A live read on data science jobs right now: how many openings are junior, mid, and senior, what companies are hiring, what salaries they advertise, and what.
The question "are data science jobs drying up?" has been bouncing around career forums for two years. Data science jobs are still out there — but the live listings tell a more complicated story than either the doom posts or the bootcamp brochures admit. The market isn't dead. It's selective. And there's a meaningful difference between those two things if you're deciding whether to spend the next six months retooling your career.
What the listings actually show, when you sit down and count them rather than scroll past them, is a field that has matured past its gold-rush phase. Demand is real. Competition is real. And the entry-level slice is a lot thinner than most career-change advice lets on.
What the Current Data Science Jobs Market Actually Looks Like
The market isn't dead — it's just not throwing doors open
The Bureau of Labor Statistics projects data science roles growing at around 35% through 2032, which is well above average by any measure. But aggregate growth projections and the reality of open listings in any given month are two different things. Growth at the field level doesn't mean every company is hiring, and it definitely doesn't mean they're hiring beginners.
What the current market looks like is a field where demand is concentrated — in tech, finance, healthcare, and retail analytics — and where employers have gotten much clearer about what they want. The casual "data science" posting that could mean anything from Excel dashboards to deep learning research has mostly sorted itself out. Employers now write more specific job descriptions, and that specificity is the tell.
What the sample says once you stop guessing and start counting
To ground this piece in something real rather than vibes, a sample of approximately 200 active U.S. data science postings was pulled from LinkedIn and Indeed in late June 2025, filtered to roles using the search terms "data scientist," "data analyst," and "machine learning engineer." The sample was limited to U.S.-based postings, with remote roles included where labeled as such. Each listing was reviewed for seniority signals (title, years of experience required, educational requirements), skill requirements (hard and soft), work model, and posted or estimated compensation.
That's not a statistically perfect census of every open role. But it's a real snapshot, and it's more useful than recycling a McKinsey trend deck from 2023.
What this looks like in practice
On any given week in June 2025, a search for "data scientist" on LinkedIn returned active postings from companies including JPMorgan Chase, Meta, Walmart Labs, CVS Health, and a range of mid-size SaaS companies. The volume looked healthy at first glance — hundreds of results. The problem became visible on page two: the same roles reposted, contract-to-hire listings dressed up as full-time positions, and a heavy skew toward mid-senior experience requirements.
A hiring manager who screens data science candidates at a mid-size fintech described it this way: "We get 300 applications for a single opening. Maybe 40 are actually qualified for what we wrote. The market isn't dry — the bar just moved, and a lot of people are still applying to the 2019 version of the job."
That's the honest read. Selective, not scarce.
Count the Junior Openings Before You Tell People to Switch Careers
Why the entry-level slice is the part everyone gets wrong
Generic career advice — the kind that says "data science is booming, learn Python and SQL and you're in" — is not lying exactly, but it is describing the field at the wrong altitude. The boom is real at the aggregate level. The entry-level share of data science openings is a different, much smaller number, and that's the number that actually matters if you're a career switcher or a recent graduate.
The false comfort comes from conflating job family growth with beginner accessibility. A field can grow 35% over a decade while simultaneously requiring more experience from new hires than it did five years ago. Both things are true right now.
What the junior, mid-level, and senior split really says
In the June 2025 sample of roughly 200 postings, the seniority breakdown looked approximately like this: around 15% of postings were explicitly junior or entry-level (0–2 years required, or titled "associate" or "junior"), around 45% were mid-level (3–5 years), and around 40% were senior or staff-level (5+ years, or titled "senior," "lead," "principal," or "staff"). Classification was based on the explicit years-of-experience requirement in the listing; when that was absent, title language and responsibility scope were used as proxies.
That 15% entry-level share is the number career switchers need to internalize. It doesn't mean the market is closed. It means the competition for those 15% of data science openings is intense, because every bootcamp graduate and every career changer in the country is also targeting them.
What this looks like in practice
A junior-friendly listing from a healthcare analytics company in the sample asked for "0–2 years of experience with Python or R, familiarity with SQL, and a demonstrated interest in healthcare data." The requirements were modest, and the role was framed around learning under a senior analyst. By contrast, a senior listing from a major retailer required "5+ years of experience building and deploying ML models in production, experience with distributed computing frameworks (Spark, Databricks), and a track record of cross-functional stakeholder influence." Same job board, same week, completely different hiring universe. The word "data scientist" appeared in both titles.
