Initiating, planning, and running collaborations

Collaborations are everywhere in research. They are the Number One way to import methods and expertise into your research group, and to export your own knowledge to help others go new ways.. They also give your team members additional training opportunities. So today’s post is all about initiating, planning, maintaining, and quitting collaborations – both for PIs and their group members.

together we can

This post is part of the open draft for the Research Group Leader Book
[about] [read more]

In this post, I address collaborations both from the view of the PI, and from the view of a scientist in training. One of my PhDs asked me to write about this topic, and it’s definitely one of the topics that should be discussed and taught during graduate training.

Being strategic about collaborations

I first collaborated at the very beginning of my PhD; of course, this collaboration had been initiated by my supervisor. Then, nothing followed for a while. When the first opportunities presented themselves toward the end of my PhD, I was thrilled and jumped at them. But I found that with my first papers out, and with becoming known in my field through conferences and talks, there were soon more opportunities than I could possibly serve.

At this point, senior scientists told me that I should be strategic about my collaborations. So, while it might be take what you get at a very early stage, opportunities for collaborations abound once you’re at the stage of Postdoc or PI. And then, each collaboration you agree to means saying no to others.

The three most important questions to ask about any potential collaborations are:

  • What am I looking for, and what is my gain? Some aspects are access to data acquisition and anaylsis methods; expanding into a new topic; getting into a new social network. You might have others.
  • What can I offer the collaborator; what is his gain? One important thing you can always invest is your time and effort.
  • Therefore, is this collaboration worth investing my and my team’s time? And, what other work, projects, or collaborations will I have to sacrifice to be able to handle this one?

Responding to invitations to collaborate

…if the PI is asked

As PI, inquiries can go three ways:

First, the collaboration might involve you directly, such as writing a review paper together, writing a collaborative grant together, or providing expertise only you as PI own in the lab. In this case, you have to answer the above questions about the value of the collaboration; probably more importantly, you have to decide whether you have the time to take part.

Second, the collaboration might rather involve someone in the lab, for instance, an expert for some fancy method. Of course, the decision whether to start this collaboration should involve that person. But step one should be for the PI to decide whether it strategically fits the lab. Step two is to discuss with the lab member whether the collaboration will advance their career. The worst way (though this happens quite a bit in real life) is to inform the lab member that, by the way, there is this new project they will soon be working on in addition to all the other stuff they are already doing.

Third, a potential collaborator might ask to work with or in your group. Although the first response to this is usually that as long as that person comes with their own money and do the work, all is fine, any such collaboration does require your supervision, and, at the latest when the resulting paper must be written, considerable time. If the student wants to join your lab, it is also important to think about whether s/he will fit into the group, and how much your lab members will need to be involved in the project. Accordingly, don’t forget to ask them.

…if a lab member is asked

Lab members are probably most often asked to contribute their expertise. This means, they will have to invest considerable time. Let your team members know whether you are open to them discussing collaborations, and at what point you want to be involved. On the one hand, they are on your pay roll. On the other hand, they are on a career path. The further they are in their career, the less they will be willing for you to impose decisions on them. To take conflict by its horns, it is best to discuss openly how much freedom they would like, and how much you are willing to give.

Your team members are likely less experienced with collaborating than you are as a PI. So share your experience with them, and help them get their priorities straight for effective career planning.

…in any case

It’s a risk to enter a collaboration with someone you do not know. I’ve read that some people advise to collaborate only with people you know well. I personally don’t think this is realistic if you want to build a relatively wide network of collaborations. However, it is important to get to know your collaborator as much as possible before giving the final go.
Ask others in your network. Most of the time, someone has information about your prospective collaborator. Furthermore, trust your gut during the initial discussions. If it doesn’t feel right, consider letting it go. And make sure all important aspects of the collaborations are negotiated up front and agreed upon by everyone.

Initiating collaborations

Initiating a collaboration is probably the easiest part of the whole endeavor.
It usually takes as little as a conversation at a conference, or an email.

Your approach will be most successful if you communicate the following things:

  • What is the specific project / topic / experiment you are proposing?
  • What is your background, your lab, etc.?
  • What time frame are you thinking about?
  • Who will pay for the project (travel, living expenses, experiments), and/or does it involve writing a grant?
  • What is your investment, and what would you like the collaborator to provide?
  • What result are you aiming for? (In most cases, this will be a publication.)

You might not communicate all of these points at once; in a conversation, you can contribute them piece by piece when they fit. If you go via email, some things might be discussed only after initial contact has been made.

