Career switch · Arts and humanities to data analytics
Can BA, BSc, Arts & Humanities Students Become Data Analysts? Yes — Here Is the Complete Roadmap
A data analyst takes a question someone in a business cares about, such as "why did repeat orders fall last quarter?", finds the data that can answer it, checks that the data is trustworthy, and explains the answer so that someone can act on it. Read that again and count how much of it is reading, questioning and explaining. That is why graduates from BA, BSc and humanities backgrounds do move into this work. The coding part is real, but it is smaller than most people fear, and it can be learned in a fixed order.
What your degree has already trained
Different degrees give different head starts. This table is a guide to where you can lean on your background, not a promise about any particular job.
| Degree | What carries over | Natural first domain |
|---|---|---|
| BA Economics | Index numbers, demand and supply logic, reading charts of trends | Sales, pricing, market research |
| BA Psychology / Sociology | Survey design, sampling, research methods, sometimes SPSS | Customer research, HR analytics, social-sector data |
| BA English / Journalism | Clear writing, summarising long material, asking precise questions | Reporting analyst, content and marketing analytics |
| BA History / Political Science | Working with sources, judging reliability, building an argument | Policy research, public data, operations reporting |
| BSc (Maths, Physics, Chemistry, Biology) | Lab records, measurement error, formulas, often basic statistics | Healthcare, quality, research support |
The maths you will actually use
Most worry comes from imagining calculus. Day-to-day analyst work looks more like this: last month 1,240 customers placed an order, this month 1,116 did. The change is 1,116 − 1,240 = −124, and as a percentage of last month that is −124 ÷ 1,240 = −10%. You will also use averages and medians (and know when the median is fairer, such as with incomes), percentages of a total, and simple comparisons between groups. Later you add sampling, confidence intervals and correlation. If you can follow the arithmetic above, you can start.
Step 1: Excel, where you produce something in week one
Spreadsheets let you see every number, which makes them the friendliest place to begin. Learn tables, sorting and filtering, XLOOKUP, SUMIFS and COUNTIFS, pivot tables and basic charts.
' Look up a reader's membership type from another sheet
=XLOOKUP(A2, Members!A:A, Members!C:C, "Not found")
' Count loans of one genre in one month
=COUNTIFS(Loans!B:B, "Fiction", Loans!D:D, "July")
' Share of total, formatted as a percentage
=B2 / SUM(B$2:B$6)
The role of Excel in data analytics explains where spreadsheets stop being enough.
Step 2: SQL, the skill that gets you shortlisted
Company data lives in databases, and almost every analyst interview includes SQL. The language reads close to English, which helps humanities graduates more than they expect. Here is a small library example:
CREATE TABLE borrowings (member_id INTEGER, genre TEXT, borrowed_on TEXT);
INSERT INTO borrowings VALUES
(101,'Fiction','2026-07-03'), (102,'History','2026-07-04'), (101,'Fiction','2026-07-10'),
(103,'Poetry','2026-07-11'), (104,'Fiction','2026-07-15'), (102,'History','2026-07-20'),
(105,'History','2026-07-22'), (103,'Fiction','2026-07-28');
SELECT genre,
COUNT(*) AS loans,
COUNT(DISTINCT member_id) AS readers
FROM borrowings
GROUP BY genre
ORDER BY loans DESC;
genre | loans | readers
Fiction | 4 | 3
History | 3 | 2
Poetry | 1 | 1
The interesting part is the difference between the two columns: Fiction has four loans but three readers, so one member borrowed twice. Spotting that is analysis; writing the query is only the tool. Learning SQL step by step sets out a practice order.
Step 3: a dashboard tool
Power BI or Tableau turns your tables into a report that a manager can filter. Learn one well rather than both badly. Your writing background helps here: a dashboard with a clear title, three key numbers and a short note beats one with fifteen charts.
Step 4: Python for cleaning, when you are ready
Python is not needed on day one, but it becomes useful when data is messy. Survey data, which psychology and sociology students know well, is a good first case:
import pandas as pd
survey = pd.DataFrame({
"respondent": [1, 2, 3, 4, 5, 6, 7, 8],
"city": ["Lucknow", "lucknow ", "Patna", "Patna", "Jaipur", "Jaipur", "Patna", None],
"satisfaction": ["4", "5", "2", "3", "4", "n/a", "2", "5"],
})
survey["city"] = survey["city"].str.strip().str.title()
survey["satisfaction"] = pd.to_numeric(survey["satisfaction"], errors="coerce")
clean = survey.dropna(subset=["city", "satisfaction"])
print(clean.groupby("city")["satisfaction"].agg(["count", "mean"]).round(2))
count mean
city
Jaipur 1 4.00
Lucknow 2 4.50
Patna 3 2.33
Two of the eight responses were dropped: one had no city and one said "n/a". Jaipur's average now rests on a single answer, and a careful analyst says so before anyone draws conclusions. Your research-methods training is exactly what makes you notice this.
Projects that suit your subject
- Economics: state-wise inflation or price data from a public source, with a dashboard and a short commentary.
- Psychology or sociology: design a short survey, collect responses, clean them in Python and present the findings honestly, including the limits of the sample.
- English or journalism: analyse a public dataset behind a news story and write the explainer an editor would publish.
- BSc: analyse open health or environmental data and explain measurement issues in plain language.
For layout and write-ups, see building a job-ready analytics portfolio.
Answering "why should we hire an arts graduate?"
Do not apologise for your degree. Name the skill it gave you, then prove you have added the technical layer. For example: "My sociology degree trained me to design surveys and question samples. I have since learned SQL and Power BI, and in my survey project I found that one city's result was based on too few responses to report." Interviewers remember a candidate who can explain a limitation. Expect a SQL test either way; practise with commonly asked SQL queries.
A realistic pace
Studying around 10 hours a week, many people need about six to nine months to cover Excel, SQL, one BI tool and two or three projects. Graduates who are working full-time often take longer. Progress is measured by finished projects, not by course certificates.
Frequently asked questions
Do I need maths in 12th to become a data analyst?
No. Entry-level analysis uses percentages, averages, medians, growth rates and simple charts, which you can learn as an adult. Statistics beyond that, such as sampling and testing, is learned during the course of study, not assumed.
Will companies reject me because I studied BA?
Some job postings filter by degree, especially campus drives that ask for B.Tech or BCA. Many analyst postings ask for any graduate. Apply where your degree is accepted, and let your projects do the convincing.
Which is better for an arts graduate: data analyst or business analyst?
Both suit arts graduates. A data analyst spends more time in SQL and dashboards; a business analyst spends more time on requirements, processes and stakeholders. Learning SQL and Excel keeps both doors open.