Can a humanities graduate become a data analyst? My route from literature

I studied Polish literature at Jagiellonian University, specialising in comparative literature with elements of literary anthropology. I was interested in texts, cultures and meaning. Excel did not occupy a particularly important place in my life. I could open it, which was a modest starting point for the career that followed.

The question of whether a humanities graduate can become a data analyst is therefore personal for me. I followed that route myself. At the same time, I would not present my history as a recipe with a guaranteed job at the end. I had particular opportunities, met particular people and learned at my own pace. Someone else may encounter those steps in a different order.

I want to explain what connected my earlier experience with analytical work. I will start with spending money in Portugal, move through spreadsheets and customer service, then show how I would test an interest in data today. No declaration that I am becoming an entirely different person next Monday is required.

My first analytical question concerned my bank balance

Towards the end of my studies, I went to Portugal on an Erasmus exchange. There were trips, friends and plans for further travel. There was also a practical problem: I had less money than I expected. I wanted to hitchhike from Portugal to Morocco, and my bank balance had certain reservations about the proposal.

I did not know exactly where the money had gone. A conversation with my flatmate drew my attention to spending on cigarettes. An individual purchase seemed small. Repetition made it significant. That mechanism later became useful in thinking about data: a small event can produce a large effect when it happens often enough.

At the time, I had no plan to become an analyst. I had an ordinary problem and wanted to understand it. That makes the story relevant to someone starting to learn. It is easier to stay interested in a table when the answer matters to me than when I am completing another exercise about an anonymous company selling unspecified products.

I do not conclude that everything in life needs measuring. Relationships, rest and meaningful work do not fit neatly into one column. In that situation, numbers helped me answer a specific question about spending. That was enough to get me interested in the tool.

It also gave me a reason to return to the information. A one-off calculation might explain a disappointing balance. A repeatable way of recording transactions could help me notice the pattern earlier next time. That difference between a single answer and a useful ongoing process became increasingly interesting to me.

A budget in Excel taught me more than addition

After returning to Poland, I worked in a hostel. I began recording expenses, income and monthly balances in Excel. The file gradually became a budget I could revisit. Instead of a vague feeling that money was disappearing, I had a sequence of events that I could compare.

I disguised cigarette purchases as “crisps.” Looking back, that is quite a good lesson in data quality. I can have correct formulas and a carefully coloured table while making interpretation harder through labels that conceal the real meaning. The spreadsheet added things up honestly. The creative contribution came from the person entering the records.

A simple budget requires decisions. Does each purchase get a separate row? How do I treat transfers between my own accounts? Does a refund reduce an expense or appear as a separate transaction? Without consistent rules, I can calculate everything correctly and still answer a different question from the one I intended.

That is what I now emphasise when explaining the beginning of analysis. A formula is part of the work. Before writing it, I need to establish what a row represents, where the information comes from and what a category means. Otherwise, I quickly end up analysing the consequences of my own inconsistent recording.

The budget also provided feedback. I could see whether a change in behaviour appeared in the data. I did not need to wait for a course assessment. The question, the evidence and the decision came together in one place, giving me a reason to open the file again.

I entered corporate work through customer service

I later joined Capita in customer service. The work involved messages, calls and customer issues to resolve. A stable job gave me a sense of security, but over time I became increasingly aware of the repetitive nature of the duties. I began looking for opportunities to work differently.

When reports and summaries became part of my responsibilities, I turned to Excel. I improved how information was prepared and helped analyse data about the team’s work. I did not need a new job title before taking an interest in a problem. I had access to a concrete process and could see where effort was being repeated.

I also experimented with formulas to prepare recurring email responses. That was a practical exercise in combining information according to rules. Automation tools look different today, but the underlying questions remain familiar: what repeats, what varies and which cases need separate treatment?

Seeing the result mattered. A summary could be prepared more efficiently, and the rules became clearer. I started enjoying the search for a solution itself. That is a more useful signal of interest in a profession than liking photographs of laptops on beaches. Sand in a keyboard has limited educational value anyway.

My practical suggestion from this part of the story is to look for a small analytical task in your current surroundings. Naturally, it needs to fit your responsibilities and access to information. I would not assume every employer welcomes every experiment, but it helps to know exactly which problem I would like to solve.

Customer service also gave me context for the data. A record represented something that happened to a person or a case. That makes it easier to ask whether a category describes the process accurately, whether a missing value is expected and whether a proposed metric reflects the work people actually do.

Learning and interviews did not form an uninterrupted success story

Eventually, I decided I wanted to work with data professionally. I studied in the evenings, including Microsoft courses on edX covering Excel and the foundations of analysis. I needed to organise knowledge that I had previously encountered while solving individual problems.

Combining practice with structured learning mattered to me. Once I knew a problem, I could more easily understand the purpose of a particular function. A course also introduced possibilities I might not have discovered independently. The two supported each other as long as I returned to doing tasks myself.

I applied to UBS and was invited to an interview involving Excel exercises. I did not get the job. That deserves to remain in the story. It is easy to retell a career change as if every attempt led directly to the next promotion, especially after knowing how the story eventually turned out.

I later went through recruitment at Lufthansa. There were tests, data analysis, charts and tasks under time pressure. This time, I was accepted. My earlier degree remained on my CV, and I did not suddenly become a computer science graduate. I demonstrated skills needed in that particular recruitment process.

I do not translate that experience into a promise that every employer disregards degree subject. Vacancies and roles have different requirements. My story demonstrates that this route is possible; it does not establish a universal hiring rule. For a career change today, I would examine current opportunities and the work involved rather than build the whole plan around someone else’s biography.

