The question of how to reliably recruit brilliant students came up in a project management workshop I recently attended. Evidently I budget substantially more money and time for working with undergrads because I’ve got an ‘army’ of ~20 now. It’s working well. But I’m sure many of you are puzzled over how and why I manage this many.

Part of the HYPER lab team in the summer of 2019.

Quickly the question of ‘why’ bother managing ~20 undergrads working in the lab:

  1. There’s no better way of ensuring you have the right person for a job than years of experience working with them.
  2. There’s no better way of preparing someone to work in the lab than years of experience working in the lab.
  3. I don’t have a crystal ball for predicting when I’ll need a particular type of person for a particular job.
  4. I don’t have a crystal ball for knowing what skills or interest a person has for a particular job.
  5. There’s no more reliable contact at a company than someone who’s worked in the lab, even as an undergrad.
  6. There’s no more reliable draw for companies to come to the lab than new recruits, even undergrads.

That all sounds fine. But many of my colleagues are saying to themselves, “What a pain. Too much stress. It’s either quality or quantity and you can’t have both.” etc. Enter the question of ‘how’ to manage ~20 undergrads working in the lab:

  1. Start small, with great people. Train them well and place them well. Word will spread.
  2. Once you have critical mass — 3-5 funded grad students — it’s time to use the group to it’s fullest potential. Several years ago I adopted the Jigsaw classroom technique to try and solve the puzzle of lab management. In short, Jigsaw divides a large group into many similar teams with a set of 3-8 defined roles on each team. This allows each team to follow a uniform procedure for different aspects of a project, and you can have each of the roles meet as a group to ensure transference of information between groups. I infuse Jigsaw with Lean/Agile techniques. Each grad student becomes a product manager — they know what is required by the grant and have a general idea for how to do it. Recruit a complimentary undergraduate team with the roles of Leader/Liaison, Communicator/Reporter, Production/Builder, Analytics/Theory, and so on. Pretend you’re Gene Roddenberry and these are roles in Star Trek (e.g. Kirk, Ohura, Scotty, Spok, etc). Your grad student will run the team by maintaining a backlog of tasks that must be done (sorted easy, medium, and hard). Weekly sprints will culminate in a report out and identifying the task lists for the next week.
  3. Now that you’ve multiplied the effort to 9-25 people, build a pipeline to sustain. You have quantity, you need to ensure quality while sustaining the culture that your lab is known for. The easiest place to do this is at the front end. I have a ‘general’ team that I manage comprised of new lab members that I personally interview. The general team functions like any of the other teams but is tasked with continuously improving the general equipment, facilities, and procedures that the rest of the teams use. This ensures that all new lab members learn the general capabilities of the lab, understand the work that goes into making those systems great, and how to improve them. If someone graduates or moves off of one of the grad student product teams, we have a pool of individuals on the general team ready to immediately step up.
  4. Now standardize the process as much as possible via procedures. We have everyone sign a Memorandum of Understanding (MOU) at the start of their work: HYPER Lab Summer MOU 2019. We also have everyone develop a work specification for their time to ensure that they have made an individual contribution communicable on a resume: HYPER Lab Work Spec 2019Did you know that we’ve even developed procedures for developing procedures? Last but not least — a procedure for offboarding.

That’s pretty much all there is to it. Once you get the teams going you just have to get out of the way, although you check in once and awhile. And much like any other puzzle, once you’ve done it, you know how easy it is to do again, and again, and again.