Wednesday, March 21, 2018

Project Management for Academia 101: learn the basics


Project Management (PM) is a full-fledged discipline which focuses on the planning, execution, and completion of projects. In large corporations, projects like the launch of a new product can take a year and involve dozens or hundreds of people with different expertise in different departments (R&D, marketing, sales, production, distribution). The Project Manager's role is to plan and coordinate the project so that it is successfully completed by a defined date.

I started looking at these practices just through conversations with friends in other industries and quickly realized that many academics are already expert project managers and other academics would really benefit by learning some of the PM rules. I have always thought that your PhD is supposed to teach you how to conduct and complete a project, your postdoc is to identify and plan a project that will get you a job while you learn how to write grants and manage a few people, and your faculty position steps everything up to next level where you manage projects, people, budgets, and several other responsibilities. PM is a critical skill to develop early as it will sustain you throughout your career, and is readily translatable to any other industry.

A good place to start is the Harvard Business School Project Management Manual. It's 20 years old, but it's short and simple and a good introduction to basic practices. The major steps for managing a project are: Initiating, Planning, Monitoring, Executing and Close Out.

Initiating involves defining and organizing the project and it’s one of the most critical aspects of PM. If you don't know what the project is about, what is needed, and who is doing what, you are setting yourself up for chaos from the very beginning. You need to define objectives and the expertise needed to achieve them, including bringing people on board and giving them clear instructions on what their responsibilities are.

Then you Plan: set intermediate objectives, deadlines, and deliverables, assign specific tasks to people and define how the different parts of the process fit together, for example, whether assignments are sequential or parallel and which steps depend on completion of previous assignments. My guest blogger Duc Phan will expand on this in the next post.

Only after you define objectives and set out a plan, you start Monitoring how the different moving parts proceed and get into the Execution/Management portion. Depending on the size of the project this is where things can get hairy due to planning errors, unforeseen obstacles, or personnel issues. This is where the manager may step in to troubleshoot or find someone else with the expertise to do it, reallocate resources, adjust deadlines, etc, until "Tada!" the project is complete and the final deliverable is produced. As part of the Close-Out process, it will be important to go through the flow of the project and see where things went wrong and whether a different design is needed for the future.

I'm sure you can see how these principles could be applied in the lab, but I will give you a concrete example. You can think of projects in terms of published articles, and this type of project management is a good primer for both trainees and faculty. Managing a large project like the one outlined in an R01 application will include multiple publications.

GOAL: I want to publish a paper!
Great! How?

Initiating:
* What is the paper about? Define the title of your project/paper, the main hypothesis that you want to prove.
* What was known about this before? Read the literature to decide what are important questions left to answer and where the knowledge gaps lie.
* What do you need to do? What are the necessary steps to prove your hypothesis and what are the techniques/approaches involved? Are there techniques/reagents you need to acquire?
* How many people do you have and how many do you need? Do you need collaborators to bring different expertise? Do you need to recruit a few junior trainees to take on data collection/analysis?
* Do all the steps need to progress in an IF/THEN fashion (each step dependent on the previous) or are there parallel steps where multiple techniques can be used at the same time (e.g. biochemistry and histology on the same tissue to answer parallel questions)?
* What if my hypothesis is wrong? Develop a risk management plan: alternative approaches and when to abandon the project.

Planning:
* Lay out the steps you have identified and outline experiments.
* Assign experiments (to yourself or different people) and determine how long it would take to complete them.
* Factor experimental failure, broken equipment, illness, and other factors that could delay the process.
* Prioritize experiments and organize them based on time available, personnel, reagent generation and troubleshooting, animal breeding, equipment scheduling and other factors intrinsic to your research design.
* Define clearly who is responsible for what, who reports to whom, and how participants are supposed to communicate.
* Set deadlines for each component of the project and make everyone aware of them. This is critical when hand-offs are needed, e.g. animals to be shipped to a collaborator, summer student leaving, etc...
* Remember that most likely the manuscript will go to several rounds of review and that you also need to factor time for revisions. If you plan to graduate or leave the lab, who is going to finish it?

Note: Some people like to lay out figure plans when starting to develop a paper: they organize the outline of each figure and which data is needed to complete it. I sometimes find this difficult as the story can change a bit as you go along. I use a hybrid system where I write out the questions we are asking (e.g. Does loss of gene X alter protein levels of Y and Z?) followed by bulleted experimental outline (e.g. 1. Generate 5 animals per genotype/sex, 2. Prepare protein lysates, 3. Run Western blot for Y and Z). Then when we have enough data, we print things out and start organizing figures at the white board, then find the holes and plug them. Which brings me to...

Monitoring/Executing:
* If the plan is laid out well, the tracking becomes easier during regular staff meetings, as you can just go through the experimental outline and see what is done and what is not done.
* There are ways to monitor without meeting such as Asana which I have discussed in a previous post or the team messaging app Slack and its ToDo option. I will further discuss these strategies in a post dedicated to getting lab members on invested into projects.
* The simpler and more annoying part of managing the project will be dealing with troubleshooting and unforeseen delays leading to changes in deadlines.
* The hardest part of managing a scientific project will be reframing and readjusting your goals as you go along. Scientific discovery is not like launching a new iPhone. As your results come in, you may find your hypothesis is not true or you may find your beloved biological mechanism does something amazing you had not anticipated. The thing is that once you have this PM loop ingrained in your day to day planning, you can shift very quickly and use the same framework I outlined to redesign your workflow and continue based on new goals. Instead of single loop, doing research is more like going through these steps over and over again in a spiral.
* Overall the hardest part of managing in general is managing people, which is particularly challenging with smart and strong-willed scientists, but as I said above, this merits an entire post.

Close out:
* Once you have a collected all the data, you set up a mini-project to write the manuscript and submit, while you plan for revisions do to on the holes that you willingly left for the reviewers to find. 
* The reviewers find a few other holes and you plan for revisions. The manuscript goes back in and is accepted!! 
* Remember to celebrate and thank all the contributors for their valuable contribution. 
* The proofs of course come back while you're on vacation without internet access (I don't know how journals know that, but they do), so remember to plan for a back-up proofreader as your final Close-Out item. 
* Review your process, adjust, and repeat roughly 100-200 times throughout your career.

Now, think of how else you could apply these strategies!

1 comment: