can have. To further clarify this point, we now turn to the role that participatory research has played in the critique of traditional notions in terms of which methods lead to so-called research excellence.

Central to the critique of current definitions of research excellence is participatory action research (e.g. Glassman and Erdem, 2014; Reason and Bradbury, 2012), particularly the idea of non-extractive methods. This is brilliantly explained by Da Silva et al. (2022) as a perspective that “enable[s] the construction of knowledge ‘with’ the participants/individuals of a particular social context, rather than ‘about’ specific individuals, in a distant manner, and brought forth by the researcher” (p. 2). Thus, instead of simply leaving the research site with audio devices filled with stories told to us and notebooks full of evidence and reflections, we can do more. As opposed to merely perceiving research participants as “objects of research to be extracted, captured, measured and quantified” (da Silva et al., 2022, p. 2), we believe that, when working with D&I, we have a special obligation to do more than benefit from the hard lessons, heartbreaks, and—quite often—pain and confusion of those who have generously taken the time to help us learn about their lives, work, and organizations. Therefore, we propose that it is part of (especially) D&I research itself—and of creating research impact more generally—to try to leave our research sites in a better state than we found them in ^[2].

The purpose of this paper is to explore how we can create such more immediate impact at research sites. Therefore, we report on and discuss the impact derived from one such intervention method, which we call the GenderLAB method. The GenderLAB method was developed in a collaborative research project across Finland, Sweden, and Denmark, and in this paper, we explore how we have used this method in our research (first in a project with a large accounting firm and later with a large variety of organizations), what local impact it has had, and how other researchers can benefit from using this method.

The background of the research interventions focused on in this paper is a project that we were involved in from 2018 to 2020, which was named “Co-creating Gender Equality from Classroom to Organization: Innovations in Nordic Welfare Societies” and ran from November 2018 to March 2020. It was sponsored by the Nordic Gender Equality Fund and was realized by a collaboration between Copenhagen Business School (CBS), the Royal Institute of Technology in Stockholm (KTH), the GODESS Institute at Hanken School of Economics, and KVINFO, as well as the equality consultants Ekvalita and ConnectUz. The purpose of the project was to develop a workshop method to advance gender equality.

As a first step, the research group developed a workshop format combining norm critique and design thinking. This combination was intended to ensure critical reflection (coming from norm-critical exercises) and to foster feelings of agency and willingness to engage, in addition to producing concrete solutions for change (as a result of the design thinking process). The workshop method was tested in workshops in Copenhagen, Stockholm, and Helsinki in 2019. The participants in our workshops were selected to be as broad as possible and included business and academic leaders, educators, managers, students, volunteers, non-governmental organizations (NGOs), educational institutions, and youth/student organizations. Given the method and structure of the workshops, there was a strong collaborative theme, which allowed the participants to discuss workshop topics in a shared forum, challenge gendered assumptions and organizational norms, and, most importantly, co-create informed solutions to these issues. Consequently, the participants created particularly rich workshop outcomes, which the project partners understood as data that were useful for research purposes and for creating a general method that could address gender-based organizational challenges in a Nordic context. On analyzing the gathered data, after each workshop, the method was continuously refined, and in December 2019, the finalized outcome was presented and made publicly available in different print and digital formats at a large closing conference in Copenhagen. All tools and manuals, which were made public at the Copenhagen event, can be found on CBS’ and KVINFO’s webpages ^[3]. The data from the launch event were also documented in a research publication (Christensen et al., 2021).

The resultant workshop methodology is a combination of knowledge and processes or, more specifically, a combination of the most recent research on unconscious bias, gender, and discrimination with norm-critical approaches and design thinking. Norm-critical approaches refer to approaches that challenge the status quo and uncover where seemingly benign everyday norms create harmful effects and unintended negative consequences (Christensen, 2018; Holck and Muhr, 2017; Plotnikof et al., 2022). Design thinking is a method for addressing complex problems and generating innovative solutions (e.g. Villeséche and Teilmann-Lock, 2023). This combination yielded a workshop structure consisting of three elements: (1) educational mini-lectures on, for example, structural inequality, discrimination, and bias, (2) norm-critical exercises meant to spur reflection and foster empathy, and (3) a detailed design thinking process intended to make the participants co-create unique solutions to challenges they perceived as important. The following sections detail how we have used the full workshop format, as well as a more condensed version of it, and what local impact it has generated (so far).

