Mindset Shift: Transitioning from Academia to Industry

Special thanks to Francesco Mosconi for contributing this post.

 

office-1209640_960_720Transitioning from Academia to Industry can be difficult for a number of reasons.

  1. You are learning a new set of hard skills (data analysis, programming in python, machine learning, map reduce etc.), and you are doing this in a very short time.
  2. You are also learning new soft skills, which also require practice to develop.
  3. A mindset shift needs to occur, and your success in industry will strongly depend on how quickly this happens.

 

Learn to prioritize

When your goal is knowledge, like in Academia, it is okay to spend as much time as you want learning a new concept or completing a project. On the other hand, during this program and on the job, you will often find that there is not enough time to deal with all the tasks and assignments required of you. In a situation where you have more on your plate than you can handle, it is essential to develop the ability to decide which tasks require execution, which can be postponed, and which ones can be simply ignored. There are many frameworks and approaches to prioritization, famous examples including the Getting Things Done system and the Eisenhower Method. Most methods are good, and eventually you will find your favorite one; however, they only work if consistently applied. In other words, it is less important which prioritization method you choose but it is fundamental that you prioritize your day and your week according to the specific goals you are to accomplish.

Questions you can ask yourself to prioritize a task:

  • Is this the most important thing I could be doing right now?
  • What are the negative consequences of not doing this task right now?
  • Is there a way I can complete this task in half the time while still delivering most of its value?

In Academia, you could be tempted to say, “But I would like to learn this perfectly, there are still some things I do not fully understand,” and that’s true. However, there will always be some things you do not fully understand, and you cannot possibly know everything. In Industry it’s often more prudent to be satisfied with delivering the results and move on to working on another task.

 

Adapting to a new “objective function”

Think of your work as trying to optimize for a specific “objective function” in order to reach a goal.

In Academia your objective function might have been something like: “given these resources, try to improve the state of the art of field X by a significant amount before the end of your PhD 5 years from now.” Although you might have had internal deadlines with your workgroup, there were no real consequences for not meeting a deadline. For example, if you were working to develop a new type of wave detector and, by the 3 months deadline, the detector didn’t work, that would simply mean that the paper submission date would be postponed.

Another way of looking at this is to say that you were optimizing for absolute performance score, regardless of the time required to achieve the result. This objective function emphasizes the result (perfectionist approach), and assigns no cost to the time required to achieve it.

The objective function in Industry is very different. Time always has a cost, be it the individual’s salary or the cost of the resources allocated to that project. Besides the implications on prioritizing your work, this also has implications on what you should strive for in order to be successful.

Very rarely you will be working alone on an isolated project. Much more likely you will be working as part of a team, so meeting a deadline is very important to make sure that others are able to complete their work using the results of yours. This also implies that it is better to re-discuss a deadline early on, if conditions have changed. If you agree to deliver a result by a certain date, you have only two options: deliver it, or re-engage the other party and re-negotiate the agreement.

Keep in mind that others will rely on you to be able to correctly estimate the time it will take to complete a task and which tasks you will be able to complete in a given time. As you acquire experience in the field, your ability to correctly assess the time and resources required to complete a task will improve, but when in doubt, it is preferable to under-commit and over-deliver than vice versa.

Finally, it is very important to shift mindset from “deliver perfect results” to “deliver results fast.” Especially in early-stage companies, processes and product will be in constant evolution. In that context, it is preferable to deliver an analysis that shows some evidence with 80% confidence in 1 day, rather than a 99% confident model in 6 months.


Editor’s Note: The Data Incubator is a data science education company.  We offer a free eight-week Fellowship helping candidates with PhDs and masters degrees enter data science careers.  Companies can hire talented data scientists or enroll employees in our data science corporate training.

Related Blog Posts

incubator

Career Enablement at TDI

TDI is more than your typical bootcamp. We provide robust career support to ensure exceptional outcomes for our students. Learn more about our career enablement options here!

Read More »
is it too late to become a data scientist

Is It Too Late for Me to Become a Data Scientist?

You might think data science is a young person’s game. After all, this is a relatively new discipline that might not have been around when you were in school. But research shows nearly half of all data scientists are 40 years and older. Whatever your age, it’s never too late to pursue your dreams of becoming a qualified data scientist. Learn how to succeed in this profession in this blog.

Read More »