In 2012, Harvard Business Review cited data scientist as the sexiest job of the 21st century. Just two months ago LinkedIn shared the “25 Hottest Skills that Got People Hired in 2014” – guess what type of workers possessed these skills? This attention has been followed with a slew of articles telling budding analysts the skills they’ll need to get to the top of the data scientist food chain.
In 2012, Harvard Business Review cited data scientist as the sexiest job of the 21st century. Just two months ago LinkedIn shared the “25 Hottest Skills that Got People Hired in 2014” – guess what type of workers possessed these skills? This attention has been followed with a slew of articles telling budding analysts the skills they’ll need to get to the top of the data scientist food chain. We all know the usual list: strong background in statistics and other maths, programming skills, analytical skills, and so on. But what are the things that make an analyst great?
What’s a Data Scientist Anyway?
From small to large it is becoming standard for companies to employ professional data analysts. The title and placement of this person within the organization varies widely from “business analyst” to “data wizard” but the role is largely the same. These employees analyze the company’s data to go gain new insights into various business processes. But you can possess all the technical skills in the world and still only be a mediocre data scientist.
Here’s our list of the non-technical aspects of being a great one:
Excellent Communication Skills
That amazing new insight you’ve just figured out isn’t worth much if you can’t properly explain it to anyone outside your department. A strong data scientist is someone who can clearly and fluently translate their technical findings to non-technical teams. He or she needs to understand the needs of the business and non-technical colleagues to be able to appropriately wrangle with the data. It also may be surprising to some of you, but some of the best insights will actually come from unexpected sources. Those people in the Sales or the Marketing department are living the data you’re analyzing – it’s not so crazy then that they suddenly ask you to merge to sources in a way you’ve never thought of before. So communication is twofold: explain well, but also listen well.
Knows THE Data, Not Just About Data
It might not be heartening to some of you to hear this, but someone who is homegrown at a company is going to be better at getting insights from its data. You also need to have a solid understanding of the industry you’re working in, including what problems your company is trying to solve. Yes, sometimes a fresh perspective is good for shaking things up and looking at everything in a new light. But in most cases, someone who knows a company’s data inside and out is going to be the most adept at playing around with it or connecting it in ways that foster creative insights.
Crunch Outside the Box
Be open-minded and think outside of the box. Yes, you hear this everywhere, all the time, and pertaining to almost anything. In data science it’s important because it can set you apart. What we mean by this is don’t stay inside the “box” of your company’s own data. Your playground should be external datasets rather than internal databases and sources. Learn to use these to your advantage and develop an intellectual curiosity to uncover “truths” that will form your solution.
Knows When to Stick to the Basics
We talk a lot about creativity and finding new approaches, but at the end of the day leading with the standard and most basic KPI’s of a company is a winning strategy. Good data scientists know to head for the core metrics when starting a project, or trying to figure out an issue. For example, before you head to the more inventive calculations or comparisons, remember that you’ll most likely to help Marketing figure out that their campaign failed by simply connecting CRM data to Adwords data.
Don’t Reinvent the Wheel
Countless hours have been wasted on data projects at many an organization only to then find out that someone has already done everything before you. Oftentimes it’s quite easy to find solutions from people in your own organization or online that will solve your problems in a few minutes. For example, there are numerous places online and in print to find production codes or simple data manipulation tips and tricks. Don’t be silly, search before you start.
Have a Process in Place
Once you do start a project, it’s important to have a workflow that you’ll stick to every time. This might differ from person to person, but in general there should be a set way of approaching data projects. The key point that we think are fundamental to this process for any data scientist is to present and discuss your questions and research along the way as often as you can because feedback is essential to finding creative insights. As we said before, it’s important to be a good communicator, and this is part of the reason why.