The first research case to use the GenderLAB format upon its completion was a Danish branch of a Big Four accounting firm, hereafter called Blue. One of the authors had been engaged in a long-standing research collaboration with Blue, as she had conducted 30 interviews in the organization during 2009–2011. Similar to other accounting firms, Blue had already been struggling with recognizing, valuing, and promoting women talents. The research conducted from 2009 to 2011 prompted awareness of the fact that the near-complete absence of women in leadership positions, especially at the partner level, was not natural. Furthermore, it was not because women were lacking in talent, skill, or ambition vis-à-vis their male colleagues; rather, there was a masculine work culture that had created norms that kept women from climbing the ranks. However, in 2019, just before the COVID-19 pandemic, the firm had still not managed to move the needle on the number of senior women in the organization to a notable degree. Despite Blue’s recruitment at the lower levels having reflected gender equality for years (40/40/20), it had only approximately 10% women partners. In 2019, a global leadership directive demanded that all local branches work toward what was announced as a new requirement for gender equality, namely a minimum of 25% women in partner and board positions.

This was perceived as a substantial challenge for Blue, and the firm contacted the researcher it had been involved with in 2009–2011. Together, we agreed that merely knowing that masculinist norms create barriers to women’s advancement was not enough and that it was necessary to disturb those norms widely in the organization and create practical tools and methods to change daily working practices and habits. Therefore, we determined that the GenderLAB format was ideal. Furthermore, it was also an ideal way for us, as researchers, to test it more broadly within a firm that had not been part of developing it. However, because Danish workplaces can often be characterized by a disbelief in gender norms and gender biases and because at that time, a certain fatigue characterized the gender equality discussions in Denmark more overall, the researchers and the firm agreed to focus on “diversity of thought” and change the name of the method from GenderLAB to DiversityLAB. This was intended to minimize the potential (anti-woke) backlash (see, e.g. Roberson et al., 2024) arising from a discussion centered specifically on gender and to focus more broadly on the many different biases and types of discrimination that characterized Blue’s advancement systems. Gender was certainly not the only organizing mechanism of advancement; racialization, nationality, age, and class, among other factors, also played important roles in determining who could become a leader, seeing as most partners were Danish white heterosexual men from similar socioeconomic backgrounds. To meaningfully disrupt this status quo and to plan the intervention process, the chief executive officer (CEO), the human resources (HR)-responsible partner, and the first author agreed to collaborate.

Analyzing the empirical material previously gathered at the firm as well as the literature from the D&I field made it evident that the lack of diversity and the persistent standard of preferring white Danish heterosexual men over others for promotions, especially at higher levels, were rooted in exclusive norms concerning what it meant in this firm to be an “ideal accountant”. In line with previous literature on accountancy, we found the “ideal accountant” in this firm to be white, middle-aged, upper middle class, abled-bodied, Christian or non-religious, and male (Kornberger et al., 2010; Lehman, 1992; Storm and Muhr, 2023). Using this analysis as a starting point helped inform how we collaborated with Blue’s leadership, and together, we planned an intervention to achieve more meaningful and sustained change.

As the next step, we collectively agreed that more voices within the firm needed to be heard—and involved—in this intervention. To further prepare for the intervention, we decided to collect more recent data from the firm across hierarchy levels and departments. Therefore, the HR department, in concert with us, started a new data collection process by conducting ten qualitative interviews and nine focus group interviews across the firm’s five offices in Denmark. These interviews and focus groups centered on the following questions: When have you felt included or excluded in this firm? To what extent do you feel (or not feel) supported in pursuing your career at Blue? Are there any barriers you face when it comes to promotions? When do you feel that you have (or do not have) your manager’s support? How can Blue become more embracing of the diversity that is already there among the employees? The focus groups consisted of approximately 5–6 employees, each facilitated by someone from lower management levels; there were participants from all lower levels as well as representatives from all departments. Considerable care was taken to compose these groups in a way that mixed employees across a wide range of units to avoid them having to disclose private information to their closest colleagues. Furthermore, a firm-wide call encouraged employees to anonymously submit their own experiences with feeling included or excluded in the workplace to us. We received 26 (approximately 6,000 words in total) such stories, which we (if it had not been done already) anonymized and added to the data pool.

