4 Costly Data Science Mistakes
January 27, 2020 by Jon O'Keefe

From product development and marketing to staffing decisions, businesses are using data to fuel nearly every decision. This means that data itself is more valuable than just about anything – thus making data science a prominent focus in today’s business world.


With that being said, data scientists must be careful when it comes to best data practices as making a few common mistakes could endanger overall outcomes for your business. Below are some of the most frequent mistakes made by data scientists that could cause a problem:


  • Forgetting the bigger picture – It is important that data scientists always remember to take the time to fully understand the larger scale of the business in which they are working. Understanding the full context of the data is not only one of the top data science best practices, but also crucial to the providing of key business insights. If you do not fully understand how the company works, it is impossible for you to do your job to help it fun better: more than just understanding the data, you must understand what it represents. Without this fundamental understanding, your otherwise perfect data models will still encounter real-world problems that will throw everything off.
  • Ignoring practice – Along with forgetting the bigger picture, data scientists must remember that the practical side of data science is just as fickle as theory. Data scientists must be ready to balance both sides of the job while keeping up with the most relevant technologies in their jobs. There are, after all, certain aspects of a job that can only be learned through experience, and every data scientist must realize this.
  • Overlooking colleagues – Sometimes just asking “why” can help better communications between data scientists and other departments. By understanding the motivation behind a request, you will be able to build a truly useful model that will help other employees solve their problems more easily thanks to your data. This level of open communication helps everyone do their jobs better. Sometimes data scientists can get wrapped up in the technical aspects of their job and forget that this data will need to be presented to another human being in order to help the business grow and improve. While technical skills are incredible valuable in roles such as these, so are more “traditional soft skills” such as communication and presentation skills.
  • Assuming perfection – Many data scientists claim that most of their job consists of cleaning up data, with only a small amount actually consisting of using machine learning to create models and so on. When receiving a data set, you must decide how much you can rely on it and what will have to be done to make it usable. While data is never truly perfect, imperfect data can still be made usable and help you accomplish your goals.


As data science continues to grow in importance and use cases in today’s business world, we can only expect to learn more about how to better enhance these roles through the beauty of mistakes.


Data scientists: What kind of mistake have you made and learned from in your role? Let us know in the comments below.