Data Scientist

You find the signal in a company's data and turn it into decisions — sometimes with machine learning, more often with a clear analysis and a persuasive chart. The title got glamorized, then AI-complicated. Here's what the job actually is now.

Median pay (US)

~$112k / yr

Typical range

$75k–$165k+

Degree required?

Often — but proof can substitute

What the job actually is

Data scientists answer hard questions with data: why is churn rising, which customers will convert, what should we charge. The real week is less model-building than the brochure promised — it's SQL, cleaning messy data, talking to stakeholders about what they actually need, and packaging findings so decisions get made. Machine learning appears when it earns its keep; a surprising amount of the value is rigorous analysis plus communication.

Is it actually you?

You'll probably love it if

You genuinely enjoy statistics — not just the idea of them
Ambiguous questions excite you more than clean assignments
You can tell a story with numbers to people who fear numbers
You're patient with messy, incomplete, contradictory data
You like your work changing what the business does

Maybe not, if

You want to build models all day and skip the meetings
Data cleaning feels beneath you (it's half the job)
You need immediate visible results — analyses die in drawers sometimes
Statistics was a class you survived, not a thing you like

The real day-to-day (no hype)

It's 70% plumbing and persuasion. Getting the data, cleaning it, and getting anyone to act on the answer takes most of the time. The modeling montage is a small slice.
AI raised the bar, not removed it. LLMs write the boilerplate code and first-pass analyses now. What's scarce: framing the right question, catching subtle data traps, and judgment about causality. Juniors who only ran notebooks are exposed; thinkers aren't.
Title inflation is everywhere. Some "data scientist" roles are analyst jobs with better branding; some analyst jobs are data science. Read the description — SQL+dashboards vs experiments+models — and price accordingly.
The degree question is real. This field leans credentialed more than most of tech — many roles come from quantitative masters/PhDs. But a portfolio of rigorous public analyses with real conclusions can substitute, especially at startups.

How people break in — or switch in

Three live routes: the credential route (quantitative degree → internships), the analyst route (start as a data analyst, add statistics and Python, grow the title), and the domain route (be the person in marketing/ops/finance who got dangerous with data, then formalize it). For switchers, the analyst route is the honest on-ramp — same skills, lower gate, and a year of real business data beats a certificate portfolio of Titanic datasets.

The strongest data scientists know a domain cold. If you're switching from marketing, health care, or finance, that context is your moat — lead with it.

Your application, already half-written

Here's a question every Data Scientist application asks, answered the way pirch would — in a real voice, grounded in real experience:

“Describe an analysis that changed a decision.”

Leadership wanted to cut our lowest-converting marketing channel. Before we did, I broke conversion out by cohort and found the channel's users converted late — 60+ days — so last-click attribution was starving it of credit. When I rebuilt the numbers with a 90-day window, it was our second-best channel per dollar. We kept it, reallocated from a channel that only looked good, and revenue per acquisition improved 18% the next quarter. The model wasn't fancy; the framing was the work.

pirch's co-pilot writes answers like this for your background and the exact job — try it free →

pirch finds the data roles that are actually you

Tell pirch who you are — your analyses, your domain, your actual level — and it hunts down real, still-open data jobs that fit the whole you, with a tailored cover letter already written.

start your free hunt

first hunt free · we never auto-apply · you stay in control

Common questions

Data scientist vs data analyst — what's the difference?

Analysts answer "what happened" with SQL and dashboards; data scientists add "why and what will happen" with statistics and modeling. In practice the line blurs, and analyst → data scientist is the most common promotion path.

How much do data scientists make?

US median around $112k, ranging $75k early-career to $165k+ senior at tech companies. Quant-heavy finance and big tech pay above that band.

Do I need a master's or PhD?

For research-heavy roles at big companies, often yes. For most product and business data science, a strong portfolio of rigorous analyses plus real domain experience can substitute — especially at startups that need answers, not papers.

Is data science still a good career with AI?

Yes, but it's consolidating. AI automates routine analysis, which raises the value of question-framing, experimental judgment, and communication. The "notebook operator" version of the job is fading; the "decision scientist" version is growing.