How to choose a thesis topic — 7 criteria and 40 ideas (2026)

Deciding how to choose a thesis topic is the single decision that shapes the next four to six months of your final year. Pick something too broad and you drown in literature; pick something too narrow or too fresh and you cannot find a single source. The good news: topic selection is not a flash of inspiration — it is a checklist. This guide gives you seven concrete criteria, a one-sentence feasibility test you can apply in ten seconds, a worked example of narrowing a bloated idea, 40 topics across five popular majors, and the traps that quietly ruin otherwise smart proposals. Work through it in order and you will walk into your supervisor meeting with a defensible topic, not a vague theme.

Why the topic decides half your grade before you write a word

A thesis is graded on the quality of your argument and evidence, but both are capped by the topic you chose. If the topic has no accessible data, no amount of effort produces an empirical chapter. If it is so broad that any textbook already covers it, you have nothing original to add. Examiners can tell within the first two pages whether a topic was scoped deliberately or grabbed in panic the week before the deadline.

Choosing well also protects your motivation. You will read dozens of papers and rewrite the same paragraphs many times. A topic you find genuinely interesting is the difference between finishing and stalling in March. Treat selection as the most leveraged hour of the entire project.

The 7 criteria for choosing a thesis topic

Score every candidate topic against these seven criteria. A strong topic scores well on at least five; if it fails on data availability or method feasibility, drop it regardless of how exciting it sounds.

1. Genuine interest

You need enough curiosity to survive the boring middle. Ask whether you would still read an article on this topic if it were not for the grade. Mild interest is fine; indifference is fatal.

2. Data availability

This is the criterion that eliminates most doomed topics. Before committing, confirm you can actually reach the data — a public dataset, a company willing to share figures, a population you can survey, or a body of documents you can analyze. “I will study employee motivation at Google” fails here unless Google lets you in.

3. Scope

The topic must be finishable in one semester at the expected length. A bachelor’s thesis answers one focused question, not a whole field. If your title contains “and”, “the impact of everything on”, or a whole industry, it is too wide.

4. Novelty

You do not need a Nobel-worthy discovery — you need a small angle nobody has framed exactly your way: a new population, a recent time window, a fresh comparison, or a local context. Replicating a known study in your own city is a legitimate contribution at this level.

5. Method feasibility

Match the topic to a method you can execute. If you cannot run regression analysis, do not pick a topic that demands it. Qualitative interviews, content analysis, case studies, and simple survey statistics are all defensible when done properly.

6. Supervisor fit

Your supervisor’s expertise is a resource. A topic close to their research area means sharper feedback and faster approval. A topic they know nothing about means you are on your own.

7. Career relevance

When two topics tie, choose the one that builds a portfolio you can show an employer or a future master’s program. A thesis is also a writing sample.

Criterion	Question to ask	Deal-breaker if…
Interest	Would I read this for fun?	Total indifference
Data availability	Can I actually get the data?	No realistic access
Scope	Finishable in one semester?	Covers a whole field
Novelty	What is my fresh angle?	Pure textbook rehash
Method feasibility	Can I run this method?	Requires skills you lack
Supervisor fit	Does someone here know this?	No qualified supervisor
Career relevance	Does it help my next step?	(soft — tie-breaker only)

The one-sentence feasibility test

Once a topic passes the seven criteria, compress it into a single sentence in this exact shape:

“I will compare X and Y using Z data to answer question Q.”

For example: “I will compare customer churn between subscription and one-off buyers using two years of a SaaS company’s transaction data to answer whether pricing model predicts retention.” If you cannot fill in every slot — especially Z, the data — the topic is not ready. This sentence becomes the backbone of your introduction and the first thing your supervisor will react to. Anyone who can say it out loud in one breath has a topic; anyone who needs a paragraph does not yet.

How to narrow a topic that is too broad (worked example)

Most students start too wide. Narrowing is a mechanical process of adding constraints until the topic is finishable. Watch a typical idea shrink:

1. “Social media marketing.” A whole discipline — unwritable.
2. “The impact of social media marketing on sales.” Still every platform, every industry, forever.
3. “The impact of Instagram marketing on sales in the fashion industry.” Better — one channel, one sector.
4. “How Instagram Reels affected online sales for three Polish fashion SMEs in 2024–2025.” A thesis: one format, one channel, a defined population, a fixed time window, and data you can request from three cooperating firms.

Each step added a constraint: channel, industry, format, geography, time, and sample size. Keep adding constraints until you can answer the feasibility sentence. A topic that feels almost too small is usually exactly right.

40 thesis topic ideas by field (2026)

Use these as starting points, not final titles. Each still needs the narrowing process above before it becomes a thesis. They are deliberately framed as answerable questions.