Stop Treating Every Data Science Role Like the Same Job
Data scientist, data analyst, and machine learning engineer are not interchangeable
One of the most consistent mistakes career switchers make is applying to data scientist jobs when their skills and background actually fit a data analyst role — or vice versa. These are not the same job. They share some vocabulary and some tooling, but the day-to-day work, the seniority expectations, and the hiring criteria diverge in ways that the job titles don't always make obvious.
A data analyst is primarily answering business questions with existing data: building dashboards, running queries, producing reports, and communicating findings to non-technical stakeholders. A data scientist is expected to build predictive models, design experiments, and increasingly, deploy those models into production systems. A machine learning engineer sits closer to software engineering: they build and maintain the infrastructure that serves models at scale. The skills overlap, but the emphasis is completely different.
The skills tell you what kind of job it really is
The clearest signal is in the skills section of the listing, not the title. A data analyst posting will emphasize SQL, Tableau or Looker, Excel, and business communication. A data scientist posting will ask for Python, statistical modeling, and increasingly some familiarity with ML frameworks like scikit-learn or PyTorch. An ML engineer posting will add MLOps tooling — MLflow, Kubeflow, Docker, CI/CD pipelines — and will often require software engineering fundamentals that pure data science roles don't.
When a listing says "Python preferred" and lists Tableau as the primary tool, it's a data analyst role with an aspirational title. When it says "experience deploying models to production is required," it's not a beginner data scientist role regardless of what the title says.
What this looks like in practice
Three listings from the June 2025 sample illustrate the divergence. A "Data Analyst" posting at a regional bank listed SQL, Excel, and Power BI as required, with Python listed as a plus. A "Data Scientist" posting at a healthcare SaaS company required Python, scikit-learn, A/B testing experience, and asked for "experience communicating model results to clinical stakeholders." An "ML Engineer" posting at a logistics company required Python, Spark, Docker, and "experience deploying and monitoring models in a cloud environment (AWS or GCP)." Three roles, three different jobs, all appearing in the same search results for anyone who typed "data scientist jobs" into the search bar.
According to O*NET's occupational data, data scientists and data analysts are formally classified as distinct occupations with different task profiles — a distinction that live listings are now reflecting more accurately than they did five years ago.
Follow the Skills Employers Keep Repeating, Not the Ones People Brag About
Python and SQL are the gatekeepers, not the bonus points
In the June 2025 sample, Python appeared in approximately 85% of data science hiring postings and SQL appeared in roughly 80%. These are not differentiators — they are table stakes. The candidate who lists Python and SQL as headline skills on their resume is not standing out; they are clearing the minimum bar to be read at all.
That matters because a lot of career-change advice still treats Python as the thing you learn to break in. It's not. It's the thing you already need to know before the conversation starts. The question employers are actually asking is: how well do you know it, and can you use it on a real dataset without being supervised?
Business judgment keeps showing up for a reason
The skill that genuinely surprised in the sample was how often "business judgment," "stakeholder communication," and "translating analysis into decisions" appeared — not in soft-skills sections, but in the core requirements. A data science hiring manager at a mid-size retail company put it plainly: "I can teach someone a new model. I can't teach them to care whether the business question is actually the right question. That's what I screen for in the first interview."
This is a structural shift from the early years of the field, when being good at modeling was enough. Employers have been burned by data scientists who produce technically impressive work that doesn't influence anything. The listings now reflect that lesson.
What this looks like in practice
From the sample, a representative senior data scientist posting listed requirements in roughly this order: Python (required), SQL (required), experience with ML frameworks (required), "ability to frame ambiguous business problems as analytical questions" (required), Spark (preferred), Kubernetes (preferred). The must-haves were Python, SQL, and business framing. The nice-to-haves were the infrastructure tools. That ordering is consistent across most of the postings that specified required versus preferred. The candidates who spend their prep time on Kubernetes before they can clearly explain a regression model to a non-technical audience are preparing in the wrong order.
Read Remote, Hybrid, and Onsite Signals Like They Actually Matter
Remote is still on the table, but not as the default fantasy
Remote data science jobs exist. In the June 2025 sample, approximately 30% of postings were labeled fully remote, around 40% were hybrid, and the remaining 30% were onsite only. The remote share is real but not dominant — and it skews heavily toward mid and senior roles. Of the junior and entry-level postings in the sample, the remote share was closer to 20%. Employers who are taking a risk on a less-experienced hire generally want them in the building, at least part of the time.
The narrative that data science went fully remote and stayed there is not what the listings show. Hybrid is the plurality model right now, and for career switchers targeting junior roles, onsite or hybrid is the more likely reality.
