The Soft Skills for Data Science


The Soft Skills for Data Science

Machine Learning algorithms can be classified into:

Mr. X just spent years becoming an expert programming in R, Python and Java, and now everyone telling him that he is not hire-able because he don’t have good enough soft skills?

Why should it matter what he is like a person if he can analyse Big Data, crunch numbers at record speed, plot beautifully complex graphs, and transform statistics into predictions?

Well, it matters

Even if you can do all those wonderful things, none of it means anything unless you can communicate it.

In all honesty, no-one will hire you in business if you haven’t considered improving your soft skills.

While it may seem harsh and unfair that thousands of smart, dedicated and budding new data scientists are being left behind, it makes total sense when you look at how they are just unable to communicate their world to the rest of us.

“What are soft skills anyway?”

Good question. The term is annoyingly vague enough to simultaneously make you think you know what it means, while yet evasive enough for you to still need a “Soft Skills for Dummies in Data Science” handbook.

Here are the best tips we here at SuperDataScience can give for both new data scientists (and for a large portion of experienced ones who maybe slipped under the radar) keen to build their softer side.

And, for anyone interested in developing their soft skills with us, be sure to check out our careers page as we’re hiring! Here is the link you’ll need:

1. Empathise with Your Non-Data-Science Audience

It sounds patronising, yes, but too often data scientists get carried away when given their newest project.

Imagination runs wild with elaborate ways to present the data in the coolest, most complicated and impressive manner, getting so lost in excitement that they completely miss the point of what they’ve actually been asked.

Empathy is key in this pointer.

If you, the data scientist, were able to put yourself in the shoes of your corporate-minded stakeholders, you’d see the problem from their perspective and work with a more human approach to the situation.

You’d start to understand why you’re being asked to manipulate the data and what it means for your company; it’s not just an opportunity for you to prove you are an excellent programmer (they already know that, else you wouldn’t have the job).

Try to recognise when you’re viewing the world like Arnie in the Terminator.

Remember that you’re dealing with people who act and think very unlike you.

More often than not, your audience isn’t going to be other data scientists who will drool at the newest features of Python you’re showcasing. It’ll be leaders, analysts, consultants, and sales specialists wanting to understand how to improve efficiency in their company, or predict trends for economic growth.

Most people who will look at your work won’t understand what “clustering” is or what on earth a “fuzzy algorithm” does.

It isn’t your job to confuse them with all the complicated things you can do: you need to learn to be your best at the “hard” side of data science while bridging the gap between yourself and your non-data science peers.

More often than not, your audience isn’t going to be other data scientists who will drool at the newest features of Python you’re showcasing.

Take-away: Put yourself in the shoes of your audience when presenting data and people will start listening to you.

2. Get Familiar with the World of Business

If you’re entering a business industry and find yourself collaborating across various fields of expertise, you’ll likely feel out of the loop when your colleagues start talking in their fancy, exclusive business lingo (scalability, core competency, leverage…).

You need to break the cycle of gormlessly nodding along and hoping it’ll all become clear what they meant once you have the data in front of you.

The best way you can overcome being left out is to ask, ask again, and keep asking.

Sure, you’re not a business specialist and you didn’t study data science for years only to find yourself having to care about product marketing. But, you are working in a business and, like it or not, that makes you part of the corporate club.

Start bridging the communication gap between science and business.

Get going by doing some research within your company and understand their core values, what drives them and who they are working for.

Yes, you don’t have to be passionate about women’s clothes to work on an online shopping algorithm, but it helps to have read about what online shoppers want to see when they use your software, how it helps them and why you’re doing it in the first place.

Don’t shy away from being part of a corporation: embrace it and add your new business acumen to your growing list of skills.

Believe me, you’ll go much further dedicating time to business skills in data science than if you choose to only ever work on your programming skills.

Take-away: If you’re working in business, make yourself interested in it and clued up on how it works, bridge the gap, and stop alienating yourself.

3. Make Your Work Beautiful and Read More Literature

Scientific writing has its own beauty in regards to its concise, matter-of-fact register. Most likely, you practised writing like this throughout University and found it a comfortable way to present your work to your examiner.

This style needs to stay where it belongs: in research and academia. It has no place in business.

What does have a place in business is:
- universally understood terminology,
- summaries of key points,
- trends and predictions,
- easy to understand graphs.

Essentially, “dumb it down” a bit

That is not to say give it 40% and patronise your audience as if they were toddlers.

But when you can make a graph look bigger, prettier, and clearer instead of more complicated and with multiple axes at once — do it.

If it means you need to add another slide for clarity and constantly have keys for common acronyms like NLP, DL… (“Natural Language Processing”, “Deep Learning” which seems so fastidious, right?) — bite your lip and do it.

An author of an historical novel does not expect their reader to have researched the era they choose to write in.

Instead, they give the necessary context and allow the reader’s eye to flow across the page effortlessly — give your readers’ eyes a break and make your work accessible without removing the core of your message.

Reading more literature will help you grow your lexicon and your storytelling skills will blossom.

Believe me, no one is impressed when you annotate your graphs with the equations you used or the fancy algorithms you wrote.

Just like no-one is interested in reading the author’s messy first draft — we want clarity and ease of use.

No one wants to see this kind of graph.

Reading more literature will help you grow your lexicon and your storytelling skills will blossom. Maybe you’re not interested in Jane Austin or Tennessee Williams, that’s fine.

Find some other form of written art (poetry, prose, lyrics) and study the beauty of how the writer is able to capture audiences.

Use these storytelling tools in your own work when presenting findings — lessen the monotony of data and incorporate the excitement of your findings, the sentiment and the emotion of its potential

Take-away: Study how to write to capture readers’ eyes, beautify your scientific work for your audience and resist the urge to show off your hard skills over your newfound soft ones.

4. Uncover Your Creativity

It’s likely if you’re reading this that you are more adept at problem solving and analytical thinking than flourishing with creativity. If that’s the case: do not give up with trying to be creative.

It is not acceptable to just be good at programming.

It’s true that creativity is not just something you’re either born with or not

Sure, some people emerge from the womb already playing the guitar or singing with perfect pitch, but most of us had to practise until we got it.

If you are terrified at the sight of a blank canvas and used to dread poetry exams — you’re not alone. But you should stop using it as an excuse to shy away from creativity.

It is not acceptable to just be good at programming. There are many interesting ways to work on discovering your hidden talents, and believe me, you’re so creative deep down, you just haven’t found it yet.