Authors:

Emily Truelove – MIT Sloan School of Management

Katherine Kellogg – MIT Sloan School of Management

Interviewers:

Kevin Lee – Leonard N. Stern School of Business, New York University

Laura Dupin – EMLYON Business School

Article link: http://journals.sagepub.com/doi/full/10.1177/0001839216647679

Question 1. Through your study, you contribute to a wide variety of literatures, including literatures on innovation, cross-occupational collaboration, and conflict. However, near the end of your findings section, you note several alternative explanations which didn’t make it into the final narrative. What difficulties, if any, did you face in framing the paper? Were there any competing contributions that didn’t make it into the paper? How did you decide on this narrative at the expense of alternative explanations? In short, how did the paper evolve over time?

Your question about how the paper evolved over time is a great one, because the final one in print went through countless drafts! For sure, framing was a challenge. Initially we framed the paper in the literatures on incumbent firm adaptation and interorganizational power struggles. In this first draft, we also motivated the paper around an empirical puzzle, which was why did the marketers who had credentials aligned with the new environment’s demands (the radical marketers) largely fail at implementing strategic marketing projects. We got several pieces of feedback on this framing. First, the reviewers asked us to motivate the paper with a theoretical puzzle, not an empirical one. Second, they did not like the incumbent firm adaptation literature as we were using it, because our data—which was at the project level—did not really allow us to show firm-level adaptation, which is where that literature is focused. In addition, while we collected secondary source data on the changing environment, our best data was on practices inside the firm, so some reviewers were concerned that we were highlighting the environmental change piece so strongly. Finally, the reviewers asked us to focus on cooperation in addition to competition, and also to look into the professions literature.

This was quite a bit of feedback to get—and that was just a few highlights on framing! In the revised version of the paper, we framed the paper in the cross-occupational collaboration literature. We did this because we saw a clear gap in that literature: while we know a good deal about how cross-occupational collaboration can be accomplished in stable settings, we know relatively little about how cross-occupational collaboration happens in dynamic settings where the traditionally dominant group comes under threat, yet is expected to collaborate with a challenger group that is rising in power. This literature seemed more in line with our data—which was about two occupational groups working on projects together—and also helped us to play up the dynamics of cooperation in addition to competition.

Interestingly enough, the feedback we got on this second draft was that the reviewers liked the move to cross-occupational collaboration, but they wanted some of the original framing back! They believed the focus on cross-occupational collaboration made the paper feel a bit narrow, and could limit the audience for the paper. One reviewer called for a revision that was “70% your current framing but 30% your old framing.” This reviewer believed we had something important to say about the specific processes that enable and inhibit adaptive change in organizations, and we agreed.

With that, we moved to the framing you see in the final paper, which is framed in both the literature on adaptation and inertia in technology-based organizations and the literature on cross-occupational collaboration. In the spirit of taking the initial feedback seriously, in our newly crafted “adaptation and inertia in technology-based organizations” literature review, we focused our citations not on studies about how organizations adapt to environmental change, but on the internal dynamics associated with technology development when a dominant group comes under threat from a challenger group. So while we did in a sense “go back” somewhat to an earlier framing, we were able to use the reviewers helpful suggestions to make it much more tailored to our specific study.

Question 2. The tension between occupational and organizational values is prominent in your study. Interestingly, the majority of marketers who have worked more than three years at Transco are moderates (60%), whereas the majority engineers who have worked more than three years at Transco are radicals (83%). This difference is partly explained by engineers being the dominant group at Transco in conjunction with having high occupational credentials, in contrast to low formal training of marketers. Thus, engineers have greater mobility possibilities outside of Transco. How do you understand this lack of organizational loyalty (putting organizational values below their own occupation) on the part of the engineers who began working with Transco in its early years?

