Justin M. Berg – Stanford Graduate School of Business
Chen Zhang – University of Michigan Ross School of Business
Laura Sonday – University of Michigan Ross School of Business
Question 1. One central focus of your paper is how the roles of creators and managers influence creative forecasting. The separation of creator and manager roles seems to be something that people could easily take for granted in organizations and would not naturally problematize. From an idea generation perspective, how did you come to pay attention to the separation of creator and manager roles and sense that there might be interesting issues and paradoxes to investigate here?
My work on creative forecasting started with an interest in explaining why so many great ideas are rejected in organizations. I call these errors “false negatives” in the paper. My focus on the distinction between creators and managers was first inspired by Elsbach and Kramer’s (2003) study of Hollywood pitch meetings, in which screenwriters (creators) pitch their movie and TV show ideas to executives (managers). I’m a big fan of this paper. As Elsbach and Kramer point out, Hollywood executives often commit false negatives, passing on scripts that prove to be mega-hits when someone else takes a chance on them. For example, the scripts for Star Wars, Titanic, The Truman Show, and Seinfeld were all rejected multiple times.
The Elsbach and Kramer piece made salient the key distinction between creator and manager roles that I ended up highlighting in my paper: creators focus on generating and evaluating their own ideas, while managers focus on evaluating others’ ideas. I found it fascinating that Hollywood managers develop fairly elaborate criteria for forecasting the success of writers and their scripts, but that no one really knows if their criteria are actually useful for accurately predicting success. They may be rejecting a bunch of great ideas without ever knowing it, especially if no one else picks up the scripts they reject.
Reflecting on this helped me realize that if your job is to evaluate ideas but not generate them, the criteria you use to evaluate ideas may become too rigid and idiosyncratic, leading you to undervalue novel ideas. In this way, specializing in idea evaluation may backfire and make managers worse at evaluating ideas. This led me to wonder if creators would be better than managers at forecasting the success of novel ideas thanks to creators’ focus on generating ideas, which could help them stay open-minded and flexible about new ideas. Given that many organizations and industries separate idea generation and evaluation into creator and manager roles, I thought this was an important topic to explore.
Question 2. You chose a fascinating context for the field study to explore your research questions. What was the process like for you in terms of looking for the context of this study and collaborating with industry insiders? What are some lessons and suggestions regarding looking for field research contexts and collaborating with practitioners that you could share with doctoral students?
As I was considering different contexts to study creative forecasting, I met James Tanabe and Lena Gutschank, two veterans of the circus arts industry who became my collaborators. Before I met them, I had done some theorizing on the effects of creator and manager roles in creative forecasting and had thought a bit about possible study designs, but I was not committed to any particular set of hypotheses or study design. This proved to be serendipitous, as it enabled James and Lena to co-construct the theory and study design with me. If I entered our collaboration with a fully formed theory and study design, I think their expertise would have been underutilized. This experience taught me that when starting a new project, it might be ideal to engage with practitioner collaborators early on, before your thinking is too cemented to learn from their expertise.
Question 3. The boundary conditions you proposed about creators’ forecasting accuracy (e.g. the ideas coming from creators themselves versus others, and creators’ past success with bad ideas) are really interesting. How did you first come to think about these boundary conditions? Did you get any inspiration for these hypotheses from observing actual phenomena?
My circus industry collaborators, James and Lena, were pivotal in opening my eyes to the boundary conditions related to the past success and quality of creators’ own ideas. James and Lena both had several years of experience watching online videos of circus and other performing arts. As we were planning our study, they mentioned their frustration that videos with the highest-quality ideas were not always rewarded with the most success in terms of views. Our conversations about these themes inspired the hypothesis that achieving success with a low-quality idea undermines one’s forecasting accuracy going forward. This hypothesis is one of my favorite aspects of the paper. I think it serves as an important reminder about the dangers of ignoring the role of luck in our success.
Question 4. In describing your studies, you explained how having hybrids–people in roles with both creator and manager duties–as a comparison group could help to shed light on the theoretical distinction between creators’ and managers’ roles. Hybrids were present in the empirical investigation of both studies, but they were not part of the hypotheses. Were you surprised by any of the findings related to the hybrid role? What were some considerations you had when deciding on the empirical and conceptual roles of hybrids in the paper?
In the circus study (Study 1), we expected that most of the participants would be in either creator or manager roles, but we did expect some participants to be in hybrid roles that combined creator and manager duties. Because we ended up having a good number of hybrids (n = 42), we figured it made sense to include them in the analyses.
To be honest, I wasn’t sure what to expect from the hybrid results in Study 1. In Study 2 (the lab experiment), hybrids experienced the creator and manager manipulations one at a time and the amount of time spent on each was controlled. But in Study 1, the circus hybrids were mostly former creators who had been promoted to a managerial position but continued to be a creator as well. Thus, it was difficult to know how much their role emphasized idea generation vs. evaluation.
It turned out that hybrids scored in between creators and managers on both forms of forecasting accuracy. This was consistent with the proposed theory since hybrids presumably do less idea generation than pure creators but more than pure managers. Although I didn’t predict these hybrid results specifically, I was pleasantly surprised that they were able to provide additional support for the proposed theory.
Question 5. Selecting certain ideas to pursue based on forecasting their success is something important and relevant in our own academic work too. Do you see any implications of this research for creative forecasting in our scholarly work?
In general, I think the theory and results support the efficacy of the peer-review system we use in academia. However, the theory would predict that the best peer reviewers are those who are currently engaged in their own research. If someone is doing a lot of reviewing but no research, they may think more like a manager than a creator.
It’s tempting to think of intellectual taste as a trait, but this work suggests it may be better to think of it as a state. Good taste in ideas may be more fragile than we realize. If we stop generating our own novel ideas, we may also stop seeing potential in our peers’ ideas.
Elsbach, K. D., and R. M. Kramer
2003 “Assessing creativity in Hollywood pitch meetings: Evidence for a dual-process model of creativity judgments.” Academy of Management Journal, 46: 283-301.