What I’ve Learned About Managing Failure in the Lab

Apr 4, 2019, 10:00:00 AM / by Alex A. Goldberg, Ph.D.

Laboratory  in ChaosAnyone who’s ever worked in a research lab knows that failure is as inevitable as death or taxes. Imagine having to perform a finicky technique from scratch, without previous experience  or the help of anyone well-versed in it either. After the hundredth time—I swear it took a lot more than that to finally get it right—the stress can become difficult to manage and solutions start to feel like a far-off fantasy. Fortunately, there are many things you can do to manage failure in the lab and come back stronger. 

Experimental failure

This is the most frequent type of failure in lab-based research. I have yet to meet the scientist who has evolved beyond mere human limitations and performs every experiment right the first time he or she tries it. A simple fact of research is that, at some point, your experiment will fail on you. There’s a reason consistency is an issue in research and that’s because, with so many steps, reagents, and experimental environments, it’s nearly impossible to optimize a method on your first try or to be so consistent that you never mess up an assay, even slightly (like adding the wrong growth media to your cells before treatment, boy did that cause a stir). 

Often, repeating experiments that previously worked may cause an even greater distress, as you know you did everything right and still might have gotten a weird result. For those following in the footsteps of others from your lab, repeating their results may mean a fruitless search for information in scribbled-on lab books and notes, with an out-of-serve email—who deletes their email account?!—as your only way to contact the previous scientist.

Solutions: Experience and consistency are keys to overcoming these types of failures. Experience is important, because you need to learn from your mistakes before you can consistently apply the technique correctly again and again. First, you’ll want to rewrite the protocol you’ve used as clearly and as detailed as possible, probably as a flowchart, and assess where you think the problem is. If you haven’t already, you’ll want to pore over everything there is to know about the assay in the literature that might address the weak spot(s), and when this inevitably does not reveal why your antibody isn’t working, you can request the help of others who have performed the same technique and can help you troubleshoot. If the error is simply mixing up samples or making crucial errors in preparing reagents, referring to standard operating procedures (SOPs) as guides and using appropriate labels will help you avoid this. Oh, and if you’re not sure about your buffers, just prepare them fresh each time, if possible. There’s little point in using materials you’re not 100% sure about. Once you’ve found a way to get the protocol to work (hopefully), your next task should be to standardize it as much as possible. Start by writing a detailed SOP, so that you have a source for the exact conditions of the experiment, including the storage conditions of the reagents and the controls used. Use labels and barcodes wherever possible to track your samples and to help reduce the chance of unwanted errors. For labs that use a LIMS, ensure your protocol is uploaded into the system, making managing the workflow and analysis more efficient and less labor intensive. It also allows you to better concentrate on the experimental design of the assay.

Disproving your hypothesis

Have you got the proper controls, made sure the samples were all processed appropriately, and performed the experiment three or more times with each attempt giving you the same unexpected result? Then it’s probably time to revisit your original hypothesis.

Solutions: While this kind of “failure” may feel deflating at first, it is the best kind to have. Acceptance that your null hypothesis is true may give your research new direction and meaning. There has even been a growing movement to publish negative results for several reasons:1

  • They often challenge scientific dogmas that may not be 100% accurate
  • They may help interpret positive results
  • They’re necessary to obtain a full, uncensored scientific picture
  • They help others adjust their own experimental methods and hypotheses

If you’re faced with a situation where the entire basis of your project appears to be invalid, it may be harder to spin the situation around, but it’s best to avoid the pressure to force the issue and publish biased or outright false results. You can always look for creative solutions and concentrate on new avenues of research with the resources that are left. So, if your results aren’t what you thought they were, don’t fret. There’s a reason you saw what you did and it’s simply a matter of using your curiosity and experience to figure out why that is and where to go next.

Failed infrastructure  

Many things can be included in lab infrastructure: administration, resources, lab equipment, and lab environment. The most common problem is not having enough infrastructure to properly address your research questions. This is where scientists usually have the least amount of control, as factors dictated by outside agents will restrict the scope of your studies and make it difficult to press on.

Solutions: This is the hardest type of failure to overcome, as many decisions are made before you even step foot in the lab (I definitely wasn’t consulted before we bought our useless sonicator that set me back four months). Collaboration is a great way to bypass the limitations of your work environment, as other scientists might be interested in sharing resources, such as antibodies or equipment, making it easier for everyone to expand the scope of their research. Networking comes in handy here, not just for your future job prospects, but also for the ability to form new partnerships within your own institution and abroad. One great way to connect with other like-minded scientists is to attend lectures and product shows, commonly held within research institutes. Another option is with social media websites, like Twitter, Facebook, LinkedIn, Reddit, and ResearchGate, while taking part in conferences will help you find collaborators within your specific field of research.

Failure to communicate

This is probably the easiest type of failure to manage. Research requires communication on a global level; from going to international conferences and publishing in journals to working alongside colleagues and scheduling times for using equipment, communication is necessary to succeed as a scientist. Breakdowns come in many forms, such as failure to properly communicate the fine points of a protocol, requiring specific equipment at a given moment and not being able to use it, and not clearly addressing your hypotheses in presentations and article submissions. Each breakdown in communication makes it harder to obtain and convey results.

Solutions: As a scientist, you need to be precise, clear, and honest with the person you’re conveying information to, whether it’s a graduate student in the lab or the audience you’re targeting with your publications. Unfortunately, working day-to-day alone at your bench on an individual project isn’t the best setting to refine your communication skills. Here are a few things we, as scientists, can do to improve them:

  • Go outside your comfort zone and practice presenting your work at lab meetings and small, local conferences to fine-tune the message you want others to understand.
  • Practice your written language skills, either by writing small abstracts as you work or by applying for grants and fellowships. This is enormously helpful when submitting papers to major scientific journals, as they’re more likely to take your work seriously if the message is as clear and understandable as possible.
  • On a more personal level, getting to know the people in your lab means healthier workplace relationships as well as more emotional and intellectual support. It also helps with workflow in the lab, as lab members who communicate effectively can coordinate their tasks efficiently too, particularly when it comes to using the same reagents and instrumentation.
  • Teach! Teaching and mentoring others is a great way to recapitulate important facts, theories, and methodologies and to figure out ways to express them so that even those who don’t work in your field can understand and apply them.
  • Keep a thorough book of SOPs and update it regularly. This avoids lapses in communication when disseminating information regarding a laboratory technique, as everything you need to know about the procedure should be written out as clearly and as detailed as possible before even beginning an experiment.  

Working in the lab can be as stressful a job as any workplace. There’s enormous pressure to develop new targets, therapies, and interesting data, to publish in high-impact journals, and to obtain the funding to keep going, either through corporate means or academic grants. All scientists go through tough times; it may seem like research is nothing but one letdown after another and that nothing will ever get resolved. However, research is largely built upon scientists learning from failure. If you’ve never failed before, you most certainly wouldn’t develop the creativity to apply elsewhere in your work or the perseverance to endure projects that take years to complete. Just remember, there’s always a solution to your problem—it’s just a matter of being patient and figuring out what it is.


  1. Weintraub PG. The Importance of Publishing Negative Results. J Insect Sci. 2016;16(1):1-2.

Subscribe to the Blog


GA International is a leading manufacturer of high-performance specialty labels and a supplier of identification solutions used in research and medical labs as well as healthcare institutions.

Topics: Science, Laboratory