The world’s leading publication for data science, AI, and ML professionals.

Simple Solutions to Make You a More Productive Data Scientist

Or a more productive anything for that matter.

Photo by Dan Barrett on Unsplash
Photo by Dan Barrett on Unsplash

There is no better time like a new year to tackle one of the biggest problems data scientists can have: productivity.

Data scientists have a lot on their plate, with demanding job descriptions and a constant need to self-educate. It’s no wonder that data scientists can feel burnt out. However, there is a way to see the light at the end of the data (or tunnel if you prefer).

By implementing some productivity solutions and tools, data scientists can work towards alleviating the crush of work that never seems to stop. You’ve probably seen some of these tips before, but sometimes it doesn’t hurt to be reminded of everything you have at your disposal to help you become more productive. Why not take the new year to experiment and see what works for you?


How to be more productive with work demands.

Automate the tedious tasks.

Data scientists spend up to 70% of their time on low-level tasks such as collecting, cleaning, and organizing data.

When you think about the 80/20 rule which dictates that 80% of your results come from 20% of your efforts, spending this much time on low-level tasks doesn’t seem that efficient. Furthermore, wasn’t there once a famous saying that suggested that you should work smart instead of working hard?

Many of the processes involved in collecting, cleaning, and organizing data are extremely repetitive and not a good usage of your time. Therefore, instead of wasting time on routine tasks, make the code work for you, and automate the monotonous stuff.

Check out these articles on automating Data Science tasks to get you started:

The Lazy Mindset of Effective Data Scientists: How Automation Can Help

Automation in Data Science

If you’re working remotely, take advantage of scheduling your workday around when you’re most productive.

Depending on the company you’re working for, this may be an impossible solution. However, if you’re working for a more free-flowing company or yourself as a freelancer, why not schedule your working time around when you’re most productive?

With the work-from-home movement looking to take hold after the pandemic is over, many data scientists will be allowed to work from wherever they want. Not only that, but the working hours become more flexible as it becomes acceptable to email your coworkers at 3 AM.

So, if you’re most productive in the morning, plan your workday around the morning hours. Or, if you’re a night owl, work late at night. You may be surprised about how your Productivity and overall happiness and satisfaction with life improves in tandem.

Use the IDE that works for you, not the one that most people say makes you "more productive".

Just because an IDE makes one data scientist more productive, doesn’t mean it will make you productive.

First and foremost, when choosing an IDE, pick the one that you’re most comfortable with. You won’t be productive if you’re fiddling with endless little add-ons or features you’re unfamiliar with.

Luckily, you have several IDEs to choose from if you haven’t already picked your one and only. Spyder, Pycharm IDE, Jupyter Notebooks, R Studio, and Visual Studio Code are just a few of the IDEs to choose from.

From personal experience, Visual Studio Code and Jupyter Notebooks have been my favorite IDEs to use, due to their clean layout, easy-to-use productivity-enhancing plugins, and small requirement from my computer in terms of size.

Film yourself working.

Kind of a weird suggestion, I know. But hear me out.

There is something oddly satisfying about watching time-lapses of people working or coding. Just take a look on Youtube and you’ll find thousands of videos of people doing exactly that.

Not only do you get a cool video at the end of the day, but you also become motivated to work during the time that you’re filming. Furthermore, you can also see where you’re spending most of your time. If you find that you’re spending three-quarters of your time coding features, but think that more of that time would be better spent selecting and tuning your models, you may adjust in the future to better your productivity.

Make task-batching work for you.

Data science is one of those disciplines where you’re expected to do lots of different tasks perfectly. However, it’s difficult to jump around from task to task that each requires a completely different skillset or type of work. This also applies to the mental shift necessary when working on different projects.

The solution for this is to make task-batching work for you. Task-batching is a productivity theme where you set aside time where you’re working on the same type of tasks and only those types of tasks.

For example, if you plan out your workday using task batching, your schedule may look similar to this:

  • 8 AM-9 AM – Checking emails.
  • 9 AM-11:30 AM – Data cleaning for Project X.
  • 11:30 AM-12:30 PM – Lunch
  • 12:30 PM-2:30 PM – Meetings about Project X.
  • 2:30 PM-5 PM- Running preliminary analyses and developing models for data for Project X.

These tasks are not only batched by type, but also by focus. For instance, you’re likely working on multiple projects at a given time. For some people, it can be difficult to switch from one project to another throughout the day. To make it easy on yourself (and if this works for your work environment/process), try focusing your entire day on a given project to avoid having to do mental gymnastics to switch your thinking to a different project midway through the day. From there, you can batch tasks specific to the project in such a way that you focus all of your attention on completing a given task instead of trying to multi-task and not get anything done.

Use productivity tools like Trello to track your progress on projects.

Productivity tools like Trello were lifesavers when I was doing my capstone project. Not only did it allow multiple team members to update their progress, but it also gave you a simple way to visualize where your team was at in the process.

Simple productivity tools that allow you to track your progress on projects are not given the credit they deserve for creating a productive environment. Whether your preferred method of tracking progress is a whiteboard, a journal, or a productivity tool, tracking your progress is a surefire way to meet deadlines and reduce stress.