Your collaborator will have his own agenda and ideas, and might propose alternatives to your own ideas. Don’t be too quick to say yes to everything. Remember to evaluate the collaboration strategically. Therefore, before you make initial contact, know what you really want out of the proposed collaboration.

Planning the collaboration

Once it’s clear that both sides are generally interested, it’s time to do some real planning:

  • In your own lab, make sure people know who is involved in what way.
  • Specify the project. Decide on the exact experiment or product. Make sure everyone is on the same page.

  • Discuss who will provide funds. If the collaboration involves someone staying in the other’s lab, this includes travel and housing cost. Make sure to think hard about all types of cost that will arise: experimental subjects, materials, publication cost etc. If grant writing is necessary, agree upon who will provide which parts.

  • Discuss a timeline.

  • Discuss authorship of a prospective publication. This can feel very weird, especially when you do it the first time. But it really helps to clear up everyone’s role in the project ahead of time. For instance, the first author will be expected to do pretty much all the work noone else feels responsible for. The last author will probably be the one to provide funds for experimentation. But these things will depend on your field, and on the specific situation.
    Discussion about authorship is especially important if junior members of both collaborating teams are involved: both might want to use this work toward their qualification (e.g. PhD), which is often only possible with a first authorship. Equal contributions are becoming more common, but might not be admissable in all graduate programs. As PI, be sure such administrative aspects are cleared up well in advance to avoid dark hours for your PhD later.
    Discussion about authorship is also relevant for the seniors of the collaborating labs, as last authorship may be important for tenure, success-oriented funding, and salary increases. Here, too, shared authorship is becoming fashionable. Again, it’s smart to clear up the consequences of author order in advance.

  • Distribute tasks. Where will what work be done? Who will program the experiment? Who will do what analysis? Who will write? Will there be student assistants who can support data acquisition? It’s good to know these things early on.

Maintaining the collaboration

Schedule short status checks to show your collaborator that you have the project in view. If the work is done in your collaborator’s lab, ask for status updates if they are not provided. If the work is done in your lab, be proactive: let your collaborator know how it’s advancing regularly. If it is not advancing, let him know too, and explain the reasons.

If your lab members are involved, make sure they represent your lab in the way you expect. For instance, ask them to give you and the collaborator regular updates. Make sure they keep their deadlines, and communicate well in advance when they see that they can’t keep them.

If the collaboration goes well – and this mainly means: if everyone gets a long well, and mutual trust has developed – one collaboration will lead to another.

Handling problems and taking the exit

Not every collaboration goes the way you imagined. There might be personal issues. You might learn that you do not trust your collaborator, or that their work ethic differs from yours. It’s possible that your collaborator does not deliver what he promised, or passes deadlines by months.

Whatever the reason: if you notice early, bring it up in your meetings. Your collaborator might not be aware that he is not meeting your expectations. Or he might think that you are fine with how it’s going as long as you don’t complain. As PI, let your collaborating student know that you will support him if the collaboration becomes difficult. As a junior researcher, claim your supervisor’s support if you don’t get enough.

Yet, sometimes all the talking does not result in the changes you want to see. There can come a point at which the value of the collaboration has become too low for you to maintain it. Or, to say it in more colloquial words, at some point, it’s just annoying without benefits.

At that point, consider ending the collaboration. There might be some reasons that make you feel you can’t, such as if your collaborator is someone really important in the field whom you do not want to cross. But even then, think hard whether you can find a way to exit gracefully. If you can’t come up with a good solution, ask a mentor or a colleague you trust.

The more work has gone into a project, the less we are inclined to let it go. Yet, if the situation in the collaboration is such that success does not seem probable, any additional effort going into the project is too much.

Celebrate your wins

And because I don’t want to end with such depressing things as quitting collaborations, let’s instead end with something happy: when a collaboration ends successfully, don’t forget to thank your collaborators. A phone call or email can be adequate. Or a dinner with all project members at the next conference. Or raising a glass of sparkling wine over skype.

Now, all the best for your collaborations!

I’m sure there are more tips for collaborating, both from the view of the PI and of junior scientists. Please share them below by leaving a comment!


Others have discussed how to quit a collaboration if it doesn’t work anymore.

The book Essentialism: The Disciplined Pursuit of Less is a plea for reducing the number of things we do. Certainly something to consider when planning collaborations. Then again, don’t forget that collaborations build your network like nothing else.

Photo credit: DonkeyHotey / / CC BY


This post is part of the open draft for the Research Group Leader Book [about] [read more].

Leave a comment...