Essential Non-Technical Skills for Thriving as a Freelance Data Scientist

Because your most valuable skills as a freelancer might not be technical at all

This week’s a guest post from Josep Ferrer, a freelance data scientist and technical writer I’ve learned a lot from - he has built a sizeable audience on medium and substack. Today, he shares the underrated skills that make freelance data science work.

---

Big thanks to Claudia for having me on AI Weekender! 👋🏼

I’m Josep, a freelance data scientist and technical writer who (accidentally) built a personal brand and a thriving freelance career without ever setting out to do so. (Curious? Here’s my story.)

Image by Josep

And today… I’m here to talk about freelancing.

Freelancing in data science sounds like a dream, doesn’t it?

You get to choose your projects, set your own hours (and terms!), and skip the endless sprint planning meetings.

One month you’re helping a startup optimize customer retention, the next, you’re prototyping dashboards for a government branch.

The flexibility and variety are genuinely exciting—and they’re what draw many of us in.

But here’s what most people don’t tell you: it’s not the tech skills that make or break you.

Sure, knowing Python and SQL will land you the job, but your ability to communicate clearly, scope projects realistically, and understand what clients actually need is what keeps you in the game (and keeps the invoices paid).

This post is about those unexpected, non-technical skills, every successful freelancer relies on.

Let’s dive in! 👇🏻

#1. Effective Communication

One of the most underrated superpowers in data freelancing? Explaining things simply.

Your client might be a marketing lead, a founder, or an operations manager, smart and capable, but not necessarily fluent in model accuracy, confidence intervals, or the difference between supervised and unsupervised learning.

If you can translate complex ideas into clear, actionable insights, you’re instantly more valuable.

Clear communication doesn’t just avoid confusion, it builds trust, speeds up decisions, and often leads to more repeat work.

And here’s the golden rule:

Use everyday language. If someone hides behind buzzwords or convoluted jargon, it’s usually ego, not clarity.

Simple words get results. Complex ones just slow everyone down.

#2. Project Scoping and Management

Freelancing without solid scoping skills is like sailing without a compass—you’re guaranteed to drift.

Clients don’t always know how long things take, what’s realistic, or where the real complexity lies. That’s your job. A clear scope prevents misunderstandings, helps price your work fairly, and sets healthy boundaries.

A client once assumed “a quick churn model” could be done in a day. After asking a few clarifying questions, I learned they had no labeled data, no tracking IDs, and a very rough idea of what "churn" even meant.

The project turned into a scoping session, then a data-cleaning sprint, then a modeling phase.

Because we scoped honestly, we avoided frustration later on.

What works well:

• Use simple proposals with bullet points and deadlines.

• Define what’s in scope and what’s not.

• Break delivery into phases and share updates regularly.

#3. Understanding What Clients Really Need

Clients rarely come with a “clean brief.” More often, they come with symptoms.

Your job is to ask the right questions and get to the root of the problem. That takes curiosity, empathy, and a willingness to say:

Let’s slow down, what problem are we really solving?

I once had a founder ask for “a recommendation engine like Netflix.” After digging deeper, I realized they were drowning in unused product data and wanted any way to personalize user experience. We pivoted to clustering and a basic rules engine. It wasn’t glamorous, but it made a real impact, and fit their current stage.

Understanding your client’s why is often more valuable than delivering the latest fancy model.

#4. Adaptability & Continuous Learning

In freelancing, no two projects are the same.

You might be wrangling retail data in one engagement and tweaking time-series models for energy forecasts in the next. That variety is exciting, but only if you’re willing to keep learning.

I didn’t plan to work in tourism, but a mid-sized private-public company needed help analyzing mobility data for carbon reduction strategies. I knew the math but had to learn new tools, new metrics, and a new domain in days. That project opened the door to a whole new line of work I now love.

So how do you keep up?

• Block out a few hours a week to read, test, or explore.

• Subscribe to a few high-signal newsletters (like this one).

• Reflect on each project: What did I learn? What should I improve?

Freelancers who stay curious and flexible don’t just survive—they thrive.

Wrapping Up: What Really Keeps You in the Game

In the world of freelance data science, technical skills might get you in the door, but they’re not what keep you there.

The ability to scope a project clearly, communicate without jargon, adapt to new challenges, and genuinely understand your client’s needs?

That’s what builds trust.

That’s what gets you rehired. And honestly, that’s what makes the work more meaningful.

If you’re thinking about freelancing (or are already in the thick of it), take a moment to assess where you stand with these nontechnical skills.

Which ones come naturally to you? Which could use more intention or practice?

The good news is that they’re all learnable.
And just like coding, the more you use them, the stronger they get.

Your takeaways from this article:

• Clear communication doesn’t just avoid confusion, it builds trust, speeds up decisions, and often leads to more repeat work.

• Clear scoping isn’t just project hygiene, it’s how you build trust, avoid burnout, and get paid for what the work is worth.

• Understanding your client’s why is often more valuable than delivering the latest fancy model.

• Freelancers who stay curious and flexible don’t just survive, they thrive.

---