SQL taught me to ask the data questions. Python is where I learned to shape it — and, honestly, it’s the one I was most intimidated by.

“Learning to code” sounds like a wall when you’re a marketing student. I put it off for a while. Then I hit a problem SQL couldn’t solve on its own: I needed realistic datasets to practise on, and I needed to clean messy exports that arrived with dollar signs, mixed date formats and half-empty columns. So I took Python Fundamentals for Beginners and started from genuine zero — print("hello") zero.

The first week was ugly

My first scripts were held together with hope. I forgot colons. I fought with indentation. I got the same red error message so many times I started reading it instead of panicking — which, it turns out, is the actual skill.

baby_steps.py
# where everyone startsname = "Vrushank"for i in range(3): print(f"attempt {{i}}: still learning")

The loop was the moment it clicked. Once I understood that a computer will happily do the boring thing 10,000 times without complaining, I stopped seeing Python as “coding” and started seeing it as “a very patient intern.”

Then I met pandas

pandas is where Python earned its place in my toolkit. The messy campaign export that used to take me an hour of manual cleaning in Excel became a few lines that ran in a second — and, more importantly, ran the same way every time.

clean_it.py
import pandas as pddf = pd.read_csv("raw_campaigns.csv")# strip the $ and turn text into real numbersdf["spend"] = (df["spend"] .str.replace("$", "").astype(float))df = df.drop_duplicates()

The datasets behind the dashboards in my projects are generated and cleaned with Python exactly like this — correlated, realistic, and reproducible, so the insights are real patterns and not lucky noise.

The trick that finally worked for me: don’t “learn Python.” Pick one annoying, repetitive task you do by hand, and refuse to do it by hand again. You’ll learn the exact Python you need, and you’ll remember it because it saved you.

Where it pays off now

I’m not a software engineer, and I don’t pretend to be. But I can generate a dataset, clean a messy file, and automate the boring parts of an analysis — and for a data-minded marketer, that’s a genuine superpower. If you’re coming from the data side first, start with how I learned SQL — the two together are far stronger than either alone.