The interviews and focus group discussions were recorded by the HR department, and we transcribed and coded these as well as the anonymous stories. Based on these codes, we produced a preliminary analysis of the diversity and inclusion (D&I) challenges faced by Blue. This analysis clearly showed the specific social norms that employees had to contend with, for example, when they wanted to be recognized by their superiors or when they were evaluated for an upcoming promotion round. Furthermore, we could show how these norms emphasized being white, heterosexual, physically fit, Danish-speaking, male, and a parent—but one free of any substantial child-care obligations. In addition, it was evident that these norms were irrelevant in terms of “good” accounting work, with objectivity, correctness, completeness, and timeliness being some primary hallmarks of that.

We presented these results to the HR-responsible partner and the CEO, and as a group, we decided to use these findings to frame the planned intervention to ensure that the intervention would be conducted with Blue’s D&I challenges in mind. It was decided to use many direct quotes from the empirical material to give the problem a clear voice and ensure that no one could claim that discrimination was not an issue at Blue. The frame “diversity of thought” would serve as the intervention’s overall theme to broaden the focus from gender to a more multifaceted and intersectional view of diversity.

To reach all leaders in the firm, the HR-responsible partner and the CEO decided that the intervention would be staged at Blue’s yearly leadership summit, for which all leaders would be called to the company’s headquarters for a full day. Participation in this summit was mandatory; no client or other work emergency was an acceptable reason to not attend. Finally, it was agreed that the intervention should not only help initiate a far-reaching cultural change but also produce more insights about the firm and its struggles and opportunities. In the next section, we detail how this was achieved with the DiversityLAB methodology.

The intervention with Blue was conducted in December 2019. It lasted a full working day and occurred as part of the annual leadership summit for the 140 most senior leaders in the firm (Blue has approximately 3,000 employees in Denmark). We designed and facilitated the intervention, in addition to aggregating and analyzing all the data acquired. We were not paid to conduct the research. However, everyone received a fee for taking a full eight-hour working day to facilitate the intervention and for producing a report documenting the results produced during the intervention. After the report was delivered, it was agreed that all data generated before, during, and after this intervention could be used freely for research purposes, on the condition that the data would be anonymized. Therefore, this article constitutes a research output, which as such is fully separate from any control by or payment from the company.

Due to the large number of participants in the intervention, we teamed up with a group of colleagues (whom we instructed in the GenderLAB method) to guide the intervention during the day of the summit, answer any questions, and ensure that the participants stayed engaged and on their tasks. The researchers involved were also tasked with moderating the discussions that occurred once the participants were divided into smaller groups. The moderators were specifically briefed on the harmful social norms that had been identified previously. They were instructed to challenge the participants if they argued in a way that seemed to reinforce those norms and also asked to prompt the participants to reflect on what they perceived as “normal” or “natural”.

After a brief welcome speech by the CEO, the intervention began with an introductory presentation by the HR-responsible partner, who introduced and detailed some of the key findings of the focus groups, providing quotes and emphasizing how the norm of the ideal accountant was a problem that was not only directly at odds with Blue’s objectives but also hindered the company in attracting future talent.

This was followed by a short presentation by the first author who introduced themselves as the workshop leader and timekeeper, directing and guiding the intervention from the main stage, while the facilitators would provide immediate guidance and help troubleshoot any issues directly with the participants. The first author explained the idea of norms as social constructs, how norms can inadvertently create exclusion and hinder diversity, and how such discriminating norms can be changed using norm-critical exercises to foster D&I within Blue. It was further explained that the participants were expected to actively participate in the intervention and that any materials they created should be anonymous (or, if they nevertheless shared their names, would later be anonymized), as those would be used for giving recommendations to Blue’s leaders and for publishing research articles such as this one.

Subsequently, the audience was divided into groups of 5–7 participants, with each group sitting around a table. With this, the intervention moved into the phase in which the participants actively engaged in a number of timed activities. The first activity involved filling in information about themselves using a handout titled “Circles of Multi-dimensional Self” (see  Appendix 1); pens were provided to those who wanted them. In this activity, the participants reflected in writing on the identities and identity categories that they used to describe themselves or that others used to describe them in terms of, for example, their gender, age, social class, work title, family status, or educational background. Within their groups and guided by the facilitators, the participants talked about which categories (e.g. woman-married-IT or single-father-lawyer) they associated themselves with and how those were connected to certain expectations (i.e. social norms). Following this, they were asked to formulate written statements about instances in which they felt that they had broken such expectations (e.g. I am a mother, but I am not interested in working part-time or I am a senior accountant, but I am not a pedantic bean-counter). The facilitators encouraged the participants to reflect on how such expectations could be misguided and exclusionary.