Yes, the tension between occupational and organizational values was quite prominent at Transco—and something that struck us from early on in our data collection. Generally speaking, the moderates in both occupational groups were more committed to the organization than to their occupation, and the radicals in both occupational groups were more committed to their occupation than to the organization. Most of the radical marketers were highly credentialed and were hired in with the intent of making Transco a more sophisticated marketing organization. These marketers came to the organization because of the marketing challenge, not because they wanted to work in carsharing. Therefore, it’s little surprise that these marketers were more allegiant to their occupation. The moderate marketers, in contrast, generally had no formal marketing training and had joined Transco years before because of their desire to work in the carsharing industry, and at a company that emphasized things like sustainability. They had been working in an engineering-dominated organization for years, and took pride in the organization they had helped to build. No surprise, then, that they privileged organizational over occupational values.

The picture was a bit more complicated on the engineering side, because there, the longer tenured organization members tended to prioritize occupational over organizational values. The radical engineers came to Transco at a time when carsharing was totally new (at least in the US), and they were attracted to the company first and foremost for the technical challenges carsharing presented. They hailed from other technology companies. They wanted to work on innovative technology development projects—not marketing initiatives. We should add that these engineers were indeed very dedicated to Transco as an organization, and to its associated values like sustainability. They were also very proud of the organization they had built. For this reason, they largely remained at the company even as it was shifting to a marketing focus. Given the excellent job market for engineers at the time of the study, this in itself is a testament to how much these engineers cared about the organization and its values. However, as one of these engineers said, “I’m an engineer first and foremost.” In contrast, the moderate engineers had less tenure in the organization, and they generally had not played a role in developing Transco’s foundational technology. They tended to come to Transco from companies where engineering was not the highest power department (e.g. a bank), and so they were more comfortable working at a marketing company, and placing organizational values over occupational ones.

Question 3. In the paper, you label certain employees “radical” and “moderate.” Did employees remain consistent in their identity over time (radicals remain radical, moderates remain moderate) or did they change (radicals turn moderate and vice versa)? How did this affect your data analysis?

They remained consistent in their identity over the course of the study. This is explained by our answer to the question above: the members of these four different subgroups had very different backgrounds and histories with the company. Even in T2, when the power of all engineers, moderate and radical, was decreased, the radical engineers still had much more power than the moderate engineers because they were truly irreplaceable (because they had built and had in-depth knowledge of Transco’s custom technology systems, and this could not be hired for). This may be why we did not see radical engineers becoming more moderate in T2; if fact, they stepped up their resistance towards marketers in T2, in an effort to prevent Transco from becoming a marketing company.

An interesting question is why the radical marketers did not eventually become more moderate, and adopt the approach of the moderate marketers in interacting with engineers, as this had been shown to be quite effective (with moderate engineers at least). This too comes back to the issue of the different backgrounds of the different subgroups. The radical marketers were strongly committed to marketing occupational goals and they were unwilling to engage in actions that would have threatened their marketing knowledge, understandings of authority relations, and values. In addition, they had top-tier marketing credentials which afforded them attractive opportunities at other companies, and this likely made them less willing to compromise.

Question 4. The use of the radical flank effect is a powerful mechanism for explaining the collaboration of dominant and challenger occupational groups in the second phase of the study (T2). Out of curiosity, how did you become familiar with this concept? Was it in the back of your mind during your fieldwork or did you integrate it during the coding phase? In the case of the paper, the radical flank effect was productive for the organization; under which conditions would a strong threat from a challenger group be detrimental to the organization?

Like so much in inductive field research, this was a bit serendipitous! As part of the comprehensive exam process at Sloan, students need to write a series of essays, one of which is focused on the research project done in the first two years of the program, in Emily’s case, the Transco project. At the time, we had an initial draft of the paper where we were not drawing on social movement theory, but were telling a more general story of political tactics used by the marketers and engineers to press for or resist change and how these tactics had changed over time. One of Emily’s exam questions (drafted by Kate, Lotte Bailyn, and John Van Maanen) was to retell the Transco story using social movement concepts. Emily did so, drawing on social movements concepts like movements, counter-movements, framing, and collective identity. In this exam essay, she noted that within the marketing movement and engineering counter-movement, there was heterogeneity, with some members of each department being more zealous than others. These terms didn’t make it into the final paper—with good reason—but she referred to an “inner core” and “outer core” within each department. She pointed out that it was the “outer core” of each group that was able to work together. During the oral exam, when we discussed the essay, Kate noted that this dynamic reminded her of the radical flank effect.

