A huge part of being a data scientist is quantitative. One needs to be good with numbers, decently proficient at coding, and be able to analyze charts and patterns. That’s the reason why most data scientists often come from technical backgrounds like mathematics and statistics, engineering and computer science.
With that said, after spending the last few months working as a data scientist, I realized there is more than simply just being good with numbers — it is the non-technical skills that truly make someone stand out from the rest of the crowd.
Sure, having a strong technical foundation is crucial, but having the softer skills will not only make you a more well-rounded individual but more importantly provide the extra edge to help you finally land that next data science job or be set up for a promotion.
Although developing these skills may not be as straightforward as signing up for an online programming course, there are still actionable steps and mental models that you can adopt in order to improve in this particular area.
In this blog post, I want to highlight 5 non-technical skills that you should pay attention to as well as outline the practical steps that you can take to develop these skills.
1. Attention to detail
Funny enough, attention to detail was one of those phrases that I used to put on my resume just because it sounded cool but in reality, I didn’t actually know what it fully meant — until today.
Well, attention to detail can take a variety of different forms but at its core, it is ensuring that a piece of analysis or deliverable that you are working on is professional, accurate, and communicating a message the right way.
Specifically, in data science, this could be the due diligence of checking for data errors and performing sanity checks. For example, checking for missing fields, non-sensible data like customers with an age of 500 as well as ensuring the percentages of exhaustive categories add up to 100%.
Besides data checks, when creating a slide deck or designing a dashboard, it is also important to pay attention to alignment, colour consistency, spelling mistakes, choice of visualization, user experience, the list goes on and on.
And while it might sound boring to you, having excellent attention to detail not only establishes your credibility as an analyst through the quality of work that you deliver but more importantly, it prevents running into embarrassing situations with a client. The last thing you want is for a client to pick up on a huge mistake before you do.
With that, I will close this section with a piece of advice I received recently regarding the importance of attention to detail.
80% of a slide deck can be done in 20% of the time, but the remaining 20% will take 80% of the time.
This is of course the infamous Pareto principle, which you can read more about here.
I wrote in a blog post a couple of months ago regarding communication skills. Here, I want to expand upon that idea as well as reiterate some of the points that were mentioned in that post.
Having good communication skills is not unique to just data science but quite frankly, any other professional work environment. Communication in the workplace can be broken down into two types: verbal communication and written communication.
Without diving into too much detail, the main differences between the two are the speed of transmission and proof of record.
Verbal communication has a high speed of transmission but no proof of record. As a result, it is more often used in internal stand-ups or client meetings where you are simply providing an update to the team or need immediate feedback on a particular idea.
Written communication, on the other hand, has a low speed of transmission but offers proof of record. This can be in the form of emails, Slack messages or even comments in your code.
Having good communication skills can go a long way when working in a collaborative environment. Not only does it make or break a team’s efficiency but it also helps with persuading others to pursue an idea during a project.
Personally, I struggled with this early on in my career as I was so used to working on my own data science project without having to work with anyone else. But knowing what I know now, I would highly encourage anyone who is serious about working as a data scientist to start practising these skills early.
What does this look like practically? Well, this could be building a habit of writing comments when you are coding, writing a summary report or mock email after performing an exploratory analysis to summarise your findings, starting a blog that simplifies complex data science topics, taking up public speaking lessons and so on.
It is the small details like these that will add up over time to make you a more effective communicator and data scientist.
3. Empathy towards clients
One of the important things that I was taught when working as a student consultant during my university days was to think from the client’s perspective.
While some of you might not aspire to work in consulting in the future, I personally think this mindset applies regardless of whether you are working as a data scientist in consulting or pure product analytics.
The ability to step into a client’s shoes allows you to think deeply about the pain points that they are experiencing and how data can be used to solve their problems. This, in turn, can be used to formulate an approach to tackling a certain problem but also to drive your analysis and present them in such a way that adds the most value.
For example, working with government clients now has challenged me to learn more about policy decisions like I never had before, as well as issues that concern the government such as healthcare, education, economic growth and so on.
As a result, when sourcing data and performing analysis, I have become more deliberate in my approach to tracking and communicating the key metrics that will bring the most impact to my client.
In summary, before doing any piece of work, think about what the client will get out of it, and subsequently focus on that and avoid doing mindless, low impact tasks.
4. Prioritization and delegation
Much like your time at university where you are juggling between multiple assignments and exams, as a data scientist, you are often required to split your time into different projects and certainly different tasks within a project.
So, how does one manage this? The answer is prioritization and delegation.
Prioritization is the process of assigning relative importance or urgency to a given set of tasks. It is having the awareness of when tasks are due as well as their respective time requirements in order to complete them.
Practically speaking, this could range from having an up-to-date calendar, writing a list of daily goals before starting the day, or even speaking to your manager to get his or her opinion on what you should focus your time on.
On a side note, it is also important to know when to say no. As a graduate data scientist, I think it is easy to fall into the trap of taking on too much in order to somehow prove your worth to everyone else in the company. However, doing so could cause tremendous stress but also jeopardize the overall progress of the team if you don’t end up delivering on what you initially promised.
Therefore, assess what you can take on based on your current capacity and prioritization before committing to a task.
Next up, we have delegation. While as an analyst straight out of university, this is probably not the most pressing thing to worry about, but it is still worthwhile talking about it.
As your career progresses and as you climb the corporate ladder, your time will slowly shift towards people and project management more than actually being the one who is doing the technical work itself.
Delegation is more relevant for a seasoned data scientist or project manager who is responsible for the health and progress of a particular project. It is one of the keys to being a good leader.
Delegation can be a complex topic that I personally haven’t had much experience with, so I will tread lightly here. But in essence, effective delegation involves an understanding of your team’s strengths and weaknesses, providing a healthy level of support and feedback as well as establishing a clear desired outcome.
Delegation done right can yield incredible results and make a project team function like a well-oiled machine.
5. Open-mindedness and constant strive for knowledge
With the ever-changing landscape of data science, new tools and technologies are popping out like popcorns. Software, programming languages, modelling techniques are constantly evolving to serve different purposes — and it can be quite challenging to keep up.
However, to be a good data scientist is to have a curious mind and an active passion for learning. It is important to stay connected to the most recent industry trends as well as consistently expand your skillset in order to stay relevant in this field.
Luckily for us, there is no shortage of resources on the internet. YouTube, DataCamp, MOOCs like Coursera, you name it. It all comes down to whether or not you are willing to put in the time and effort to actually do it.
I like to think of open-mindedness as embracing the mindset of a student who is always chasing after knowledge regardless of where you are in your career. This will not only make you irreplaceable in your workplace, hence ensuring career longevity but more importantly, it elevates the value and competence that you can bring to the table.
My personal anecdote was when I was tasked to build a PowerBI dashboard which I had zero experience in at the time, and truthfully also zero interest in. Despite that, however, I realised the value that I can bring to the project team as well as long-term career benefits by upskilling in that particular area.
As a result, I took the initiative to reach out to more experienced colleagues for guidance in addition to following tutorial videos over the weekend to learn what I needed to know.
Looking back now, it has been such a rewarding journey to learn something completely new and it was only made possible because I had the open-mindedness to begin with and the desire to learn.
To close, the reality is that we are not always going to know everything and that is okay, but instead, it is all about having the mentality to strive for improvement that matters in the end.
Jason Chong Associate Data Analyst at Quantium | BCom (Actuarial Studies) |