Location filters are doing more work than people realize
A remote label does not mean the role is genuinely open to anyone in the country. Many "remote" postings include a state or timezone restriction buried in the requirements — "must be located in the continental U.S." or "EST/CST timezone required." Those restrictions meaningfully shrink the applicant pool and, importantly, shrink your competition if you happen to be in the right location.
Onsite and hybrid roles are even more location-dependent. A data science role at a healthcare company in Nashville is not competing with the same applicant pool as a role at a fintech in San Francisco. If you're in a secondary market, that can work in your favor — but only if you're actually applying to roles in your geography rather than exclusively chasing the remote listings everyone else is targeting.
What this looks like in practice
Three postings from the sample show the range: a fully remote senior data scientist role at a Series B startup listed "U.S.-based only, Pacific timezone preferred"; a hybrid data analyst role at a Chicago insurance company required "3 days onsite per week at our Loop office"; and a fully onsite ML engineer role at a defense contractor in the D.C. area required active security clearance. Same job board, same week, three completely different geographic and logistical realities. The work-model label is the start of the filter, not the end of it.
Salary Bands Tell You Where the Market Is Serious and Where It Is Hand-Wavy
The pay range is often the fastest way to spot the level
Salary transparency in job listings has improved since pay transparency laws took effect in states like Colorado, New York, and California. In the June 2025 sample, roughly 55% of postings included a salary range, either posted directly or estimated by the platform. The ranges were the single fastest way to identify whether a role was genuinely entry-level, mid-level, or senior — faster than the title, and sometimes more honest than the years-of-experience requirement.
A posting that says "data scientist" but lists a range of $65,000–$80,000 is telling you something: it's either a junior role, it's in a lower cost-of-living market, or the employer is not competing for experienced candidates. A posting with a range of $160,000–$200,000 is not a career-switcher role regardless of what the requirements say.
Junior pay, mid-level pay, and senior pay do not compress the same way
From the sample of postings with disclosed compensation, approximate ranges by level in U.S. markets looked like this: junior and entry-level data science roles clustered between $70,000 and $100,000; mid-level roles ranged from $110,000 to $150,000; senior and staff roles ran from $150,000 to $220,000 or higher in high-cost markets. These figures align reasonably well with levels.fyi's compensation data for data roles at larger tech companies, though smaller employers and non-tech industries tend to post at the lower end of each band.
The compression point worth noting: the gap between mid and senior is often larger in practice than the gap between junior and mid, because senior roles increasingly include equity and bonus components that don't show up cleanly in the base range.
What this looks like in practice
From the sample: a junior data analyst posting at a regional healthcare system in Ohio listed $72,000–$85,000 with no equity, fully onsite. A mid-level data scientist posting at a SaaS company in Austin listed $120,000–$140,000 with "competitive equity," hybrid. A senior ML engineer posting at a major tech company in Seattle listed $175,000–$210,000 base, with equity and bonus on top. Each range is a signal: scope of work, seniority expectation, and how seriously the employer is competing for talent. When salary is missing from a posting entirely, treat it as a yellow flag — either the employer hasn't decided what they're willing to pay, or they're collecting résumés without active intent to hire.
What This Means for Career Switchers, Students, and Active Job Seekers
If you're switching careers, stop aiming at the wrong tier
The career switcher mistake is applying to mid-level data scientist jobs because the requirements list Python and SQL, which you now know. Those requirements are necessary but not sufficient. Mid-level postings are looking for people who have already done the job, not people who have learned the tools. The better target for most career switchers is roles that explicitly value transferable skills: data analyst roles that reward domain knowledge from your previous career, or associate data scientist roles at companies that have formal mentorship tracks.
If you spent five years in finance, a financial data analyst role at a bank is a more realistic first step than a data scientist role at a tech company — and it's a legitimate path into the field, not a consolation prize.
If you're a student, the real question is whether the bet still pays off
The answer is yes, with conditions. A data science or statistics degree still opens doors that a bootcamp certificate alone does not — particularly at employers who use educational credentials as a first-pass filter for junior roles. But the degree is not sufficient on its own. Employers in the sample consistently listed portfolio projects, internship experience, and demonstrated ability to work with real datasets as requirements even for entry-level postings. The student who graduates with a degree and no applied work is in a worse position than they would have been five years ago, when the field was less crowded and employers were more willing to train from scratch.