With these reflections fresh in mind, the participants moved on to the second exercise, which involved formulating a clear problem statement. The facilitators guided the participants to formalize their own understanding of the problem of lacking diversity within Blue. The participants received two blank post-it notes each. They were asked to write one sentence that summarized the problem from their own perspective on one of these notes and to draw an image of what the problem looked like on the second note. This latter step was met with a degree of vulnerability, as some felt it embarrassing or uncomfortable to draw. Facilitators were in advance trained to handle such situations, for example, by explaining why drawing was important (it activates a different part of the brain than writing) and providing encouragement and inspiration, for example by reminding participants that stick figures were ok to use. Once everyone had written a statement and created a drawing, the participants presented these to each other. Guided by the facilitators, the participants discussed these outcomes within their groups, at the end of which each participant needed to choose one statement or drawing that they resonated with the most (they could choose their own). Drawings and statements that were not chosen were collected by the facilitators and set aside.

The next exercise was individual and fast-paced: The participants were each handed an empty sheet of paper as well as several blank post-it notes. They were instructed to paste their chosen problem statement (written and/or drawn) on top of their sheet of paper. Next, the participants were asked to come up with as many solutions to their chosen problem as possible. Each new solution needed to be written on a new post-it note and stuck onto their sheet of paper. The facilitators encouraged the participants to devise as many solutions as they could, asking them not to think but to continuously keep writing and sticking notes. Every two minutes, the participants passed their sheet of paper with their chosen problem statement and solutions to the person on their left and, in turn, received a new sheet of paper from their right neighbor, with a different problem statement. The participants were again encouraged to find as many solutions as possible to the new problem for the next two minutes, after which they once again passed their sheet of paper to their left. This process continued until each participant had written solutions to each problem statement at the table.

In the next exercise, the pace slowed down again. The participants were instructed to look at and discuss the solutions they had come up with at their table. Guided by the facilitators, they reflected on the solutions they and others had found, and the facilitators pointed out the creativity present at the table.

In preparation for the next activity, the facilitators passed an envelope around each table. Each envelope contained around 10 small slips of paper, from which each participant randomly drew one. Each paper had an identity written on it, similar to the first exercise of the day, each new identity contained a couple of random identity markers, for example Muslim-mother-accountant or homosexual-male-partner. Following this, the facilitators guided the participants to identify with this new role, asking them to imagine what it might be like to work at Blue in this role and then to provide solutions to the problem statements from this person’s perspective rather than their own. In total, 2,600 post-it notes with ideas were generated, some of which can be seen in Plate 1.

After the intervention, a few facilitators reported a degree of resistance to this last part, whereby some male participants for example rejected taking on the role of the homosexual partner. If they could not be convinced to temporarily take on this role, these participants could draw another role from the envelope and facilitators made a mental note to mention this in an anonymized way later, during the collective de-briefing amongst the intervention organizers.

Subsequently, the audience came together in plenum to listen to five short stories being read out on the main stage. The stories were fictional but inspired by the interviews, focus groups, and anonymous stories collected prior to the workshop. Each story was written from the perspective of a different imaginary person, and the wording was specially crafted to invoke empathy for these characters, as each story detailed a situation in which they experienced exclusion in an everyday interaction. To emphasize normative pressures, each story was read by someone with identity makers different from those of the protagonists of the stories. For example, a middle-aged white man read a story about a prospective mother whose superiors were suddenly encouraging her to go home early and to consider switching to a part-time position even though she had long been working toward becoming a partner. In addition, a young and visibly muscular man read out the story of an elderly man struggling with constant flare-ups of an invisible illness, which made him feel excluded from after-work activities as most of those centered on physically demanding activities, such as CrossFit or cycling. Through these contrasting traits and narratives, it was possible to affect the participants and to make them feel how what one might accept as normal behavior from a certain body might not be so normal after all. The effect of these stories was palpable, and many in the audience appeared to be moved. The facilitators later reported that this part had created empathy among the participants for what were entirely new perspectives to many of them.