Depending on your style and project requirements, you may have to switch up your productivity tool. Here are some of my favorites and other tried and true methods of tracking productivity that you can implement for your next project:


How to be more productive with self-education.

Self-education is one of those necessary evils that can leave you feeling burnt out and unmotivated when it comes time to do the work you’re actually being paid for. With data science continually evolving as a discipline, self-education is one of the unspoken parts of the job contract that employers expect you to do. The problem is that after the workday is done, little energy is left to work on personal projects or to finish a class on new technologies. To improve your self-education productivity, make the most of the time you have to learn, and to even give yourself some extra time just for yourself, try out these productivity tips and solutions.

Create a personalized learning curriculum to learn a new skill or technology.

While many concepts can be learned with a single Youtube video, others require a bit more effort. For example, if your job requires you to learn how to create a particular machine learning model, it would be hard to learn everything you need in a single video. This may require something a bit more in-depth, such as a learning curriculum.

Creating your own personalized learning curriculum is not only a great way to stay motivated in the self-education process, but is also an ideal way to learn difficult concepts without having to pay an exorbitant amount of money for an online class.

For this example, start by making a list of all the concepts you need to know to make a machine learning model. Then, add in any extra concepts that are perhaps more specific to the exact type of machine learning model you need to create for work. From there, you can scour the web for free courses, source documentation, blog posts, and video tutorials that can fulfill each of the concepts you need to learn. Some of my favorite sources for online learning content are freeCodeCamp, Coursera, Udacity, edX, Youtube, and the TowardsDataScience publication right here on Medium.

Not only will you be more inclined to see the curriculum through after you’ve put in the time to personalize it, but the lessons are already laid out for you in such a way that only a small amount of time per day is required before you have a deep understanding of how to create a machine learning model that can then be implemented at work.

Break learning concepts into smaller pieces that can be studied in less than an hour.

Not everyone has eight hours in the day where they can delve into learning a new concept or Technology. Few people even have one hour they can devote to learning. Because of this, your productivity can suffer if you feel like you need to learn something new but have no time to learn it.

Instead, think of learning as something done in smaller time chunks. Everyone has twenty or thirty minutes that they can use to learn. By breaking down the learning into smaller sections, it will be easier for you to develop a learning habit that only takes a small portion of your day.

Furthermore, it can be intimidating to think of learning an entire concept, but when it’s broken down into logical pieces that only take a small amount of time to accomplish, the task can seem much more manageable.

By breaking down your learning concepts, not only will you create an easy-to-maintain habit, but you’ll also be surprised by how fast you can learn something by simply applying yourself for a few minutes every day.

Leave thoughts of work behind while trying to learn.

Work is one of the most distracting things you can think about when trying to learn a new concept.

Between worrying about if you completed everything you needed to, or whether you’re on track with that big project your boss trusted you with, or if the email you sent to your colleague was slightly offensive because you realized too late that a joke may not have been the way to go, or if the office gossip is true, too much time that could be better spent learning will be spent wasted as your mind wanders towards what you’re actually being paid to do.

It’s only natural that when you’re sitting in front of a boring lecture on differential calculus that your mind goes on autopilot and wanders.

Therefore, to improve productivity while learning, you need to turn everything else off and focus solely on learning.

This could mean waking up at five in the morning just to learn when the world is silent and your workday hasn’t started yet. Or it could mean meditating before learning to slow your mind and make you more centered and focused on the task at hand. Or you could put on those big noise-canceling headphones to drown out any distractions.

Whatever the trick, make sure it keeps you focused on learning instead of letting you check your email every five minutes. You’ll be surprised how much you can learn when you limit your focus to just what’s in front of you.

Become aware of when you are about to burn out.

Sometimes, life can be a lot. Between work, family, personal needs, and any number of other commitments you’ve made, it can seem like one more priority would be the thing to push you over the edge.

To use some overused, but highly relevant analogies: your plate can only be so full, and you can only pour from a full glass.

In other words, if the thought of going through one more lesson on a new technology makes you feel like you want to quit your job, sell everything, and move to a remote cabin in the off-grid wilderness where you don’t touch technology for the next forty years, you need to stop. Stop forcing yourself to learn something new when it’s obvious that your brain needs a break.

Working as a data scientist is mentally taxing enough, and sometimes your brain needs a breather after an eight hour day of crunching numbers, cleaning data, and debugging code.

Listen to your brain once in a while, and give yourself a break from learning. Unless your boss is a cyborg, they’ve likely experienced the same thing too. Everyone in tech has at some point or another. Give yourself enough time away from learning that you get to the point where you’re excited to learn again. It will come back, trust me. And when that excitement for learning returns, you’ll be even more productive than before.


Final thoughts.

With everything that comes with the data scientist title, it can be easy to be left feeling overwhelmed. However, if there is one thing that fosters productivity it is setting yourself up to be productive. This sounds kind of obvious, but more often than not, life happens, and your productivity solutions fly out the window to compensate for anything that comes up out of the blue.

While you’ve likely already been aware of the solutions I mentioned in this article, it never hurts to have a gentle reminder of the other solutions that may further stimulate your productivity. Because at the end of the day, data scientists need all of the productivity they can get.


Related Articles