If you're already in the market, the play is sharper positioning
Active job seekers who are not getting callbacks usually have one of two problems: they're applying too broadly, or their resume doesn't match the specificity of what listings are asking for. The fix is to pick a lane — data analyst, data scientist, or ML engineer — and tailor every application to that lane's specific skill language. A resume that lists "Python, SQL, machine learning, data visualization, statistical modeling, deep learning, NLP" without context is not impressive; it's noise. A resume that shows a specific project where you used Python and SQL to solve a business problem, with a measurable outcome, is the thing that actually gets read.
How Verve AI Can Help You Prepare for Your Data Scientist Job Interview
Once you've identified the right tier and the right role type, the next obstacle is the interview itself — and data scientist interviews are more demanding than most. They test technical depth, business judgment, and communication simultaneously, often in the same question. The structural problem is that most candidates practice the technical parts in isolation and then blank when the interviewer asks them to explain why their model choice made business sense.
Verve AI Interview Copilot is built for exactly that gap. It listens in real-time to the live conversation and responds to what you're actually saying — not a canned prompt. If an interviewer follows up on your A/B testing explanation with "how would you have handled that if the sample size was half as large?", Verve AI Interview Copilot sees the full context and can surface a relevant response path instantly. It stays invisible during screen share, so you're not managing a visible tool while trying to think. For candidates who are switching careers and need to bridge domain knowledge into data science language, Verve AI Interview Copilot can help you translate what you already know into the framing that data science interviewers are actually listening for.
FAQ
Q: Is data science still a viable career path for new entrants in 2026?
Yes, but the entry point is narrower than it was five years ago. The field is growing, but employers are more selective, and the junior share of open postings is around 15% of the market. Viable doesn't mean easy — it means the path exists if you target the right roles and build the right foundation.
Q: How many of the current openings are actually suitable for junior candidates or career switchers?
Based on the June 2025 sample, roughly 15% of postings were explicitly entry-level or junior. That's a real number of openings, but the competition for those roles is intense because they attract every career changer and bootcamp graduate in the market simultaneously.
Q: What skills show up most often in live data science listings, and which ones are non-negotiable?
Python and SQL are non-negotiable — they appeared in roughly 85% and 80% of listings respectively. Beyond those, the ability to frame business problems analytically came up consistently in the core requirements, not just the soft-skills section. Fancy ML frameworks are nice-to-haves for most roles below senior level.
Q: How do data scientist roles differ from data analyst and machine learning engineer jobs in today's market?
Data analysts focus on answering business questions with existing data using SQL and visualization tools. Data scientists build and communicate predictive models. ML engineers build and maintain the infrastructure that serves those models at scale. The skill requirements diverge significantly, and applying to the wrong role type is one of the most common reasons qualified candidates get ignored.
Q: What salary ranges should candidates expect at junior, mid, and senior levels?
From the June 2025 sample: junior roles clustered between $70,000 and $100,000; mid-level between $110,000 and $150,000; senior between $150,000 and $220,000 in major markets. Non-tech industries and smaller employers tend toward the lower end of each band, and equity can significantly change the senior-level picture at tech companies.
Q: What should a portfolio or resume emphasize to stand out for these roles?
Pick a lane — analyst, scientist, or ML engineer — and tailor to it. Show one or two specific projects with a clear business question, the tools you used, and a measurable outcome. Breadth of tools listed without context is noise. Demonstrated ability to solve a real problem with real data is what actually gets read.
Q: Are employers hiring mostly remote, hybrid, or onsite data scientists in this market?
Hybrid is the plurality at roughly 40% of postings, with remote around 30% and onsite around 30%. Remote skews toward senior roles. Junior and entry-level postings are more likely to require hybrid or onsite presence, and many "remote" postings include geographic restrictions that meaningfully narrow the applicant pool.
Conclusion
The live verdict on data science jobs is this: the field is real, the demand is real, and the path is still open — but it rewards specificity in a way it didn't five years ago. The market isn't handing out junior offers to anyone who learned Python on the weekend. It's hiring people who know which kind of data role they're actually suited for, who can demonstrate applied skills on real problems, and who can explain their analysis to someone who doesn't care about the model.
If you're switching careers: target roles that value your domain knowledge, aim at data analyst positions before data scientist ones, and be honest about which tier of the market is actually open to you right now. If you're a student: get applied experience before you graduate, because the degree alone isn't the differentiator it used to be. If you're actively job searching: pick a lane, tighten your resume to match the listing's actual requirements, and stop applying to mid-level roles with an entry-level background.
The market is selective. That's different from closed. Know the difference, and apply accordingly.
Drew Sullivan
Interview Guidance