Following this, the audience returned to their original groups. All groups were instructed to look at all solutions created at their table and then collectively choose three solution ideas (i.e. three individual post-it notes) according to the following criteria: (1) What can be done tomorrow? (2) What should be done in the long run? (3) What is the most unforgettable/wild idea? Once these three solutions were determined, each group received three identical idea forms (see  Appendix 2). On these, they had to formalize how they would realize each of the three chosen ideas. In closing, the groups could opt to present their three solutions and the implementation strategies for these to the plenum. At the same time, using Padlet software, all participants were invited to submit their comments and ideas online; this could be done anonymously, and we received 39 extra comments in this manner. Finally, the moderating first author wrapped up the event and closed the intervention by thanking all participants and reminding them that their countless solutions would be collected, aggregated, and fed back to Blue’s top management.

The data analysis occurred in a number of iterative rounds. In a first step, immediately after the intervention ended, the authors of this article and all other facilitators engaged in a debriefing session in which the immediate observations and experiences of facilitating at the different group tables were shared. These immediate reflections were noted down by the second author. In addition, all materials created by the participants during the intervention—approximately 2,600 post-it notes with concrete solution ideas to change the existing culture and 57 filled-in idea forms—were collected by the second author.

In the next step, the second author sorted and digitized all these paper materials, which resulted in a 367-page document. Furthermore, those involved in running the intervention were called to provide written accounts and observations, and we obtained detailed notes from all 14 moderators and facilitators (approximately 50 pages of written text). All these materials, along with the submissions received on the Padlet, were read by the second author, who created a list of emerging themes; these included codes such as “work-life balance” and “locker room humor”.

Subsequently, we further refined these themes in collaboration with the facilitators and co-organizers of the intervention. During this process, the following was a central question: How can we, as researchers, use this material to create a meaningful and lasting impact within Blue? It was decided that for the intervention to have a continued impact, beyond the day spent with us researchers, it needed to create outcomes that could be easily understood, so results needed to be in Blue’s employees’ own words as much as possible. Furthermore, results needed to be not just meaningful but also formulated concisely and communicated in a way that made them easy to share, so that they could reach as many within Blue as possible.

Accordingly, we developed the entire analysis strategy, which was three-pronged: (1) providing an external view of Blue, for which the facilitators’ statements and observations were used, (2) providing an internal view of how the problem of lacking diversity was perceived, for which the problem statements were aggregated, and (3) providing practical ways in which the problem could be addressed, for which the solutions and idea forms generated on the day of the intervention were used.

First, we aggregated all facilitators’ statements and distilled five especially meaningful quotes that seemed to very aptly address how we perceived Blue on that day. For example, some praised the participants’ willingness to engage while also pointing out less agreeable tendencies, such as how leaders across different tables rejected taking on the role of the homosexual man. In addition, we wrote a short accompanying text highlighting other instances in which we observed resistance from the audience, showing how Blue’s problem was rooted in an extremely strong and very limiting normative understanding of who and what was considered welcome at Blue.

Second, all problem statements were clustered into different themes, which were refined in discussions between us and other facilitators as well as presented to Blue’s HR-responsible partner and the CEO. This yielded ten overarching themes, and we provided examples for each theme. For instance, under the theme “space for diversity”, we provided examples such as “no female role models”, “no good options for vegetarians at the canteen”, and “lack of diversity of competencies”. These were the words of Blue’s leaders and showed that its employees struggled with the very restrictive normative pressures in their workplace internally as well.

Third, we organized all solution post-it notes along a two-by-two matrix (see  Appendix 3). On the x-axis, we noted down the estimated time required to implement a given solution; on the y-axis, we noted the potential impact of a solution based on how many diversity categories, such as age, family status, gender, and nationality, it included. This allowed us to create a heatmap (see  Appendix 3) of what we deemed to be some of the more meaningful solutions: Solutions that were quicker to implement and included more diversity categories were “hotter”, which we marked in red colors. An example of this was “anonymize CVs in the first screening round”. Solutions that took longer to implement and included fewer diversity categories or in which diversity categories were not very clear were “colder”, which we marked in yellow and green colors. An example of this was “new pricing model”. As there were far more red solutions than yellow/green ones, especially such green solutions in the lower right corner are very few. The result was a visual of the solutions, similar to the temperature maps shown in weather forecasts. The process of organizing data along axes and marking clusters of data points (solutions in this case) in colors was somewhat similar to Blue’s method of organizing and processing data. In addition, we provided a full list of all the solutions created.