My first JOIN opened another door

I remember a colleague showing me SQL and how to join two tables. For someone who had mainly worked in spreadsheets, that was a powerful moment. I saw another way to ask questions of data and work with a different scale of problems.

It did not mean I suddenly understood everything. I learned, made mistakes and kept discovering things I did not yet know. My first data role brought new tasks and new reasons to expand my knowledge. Receiving a job title does not finish the learning process.

This is why I am careful with enormous lists of tools supposedly required before a first application. The scope should follow the role. Trying to learn everything at once makes it easy to collect introductions for months while still struggling to solve a modest problem independently.

For someone starting with Excel, SQL can be a natural next area. I still want to understand why I am adding it: retrieving information, combining datasets and calculating a result. My article on SQL JOIN types includes examples and row-count checks. The excitement of a first successful query is worth following fairly quickly with verification.

What humanities habits can bring to analytical thinking

I do not believe a degree automatically supplies a set of abilities. A person can graduate from any subject and still read carelessly. However, several habits associated with humanities study transfer usefully to analysis when they have actually been practised.

The first is defining terms precisely. When someone mentions an active customer, I want to know what activity means. A purchase in the last month, a login or a current contract? Each definition selects a different group. Before counting anything, I need a shared understanding of the concept.

The second is assessing sources and context. Information from a sales system answers a different question from a response in a survey. Even identically named columns may be produced according to different rules. Reading the data description carefully is less spectacular than creating a new chart, but it can rescue the conclusion.

The third is building an argument. I want to distinguish an observation from an interpretation and identify what the evidence cannot establish. If complaints have increased, I do not yet know that service has deteriorated. The customer base might have grown or the recording process might have changed. I need another check.

The fourth is clear writing. Someone may make a decision based on my work and should understand the result without guessing what I intended. Explaining a limitation in one plain sentence is a practical skill. The O*NET description of Business Intelligence Analysts includes reporting and communicating information alongside working with data.

These habits work best when connected to the calculation. I need to show how a definition changes the selected records, how a source limitation affects the result and why a recommendation follows from the evidence. A thoughtful introduction cannot compensate for an incorrect total further down the page.

What my earlier studies could not do for me

I had to learn tools and check whether I could actually use them. Understanding an argument in an essay does not automatically produce a correct formula, query or chart. That requires practice, mistakes and a willingness to revisit the basics when the output differs from expectations.

I also needed to become comfortable with numbers. Proportions, percentages, averages and denominators are useful early on. If complaints rise from one hundred to one hundred and twenty, the twenty percent increase tells me little without the number of orders. The example requires attention rather than advanced mathematics.

More demanding statistical questions appear over time. The necessary depth depends on the work I want to do. I would not tell someone that an analyst can always avoid mathematics. I would break learning into concrete questions and identify the concepts needed to answer them correctly.

Patience with technical details matters too. A date stored as text, an invisible space or an incorrect relationship between tables can disrupt an analysis. If I only enjoy presenting a grand conclusion, it is useful to discover how I feel about the part involving a search for such details.

None of this invalidates my earlier education. It simply identifies additional work. I find it easier to learn when I can say “I need to understand this calculation” than when I turn one difficult exercise into a verdict about whether I am fundamentally a technical person.

A small exercise I would use to test my interest

If I wanted to test my interest in analysis today, I would start with a synthetic budget. I would create thirty invented transactions containing a date, category, amount and short description. Each row would represent one expense. Using fictional information would let me show the file to someone else without sharing personal finances.

First, I would write the question: which categories account for the largest share of known expenses in a particular month? I would define the date range and treatment of missing values. If an amount were unknown, I would mark it explicitly. I would not turn it into zero merely to make the table look tidy.

Then I would calculate the total and each category’s share. In a small example with known expenses of one thousand units and transport costs of two hundred, transport accounts for twenty percent. I would check that myself. I want to know the expected answer before trusting a formula.

I would deliberately add a misspelled category, a duplicated transaction and a date outside the month. Then I would try to find them and decide how to handle them. A duplicate requires evidence that it represents the same purchase. Two identical amounts can belong to separate events.

Next, I would write five sentences describing the question, the result, what surprised me, the limitations and a proposed next step. I would add a chart only if it made comparison easier. The short note shows whether I can connect the calculation to the question that started the work.

Finally, I would change one input and check the effect. Adding another valid transport transaction should change both the transport total and its share. Moving that transaction outside the selected month should remove it from the result. These tiny experiments reveal more about my understanding than a screenshot of a finished chart.

After the exercise, I assess the work rather than my label

I would consider which part of the exercise felt satisfying. Did I want to understand a discrepancy? Was I curious about another month? Did fixing a category feel more rewarding than simply finishing the instructions? Those observations help me decide what to learn next.

One exercise cannot settle an entire career choice. It can show whether I want to try another, slightly more demanding problem. I would also ask someone to read my note and identify unclear points. If I need to explain half the assumptions aloud, I would improve the description before expanding the project.

My route combined curiosity, practical tasks, learning and particular professional opportunities. I did not have to erase literature from my CV. I had to add new abilities and demonstrate that I could solve problems using data. That is a more useful question than whether my previous label grants admission to the field.

The next step can remain small: a completed problem, a checked answer and an honest decision about whether I want to return to this tomorrow. My own journey began with the need to understand something very ordinary. A question about where the money went turned out to be enough to get me started.