I’ve seen so many fantastic resources about how to start from ground zero and transform yourself into a data scientist. These resources advise to learn Python or R, master data munging and cleaning, understand effective and intentional data visualization, or maybe even to study machine learning.
All of these skills are priceless. Different combinations of them are often required to become an industry-ready data scientist. However, mastering these skills does not mean the battle is over. A data scientist should possess some of the technical skills listed above. But more than that, a data scientist is expected to analyze day to day company activity from a unique perspective. When a data scientist is made aware of an event (a need for a new KPI, an upcoming feature to be released to production, a recent increase in customer drop-off, a new marketing campaign, etc.), they have a duty to ensure that any data generated by that event is accounted for, AND used to its full potential.
For example, can we come up with some creative, domain-specific ways to measure the success of that new feature release? Can we analyze data and isolate a source event or sequence of behaviors that precede the customers leaving a website? Can we reasonably prove that marketing strategy A was better than strategy B?
It is of utmost importance to maintain a data scientist’s mindset. A wonderful way to capture this mindset is through two juxtapositions that exemplify the responsibilities of a data scientist. Determined, yet adaptable. Articulate, yet inquisitive.
Data Scientist Mindset #1: Determined, Yet Adaptable
A constant reminder that I pose to myself is, “If you don’t believe in yourself, who will?” Nobody knows your work better than you do, or believes in your work like you do. Nobody will truly know when your findings are ready to be presented like you will. You were hired on for your specific skills that bring unique value to your company. It is your responsibility to make sure that all of the work you put out is honest and possesses integrity.
Being more determined
When you need to be determined and advocate for your work, you must do so. You can do this in a myriad of ways, but here are a couple examples:
- Fight for that precious meeting time with a senior engineer or manager to get the questions answered that are blocking your progress
- Explain to your colleagues how your project idea would yield insights that could benefit departments X, Y, and Z
- Remind yourself and your superiors that accurate, valuable insights come from time, integrity, and attention to detail
On the flip-side, a person with a data scientist mindset must be willing to take one for the team when it is time. Suppose there is a priority shift in the company. They should be able to switch gears quickly and transfer focus to whatever important task is at hand. A data scientist should always be willing to learn new things – whether that be a new domain (such as civic technology) or a new BI software. A data scientist should be determined. But, they should never let their research keep them from shifting focus when it’s time. Sometimes, it may even be time to drop a project completely, if there are more important priorities, or that a current project is deemed a dead end.
Being more adaptable
A data scientist should always do their best to be adaptable in ways such as:
- Try out new packages, algorithms, etc. that are suggested by colleagues (even if you are loyal to your day one technologies)
- Minimize negative remarks when your manager asks you to drop a project you are extremely excited about
- Constantly approach problems from different perspectives
- Try to place yourself in the shoes of people in different roles in the company
- Embrace incorrect hypotheses and using them to fuel your research further, and not letting them discourage, deter, or distract you
I know that being determined and being adaptable aren’t necessarily exact opposites. But, the essence of these two traits really balance themselves in a way that makes a more well-rounded data scientist.
The determination drives the data scientist to get all worth that they can out of the data at their disposal. It encourages the data scientist that those findings are accurate, valuable, and worth being heard.
The adaptability keeps the data scientist honest. It prevents them from becoming stagnant in their education and skill development. And, (in my opinion) makes them overall more pleasant to work with.
Data Scientist Mindset #2: Articulate, Yet Inquisitive
All techies learn very quickly in the industry that communicating your thoughts, concerns, and new ideas about projects can be extremely difficult… especially when trying to talk to someone who is not in the same field as you are. However, one of the most demanded “soft skills” in data science is effective communication, because data scientists are storytellers. When a data scientist develops a model, he/she must communicate to the stakeholders why that model could positively impact revenue. A data scientist must be able to uncover insights. But, that is not the end of their task. A great finding only gets you so far! They must morph these insights into full-blown stories and actionable items. These stories and items should pull business counterparts in and convince them of the insights’ actual value.
Being more articulate
Being articulate can be a tough task to master, but here are some ways to practice:
- Pay more attention to opportunities to build documentation of database tables, code bases, or processes that would benefit other employees in your company
- Take the time to explain why something works or where the data came from. (Note: this is mutually beneficial! They learn, and you practice your communication!)
- Practice your ability to communicate data concepts to people who are not in the field (family, friends, significant other, etc.)
A very important part of a data scientist’s workflow is being able to explain what you found. But, more important than talking is the ability to ask questions and listen. A data scientist with the proper mindset should always be asking questions to their business counterparts. Although they are not the ones writing the code, the business counterparts are the ones who possess fundamental domain-specific knowledge. This knowledge will help the code gain value.
The business counterparts aren’t the ones analyzing the data. But, they are the ones generating it, or at least have a part in the data being generated. They have indispensable industry knowledge that will help your analyses immensely. Likewise, a data scientist should always be asking questions about his/her own work. What will happen if I have a 0 or NULL there? Am I missing data? Do these numbers make sense? Numbers speak louder than words, so listen!
Being more inqusitve
Here’s a couple good opportunities to be more inquisitive in your work:
- Develop a list of questions for your coworkers before any sprint planning or project meeting, so that you receive as much information as possible before you get to work on a task
- Complete your due diligence of testing before you accept a data source as truth, push a piece of code to production, or present findings to the C-suite
Being articulate and inquisitive are not mutually exclusive. But, the two traits compliment each other in a special way. A data professional with a data scientist mindset becomes a force to be reckoned with.
The articulation gives the data scientist the ability to connect with coworkers and clients from a variety of backgrounds; it also allows them to provide more powerful, relatable explanations and presentations.
The constant inquisition pushes the data scientist to maintain the quality of their work, and to continually pursue excellence.
The field of data science is so beautifully broad. Every day, there are new languages, packages, software, and algorithms being released. There is no sign of that slowing down. With all of these new tools constantly popping up, there’s constant pressure to keep your skills sharp. Forget the stress about which tools to focus on. If you constantly check your data scientist mindset, you’ll be a better data scientist for it. Be determined, yet adaptable. Be articulate, yet inquisitive.