We aggregated all our findings into an idea catalog, which was a brochure created by a professional graphic designer. From this, different parts—such as the heatmap—could be easily copied and pasted into other documents or presentations. The HR-responsible partner and the CEO were extremely happy with this idea catalog and its content, and we were told that the catalog and especially the heatmap were found to be very impactful. This is detailed in the next section.

In this section, we highlight the impact of the intervention on Blue. We want to preface this by stating that we believe that, similar to how diversity and exclusionary social norms are difficult to measure, the impact of a norm-critical intervention such as ours is difficult to quantify. Our intervention aimed to pull people out of their comfort zones, and yes, some felt insecure or even embarrassed, for example, when they were asked to think from the perspective of a homosexual man. Others openly embarrassed themselves when they vehemently, and in front of their colleagues, rejected taking on such a role. There was space for all such reactions as well as reflections afterward, as explained below. This also means that we cannot capture our impact in numbers, graphs, or charts. Has the number of homosexual men working at Blue increased or decreased—and, if so, by how much—after the intervention? How many leaders have become aware of their own unconscious homophobia? Did Blue experience a decline in heterosexism after the intervention? We do not know, but we do not believe that this detracts from the impact the intervention has made. This is exactly where we believe the strength of interventions such as GenderLAB lies, as it, like diversity itself, is beyond measure.

Nevertheless, the intervention at Blue created both immediate and long-term impact. From the materials produced and the moderators’ reported reflections, it was clear that, during the intervention itself, there was a shift in the conversations among the participants as well as in the scope of the solutions that they produced. It was noticeable how, during the intervention, attitudes shifted from a lack of awareness about the prevailing social norms and their impact to more critical reflections of what was considered “normal” and how to achieve higher degrees of inclusion.

Moreover, after the intervention and in our follow-up interviews, the HR-responsible partner explained how the idea catalog and especially the heatmap served as tools for accountability, as they could be used to remind the leaders that they themselves wanted and had articulated specific ways for change. Other partners reported on bringing select exercises from our intervention to their own units to initiate conversations about norms, the diversity of identity markers, and inclusion. Furthermore, we could see that some of the more short-term solutions created and refined during the intervention had already been implemented. Regarding this, the respondents stressed how easy—and often cheap—certain changes were. For example, a “quiet room” was quickly established, using an empty office and some unused furniture, originally intended to be a place where people could pray. It was a pleasant surprise to the respondents that such changes, though originally intended to make Blue more inclusive in terms of diversity, actually benefited a majority of employees. The quiet room soon became popular, with people using it to meditate or organize their thoughts, and it was soon moved to a larger space. The room is now used for various inclusive purposes. In addition to those engaging in prayer and meditation, people who have suffered a mild brain concussion use it when they need some rest, mothers returning from maternity leave use it when they need to express milk, and neurodivergent individuals use it when they need some space. In other words, the room has made it possible for a significantly broader range of individuals to work—and feel included—at Blue. Furthermore, without any announcement and practically overnight, more vegetarian and vegan options appeared on the lunch menu, and signs were added next to each dish on the buffet, indicating the ingredients and thus helping everyone—rather than only those who did not eat meat—make more informed choices. This was interestingly one of Blue’s management’s greatest discoveries in terms of privilege blindness. Introducing more vegetarian and vegan options, with English signage, was quite easy to implement, they had just never thought about it. However, doing so had profound inclusive consequences for the employees, who could now eat more than just a salad for lunch and feel secure that they were following their dietary needs. The long-term impact of this could be (though we do not know this for sure) that the “easy” implementations, such as the quiet room and vegan and vegetarian options for lunch, would lead Blue’s employees to become curious about and investigate what their colleagues need to feel included. After all, the lesson learned here was that, if something is not a problem to you personally, you will not understand it unless you ask people who are different from you about what they need.

