The challenge of building a research group: unspoken keys to successful leadership
As researchers, we train, then train some more, then train further to become better at designing, performing and interpreting experiments. Many of us will be exposed to scientific writing and poster or oral presentations throughout our PhD thesis or postdoctoral training. If our research output in the different phases of our careers is good and we are genuine and original (a matter for another essay), we might get access to a group leader position. And then, suddenly, everything changes. All at once, our success does not depend on our hands, on our experimental skills or on the number of experiments that we can multiplex. In the blink of an eye our achievements depend on experiments done by others. And the challenge begins.
I look back in time to the moment when I started as a group leader – Sept 1st 2010, a 30 year-old youngster – and it marvels me that I made it this far with the toolkit I brought in my backpack. My experimental skills are now barely relevant, and all that matters is my capacity to connect the dots (meaning building hypotheses) and, most of all, leading others to design, perform and interpret the experiments that our research projects require. This is obviously an oversimplification, but bear with me. Leading a lab is about creating a list of values that will drive your research team (“the lab philosophy”), a structured but personalized plan for each team member to achieve their aims in the lab (“the expectations dialogue”), and a supervision strategy that is adapted to the dimensions of your group (“the social network”). I will develop these three aspects from the perspective of a group leader, but I would like to stress that we should define, nurture and refine them from early stages in our career, because we begin to lead others much earlier than the moment we become group leaders. I have excellent leaders in my research group (some of them are PhD students, postdocs and technicians) that apply these principles on a daily basis.
1. The lab philosophy
When I started my lab, I began by building the team based on repulsion forces. I identified what I did not want my lab to be, based on what I had seen and experienced up to then. I did not want my lab to be based on inter-individual competition; I did not want to have papers as the aim of the research, but rather as the consequence of well-developed research projects; I did not want researchers in the lab to be mercenaries of science, with a poor personal–professional balance. With time, this set of principles based on “what I do not want”, has evolved into a Decalogue on “what the values of the lab are”. These values are conveyed in every candidate interview and in my yearly overview (one-to-one meetings) with every lab member. They provide me with the clarity and transparency I need when I set my expectations about working dynamics in the lab. I would have preferred that someone would have mentored me about the importance of taking the time to establish the lab Decalogue when I started my lab. I would have felt that I was constructing my team based on a blueprint that I had previously meditated on and actively decided. I recommend everyone to begin writing this list of values early on, so that they can structure their leadership strategy.
2. The expectations dialogue
I believe that most of us have a clear idea of what we expect from our co-workers when we are part of a team. Similarly, I frequently know what I expect from the people I recruit. The key is to communicate it. Often, a poorly productive work relationship in research is the consequence of defective sharing of expectations. I learned the relevance of these aspects the hard way, and I have done my best to implement a system that ensures that the definition and revision of expectations is part of the supervision process. Every researcher in the lab is presented at least twice a year with a prototypical time-dependent expectation plan for each profile (PhD students, Postdocs, Research Assistants, Technicians, PIs). In addition, a mutual expectation review is done once a year in the one-to-one meetings. This structured strategy allows everyone in the lab to have a clear idea of what I expect from them, and, accordingly, what I must deliver in terms of training. Importantly, the expectations go both ways, and it is very important to ensure that we, PIs, understand or reframe the expectations of our trainees, for a healthy and productive work relationship. Setting mutual expectations is something that anyone can and should implement with people they supervise (or that anyone should request to their supervisors), and it is of tremendous value in the training process.
3. The social network
I am reading with interest the book “Sapiens” from Yuval Noah Harari. It fascinated me how he framed the principles of population size in communities, specifically how large tribes above a certain size depend on myths and beliefs that hold the group together. Certainly, these aspects do not necessarily apply to research groups, but it made me think about the determinants and working modes associated with lab size. I believe that points one and two in this essay are required for lab tribes of any size. However, when the size of a lab increases beyond a critical point (which I would set at six to eight people), there are two alternative working modes I have identified that could be implemented.
The first is the simplest (or the one that requires less energy from the group leader) and is based on the Darwinian process of natural selection. In this mode, each lab member is responsible for his/her own "survival". The group leader only needs a handful of successful projects to move the lab forward, and the rest can be assumed as unsuccessful projects, each involving a researcher that might leave the lab empty handed. This strategy is appropriate for labs with a high level of resources, with an abundance of postdoctoral researchers and with tremendous capacity to attract candidates who are highly independent to undertake all the stages of a research project from beginning to end.
The second mode is more cumbersome, since it requires knitting a network of lab members per project, and assigning leadership roles within these teams that must be recognized in the output of the research. This “social network” model is the one I have come to implement, and the one that gives me more headaches and satisfactions. I have defined that the size of my lab is the result of a complex formula encompassing: (i) the number of postdoctoral researchers that can lead a project with a high level of independence (total number of projects), (ii) the number of senior researchers that are ready to co-supervise a PhD student with me (total number of PhD students), (iii) the set of technologies and projects that require technical support (total number of technicians), and (iv) the number of researchers with whom I have established a plan of path to independence (they will account for one or two additional people under their supervision in the course of their stay in the lab). This complex structure also requires a training and recognition plan that needs to be designed and revised periodically. The recognition relates to authorship, thesis co-supervision and opportunities to apply for funding as PI. It is very important to remember that a lab strategy needs to have prototypical roles and expectations but, at the same time, we need to deliver personalized supervision. For example, this implies that there is no set time for a postdoctoral fellow to engage in a thesis co-supervision, and this needs to be tailored to every member´s profile, evolution, and contractual commitment. Some of them might instead supervise Master´s students, as a more concise and time-limited exercise. Similarly, I recently started offering PhD students co-supervision of undergraduates, when I evaluate that they are ready to undertake the challenge. With this complex structure, it is natural that the lab assembles into small functional topic-centric teams. This has been a fascinating phenomenon in the lab. Rather than leading and defining it, I saw myself participating in the creation of these topic-centric teams and shaping them to provide them with body, relevance and functionality. When defining the second working mode, it becomes obvious that it does retain a strong component of individual responsibility, and, in turn, it is not devoid of Darwinian selection pressure. However, in this context, natural selection is not an isolated and stand-alone factor, but rather the culmination of a set of tools and opportunities given to every researcher.
To wrap up, I would like to emphasize that leading a lab goes far beyond lab techniques, experimental design, grant writing, and paper preparation. I would provocatively state that leading a lab is all about operating as a human resources department, so that the time we devote to put the salt and pepper in the research projects is determinant because the rest of the aspects of the lab are solid. And we should not wait until we become group leaders to apply these principles of team-building and lab leadership, since they are a game-changer in every step of the way through our career path.
Top image of post: from the lab of Arkaitz Carracedo in November 2020