At that point, we read social movement scholarship on the radical flank effect. This was very helpful because it helped us to interpret some of what we were seeing. For example, we knew that some engineers and some marketers had become better at collaborating over time, but we had not theorized about why. The radical flank effect highlights the importance of a critical event that weakens the power of the defenders of the status quo, and leads moderate defenders to work with those pushing reform. This created an “ah ha” moment for us because we had long been struck by the large impact Transco’s release of its Q2 results had had on the company. We had seen that these negative results were used by radical marketers to persuade the CEO to give them a public mandate for more dramatic change at Transco, and to signal to engineers that they were going to step up their efforts to constrain engineers. But, we hadn’t theorized about this or made it a central point in our paper. The interpretive theory of the radical flank effect helped us understand the why behind patterns in our data. It also helped give us some more compelling language to describe what we were seeing (e.g. “radical” and “moderate”).

Your other question is about when a strong threat from challenger group might be detrimental to an organization. This is an important issue to raise, and we believe it really depends on the organizational context. At Transco, thriving in the new environment—the more mature carsharing industry—required collaboration between the challenger and defender subunits. We would imagine that the radical flank effect would be less useful—and indeed may be harmful—in cases where a moderate solution would be counterproductive to the success of the company. For example, Tripsas’ (2009) study of a digital photography company demonstrated that, while the company was a leader in much of the earlier digital technology, it engaged in a moderate response to environmental change and offered reform-oriented hybrid solutions rather than revolutionary solutions. This slowed down its execution of a digital strategy. This issue that the radical flank effect does not necessarily result in a successful solution is an important one, and social movement theorists writing about the radical flank effect have received this very critique. In our future research section, we suggest that it would be useful to explore the contexts in which moderate solutions accomplished by the radical flank effect can be productive versus counterproductive to the success of the organization.

Question 5. Do you have any advice for students doing ethnographic studies while taking classes?

It is definitely a challenge to do an ethnographic study while taking classes—or really, at any stage of a doctoral program!—because the data collection is so intense. Our advice is to take advantage of any consolidated stretches of time you have, and to spend this time at your research site(s), especially in the early days of data collection. Emily spent the spring semester of her first year securing access to Transco, and started collecting data there in June, right after classes ended. She was able to go in nearly full time during June, July and August, which was very important. First, in the early days of conducting a study, when people at the site are just getting to know you and you them, it is very helpful to put in face time and to allow yourself the time to figure out what is really going on at the site. Second, this allowed Emily and Kate to get a very good grasp of the situation at Transco before Emily returned to classes in September. This was helpful because at that point, with the study’s focus more honed, it was possible to be more selective about things like which meetings to attend. Of course, it doesn’t always work for people to spend a full three months collecting data uninterrupted, but the broader point is, use any longer chunks of time you can fine (e.g. January break, even spring break).

Three other pieces of advice to add. First, especially for your initial research project, it can be very helpful to select a site locally, if that is possible. Your first study is going to be overwhelming no matter what. If you add travel into that mix, it can make your life tougher. For example, in this paper, attending daily scrum meetings where engineers and marketers interacted was very important. Had the research site been hours away, observing these meetings regularly would have been impossible. Second, write biweekly memos about your evolving findings and meet regularly to talk about them. These memos helped Emily and Kate to step back from Emily’s day-to-day fieldwork on a regular basis and push their theoretical thinking. Finally, share what you are learning with others—fellow students and faculty too—as you are collecting data! Their comments will help guide your data collection. Emily and Kate both took part in groups of qualitative researchers who convened periodically to share their work and exchange tips. This proved invaluable for us.

Thanks for your interest in our findings and for the opportunity to answer these interesting and insightful questions!