Business and management

#	Topic idea
1	AI adoption in HR recruitment: efficiency vs. candidate trust in mid-size firms
2	Hybrid work policy and self-reported productivity in one IT company
3	Employer branding on LinkedIn and application rates for Gen Z candidates
4	ESG reporting quality among listed retail companies since 2023
5	Onboarding process redesign and 90-day turnover in a service firm
6	Agile transformation outcomes in a non-tech department
7	Customer loyalty programs and repeat purchase behavior in e-commerce
8	Leadership style and remote team engagement: a case study

Economics and finance

#	Topic idea
9	Inflation expectations and household saving behavior 2022–2025
10	Buy-now-pay-later usage and debt levels among young adults
11	Minimum wage increases and small-business hiring in one region
12	Cryptocurrency volatility and retail investor sentiment
13	Green bond pricing versus conventional bonds
14	Determinants of startup failure in the first two years
15	Tourism spending recovery in a post-pandemic local economy
16	Interest rate changes and mortgage demand

Psychology

#	Topic idea
17	Smartphone use before sleep and self-reported insomnia in students
18	Social comparison on Instagram and body image in young women
19	Perceived academic stress and procrastination
20	Remote work and loneliness among early-career employees
21	Mindfulness apps and reported anxiety over four weeks
22	Parenting style and adolescent self-esteem
23	Music genre and short-term concentration
24	Digital wellbeing habits and screen-time guilt

Education

#	Topic idea
25	AI writing tools and student learning autonomy: teacher perspectives
26	Gamification and vocabulary retention in language classes
27	Flipped classroom outcomes in secondary mathematics
28	Teacher burnout and workload after curriculum reform
29	Inclusive education support for students with dyslexia
30	Parental involvement and homework completion
31	Peer feedback and essay quality in higher education
32	Microlearning videos and exam performance

Computer science and IT

#	Topic idea
33	A RAG chatbot for university FAQ: accuracy vs. hallucination
34	Sentiment classifier for product reviews: model comparison
35	A web app for personal finance tracking (design and evaluation)
36	Image classification for waste sorting on limited hardware
37	Comparison of authentication methods for a small web service
38	Recommendation system for a niche online store
39	Static site performance: framework benchmark under real traffic
40	Accessibility audit and remediation of a public-sector website

Common traps that sink a thesis topic

Even attractive topics fail for predictable reasons. Watch for these before you commit.

No realistic data access. The most common killer. “The impact of internal culture at a Fortune 500 firm” is unwritable if you have no way in. Confirm access before approval, in writing where possible.

Topics too fresh for literature. A brand-new phenomenon may have zero peer-reviewed sources. You can check this in minutes: search the topic on Google Scholar and, for health or life-science angles, on PubMed. If nothing scholarly appears, you will have no theoretical chapter to write.

Politically or ethically loaded topics. Subjects that require sensitive personal data, or that invite bias in a short thesis, create ethics-approval delays and hard grading. Choose a version you can study neutrally.

Topics that are secretly two theses. If your question has an “and” joining two independent investigations, split it and keep one half.

What to do when your supervisor imposes a topic

Sometimes you do not get a free choice — the supervisor assigns a theme, or approves only topics inside their project. This is not a disaster; it removes the hardest part of the decision. Your job shifts from choosing to carving out your angle:

• Ask for the exact boundaries: which question, which method, which data are already available in their project.
• Propose a narrowed sub-question using the feasibility sentence, so you own a defined slice rather than a vague mandate.
• Confirm who owns the data and whether you may publish it in an appendix.

An imposed topic with a clear dataset often beats a free choice with none. Reframe it as a constraint that already solved your data-availability problem.

How AI can help you brainstorm and pre-validate a topic

Modern AI tools are excellent at the messy early stage: generating variations, stress-testing scope, and surfacing whether a topic likely has literature. Ask a model to “give ten narrower versions of this topic, each with a specific dataset and method”, then run each candidate through the seven criteria yourself. Use AI to expand and pressure-test options — but verify every suggested source manually, because models still invent citations. Our comparison of the best AI tools for academic writing shows which ones handle sources responsibly.

If you would rather move straight from an approved topic to a structured draft, the Smart-Edu bachelor’s thesis generator builds a full, chapter-structured thesis with a bibliography in 30–90 minutes from 249 zł, keeping the scope and methodology you defined here. It is a drafting accelerator, not a replacement for your own analysis — you still own the argument, the data, and the defense. For the full workflow once your topic is locked, see our guide on how to write a bachelor’s thesis with AI, and for longer projects, how to write a master’s thesis with AI.

Frequently asked questions about choosing a thesis topic

How early should I choose my thesis topic?

Ideally at the very start of your final year, before literature courses end, so you can steer your reading toward your topic. A realistic minimum is three to four months before the submission deadline — enough time to secure data, get supervisor approval, and still write. Choosing late is the most common cause of a rushed, thin empirical chapter.

Can I change my thesis topic after starting?

Yes, but the cost rises the later you switch. A pivot in the first few weeks — before you have written the theoretical chapter — is usually painless. After the empirical work begins, changing topic often means restarting. If you must change, narrow within the same theme rather than jumping fields, so your existing reading still counts.

How specific should a bachelor’s thesis topic be?

Specific enough to answer with one method and one dataset in a single semester. A good test: if you can state it in the “I will compare X and Y using Z data” sentence and name the data source, it is specific enough. If it still sounds like a chapter of a textbook, keep narrowing.

What if two students pick the same topic?

That is fine as long as your angle, data, or population differs. Two theses on “employee motivation” can coexist if one studies a hospital and the other a software firm. Talk to your supervisor early so you can differentiate the framing rather than compete on identical ground.

Summary

Knowing how to choose a thesis topic comes down to discipline, not luck: score each idea against the seven criteria, compress the survivor into a single feasibility sentence naming its data, and keep adding constraints until it is finishable in one semester. Avoid the traps — no data, no literature, or two theses hiding in one — and treat an imposed topic as a head start rather than a limitation. Spend a focused hour here and every later stage of writing gets easier.