In the medium term, initiatives that had already been successful in Blue’s overseas offices were implemented in its Danish branch: So-called employee resource groups began sprouting, and during our various check-ins with Blue after the intervention, we learned about the Muslim network, the LGBT + meetups, the parents@Blue group, and many others. In a follow-up interview in 2024, the HR-responsible partner explained,

Apart from providing financial support to help coordinate such initiatives, Blue was also one of the first companies in the country to offer equal paid parental leave to “mothers, fathers, and co-parents,” with care taken to keep the wording of the policy as inclusive as possible. Shortly thereafter, the numbers showed that fathers began taking longer parental leaves. There is great hope that this trend will continue in the long term, thus slowly dismantling gendered stereotypes and expectations about parenthood and career choices.

In addition to the HR-responsible partner, we interviewed two other partners in 2023—one leading internal services and one overseeing client-facing advisory. After speaking with them, we believe there is more awareness now of what is considered normal and what kind of working and behavioral norms Blue’s employees want and do not want to accept. Blue remains a white-collar and male-dominated firm that struggles with exclusionary conceptualizations of the “ideal employee” (Kornberger et al., 2010; Lehman, 1992; Storm and Muhr, 2023). However, this ideal is no longer sacrosanct, and many initiatives seek to change this normative ideal across the short, medium, and long term. While not all of them may have directly originated from our intervention, those we spoke to agreed that the intervention made this ideal visible and also showed that it is possible to change it. This created interest in and momentum for initiatives such as the ones detailed above, which, in turn, were taken seriously.

A critical reader may correctly note, like one of our reviewers, that the solutions were all defined by managers in a top-down effort. This is undoubtedly true and a limitation of this intervention. As discussed below, we have conducted other interventions in which broader selections of employees were present, but these are two different formats; each has its own advantages and disadvantages. The advantage of the format we used with Blue, in which all managers were present, was that it created collective accountability for the solutions. As they were developed collectively by the entire group of company leaders, they were much more easily implemented afterward. One may further note that the implemented solutions seem small in scale, especially in comparison to what the leaders had originally proposed during the intervention. However, we argue that, in a firm as conservative and careful—if not outright hesitant—about D&I changes as Blue, the “big dreams”, as captured in the heatmap, are crucial for daring to implement reasonably low-risk changes in the first place. Because that is what they are—risky: Nobody wants to be responsible for an initiative that publicly failed and, especially in the case of our participants, not in times when even the most well-meant D&I efforts can lead to unforeseeable, public, and devastating backlash. The fact that these leaders nevertheless dared to stand up for what they believed in and to learn from, yes, reasonably small changes shows that our intervention changed what these leaders believed to be possible. This is also why the title of the paper is “Accounting for Change”, defining the solutions collectively makes leaders collectively accountable in a way that then also makes change more possible.

Similarly, a critical reader may (again) correctly note that diversity practices, such as the GenderLAB method, have long been criticized for not making the structural, systemic, and cultural changes needed for organizations to become truly inclusive spaces (see, e.g. Roberson et al., 2014). While we, as D&I researchers, completely agree with this and are always hoping for more numerous and profound changes, we must admit that, as in our roles as D&I practitioners (activists), we see sustainable change in such small, collectively defined initiatives. Therefore, considering the conservative industry and the national context of Denmark—which is a very homogeneous country that exhibits high degrees of bias as well as reluctance to implement change—we perceive the changes defined by Blue as small but, by virtue of their very resistant existence, significant. Furthermore, while the solutions were identified by management, the challenges they sought to answer were identified by employees from all levels and departments across all Danish offices. We may, of course, be optimists, but we also doubt that we could remain working with D&I, as both researchers and practitioners, otherwise.

Apart from its use within Blue, we have further adjusted and refined the format of GenderLAB to make this method of changing social norms and opening up spaces to diversity accessible to as many as possible, far beyond “just” gender. We have held free seminars that were open to the public, during which we instructed well over 150 voluntary participants in the general LAB methodology. Furthermore, we have provided all LAB exercises, worksheets, and scripts to the public, free of charge (under the condition that the LAB method is used only for non-commercial purposes). We have also shown how to shorten the LAB method (called MiniLABs) and how to tweak the LABs for different purposes and audiences (QueerLABs, InclusionLABs, etc.). To continue this effort, we provide our handouts and describe our LABs in this article and hope that others will want to use our methodology, allowing it to travel beyond those we have reached with it so far.

One of the main learnings from the way GenderLAB was designed and tested originally, as well as the way we used it in Blue’s DiversityLAB, is that it is a very complex method and requires facilitators in each—or at least every other—group. Regardless of the professional context (a university, the armed forces, a consultancy, an NGO, and Blue), the groups were not able to follow the assignments correctly unless they had a facilitator close by. Consequently, it is difficult to use this method for research, as we typically do not have the budget to cover individual facilitators. Therefore, MiniLAB was developed as a reduced version and has been successfully used at least 25 times since the Blue intervention in 2019. This version requires only one facilitator, who facilitates from the stage, and many post-it notes. The format is presented on one slide, one bullet at a time, as shown in Plate 2. The times can be adjusted to fit the individual format and group size.

The main difference here is that, instead of choosing a problem statement individually and circulating the various problem statements, one of the group’s problem statements is chosen from the beginning, and they work with this for the rest of the time. For the sake of data collection, problem statements are written on large post-it notes, while solutions are written on small ones; this allows the facilitator/researcher to get an idea of all the problem statements that have been formulated, as opposed to just the one worked on. Another modification we often utilize to make the format simpler is to skip assigning a role. This step requires significant facilitation, as this is needed to make participants take on roles with empathy and seriousness and to intervene if participants make fun of or refuse a role.

Every time we have used this method over the past year, it has generated engagement, ownership, and empowerment, as well as surprise over how many brilliant and easily implemented solutions employees can generate if they are simply asked to. Thus, the method has proved extremely useful in assisting people across all hierarchies in expressing opinions that are usually not voiced. This method has been used in large private multinationals, local construction firms, NGOs, security organizations, and unions, among others. Furthermore, group compositions and sizes have varied widely, as we have used it for small groups and large groups as well as homogeneous (e.g. managers divided into their departments, as in the Blue example) and heterogeneous (where levels and departments are mixed) groups. Thus, while the method was generated for research purposes, it has already started to create change in a wide range of organizations. As a manager from a construction firm recently told us, it has made an impact by “spreading rings in the water.” Furthermore, as he elaborated, these rings originate from all corners of the organization, indicating that D&I initiatives and policy implementations become a visible result of a bottom-up process, which creates a sense of agency regarding—and ownership over—the company’s D&I agenda. It is no longer something that just comes from the top in the form of a strategy paper or a policy.

This paper aimed to show how interventions, such as GenderLAB, can be used to generate local impact and how the field of D&I research can benefit more broadly from such methods. We have done this through a detailed description of how we have conducted this intervention-based research method in Blue and beyond, in addition to explicating how impact has been created throughout this process. We have thereby contributed to the growing literature on participatory methods, particularly the streams arguing for non-extractive methods that not only take data but also give knowledge to research sites.

Considering the ethics of this particular form of impact entails the question of who we are morally responsible for in the research process and what such responsibility may look like. When we reconceptualize impact to extend beyond research excellence and the impact of finished research to include the impact of the research process itself (e.g. on the research site), it creates difficult questions of research ethics with regard to who the impact is valuable to and who owns it. In our interventions, ownership is especially interesting, as the researched organizations have the right to use and implement as many of the produced solutions as they possibly could, but we have the right to use the outcomes as well as research data for publication purposes. Thus, the entire idea of ownership over the impact has been dissolved. To us, this is where the very core of the ethics of impact lies. It is not merely a question of ownership but a moral responsibility, maybe even an obligation, to make sure that we, as (D&I) researchers, use our privileged position to ensure that the research impacts not just our careers but also leaves our research sites in a better state than we found them in. In these types of research interventions, practicing good research ethics does not require objectivity, distance, and non-interference; rather, empathy, care, and involvement are key. It requires us to be critical with ourselves and analyze our own actions from several lenses and perspectives while simultaneously being careful with the people with whom we engage, accommodating and addressing their discomfort and not giving up on them, no matter how large or small or fast or slow their progress may be.

The value of such impact is no longer only what benefits an academic career or an academic institution; in essence, it is no longer the case that a researcher’s sole purpose is to produce 4* journal publications. The ethical value of this kind of impact is to be found in the way rings spread in the water—both the ones we can measure and the ones we can sense.

Figure A1 

Figure A3