What a proposal has to prove
A research proposal has four jobs: to show that the problem matters, that the question is answerable, that the method fits and that the plan is realistic. It is read by a supervisor or committee who decides whether to approve, and approval often depends less on the novelty of the idea than on the clarity of the design. A modest, well-defined project is approved; a grand, vague one is sent back.
| Reviewer asks | Where the proposal answers it |
|---|---|
| Why does this matter? | Background and problem statement |
| What exactly will you find out? | Research question, aims, hypotheses |
| What is already known? | Brief literature review and gap |
| How will you find out? | Design, sample, instruments, analysis |
| Is it ethical and feasible? | Ethics, access, resources, timeline |
A structure with word counts
Follow your program's template if it has one. A common structure for a 2,500-word proposal is below.
| Section | Words | What it contains |
|---|---|---|
| Title | 15 words | Names the topic, population and approach |
| Background and problem | 350 | Context, evidence the problem exists, why it matters |
| Research question and aims | 200 | One main question, two or three sub-questions or aims |
| Literature and conceptual framework | 600 | Key theories and studies, the gap, the framework or hypotheses |
| Methodology | 700 | Philosophy, design, sample, data collection, analysis, validity |
| Ethics and risks | 200 | Consent, confidentiality, data storage, risks and responses |
| Timeline and resources | 150 | Milestones, budget, access |
| Expected contribution | 150 | What changes if the study succeeds |
| References | Not counted | Cited sources |
Make the problem concrete
Start with evidence, not generalities. Quote a statistic from a credible source, a trend, a documented organizational problem or a conflict in the literature. Then narrow to the focus of your study.
| Vague opening | Concrete opening |
|---|---|
| Employee engagement is important to organizations. | Surveys of mid-sized professional service firms report engagement scores that fall after the first year of employment, yet little research explains why in firms with flat hierarchies. |
| Digital marketing is growing. | Small retailers spend a growing share of their marketing budget on social advertising, but there is little evidence on which formats generate repeat purchases rather than one-off clicks. |
Use your own sources for the evidence and make sure each claim can be traced to one.
Write a research question you can answer
A research question should be clear, focused, answerable with data, and linked to the literature. It also determines the method: how and why questions suit qualitative or case designs; what is the relationship or to what extent questions suit quantitative ones.
| Question | Type | Likely design |
|---|---|---|
| How do managers in family firms decide when to hire outside executives? | Exploratory | Interviews, thematic analysis |
| To what extent does supervisor support predict intention to stay among nurse managers? | Relationship | Survey, regression |
| Do customers who receive personalized offers buy more than those who receive general offers? | Comparison | Experiment or A/B test |
- Avoid yes or no questions They produce thin answers.
- Include population and setting Who and where.
- Keep to one main question Sub-questions should support it.
- Check it is not already answered A quick search may show a decisive study.
Show the method and why it fits
The methodology section is where proposals are won. Name the design, justify it against alternatives and describe each element precisely: population, sampling approach and size, instruments (survey items, interview guide), procedure and analysis technique. For quantitative studies, include a sample size calculation (see our guide to survey design and sampling). For qualitative studies, state how you will judge when you have enough data and how you will keep the analysis rigorous.
Say how you will protect the quality of the findings: validity and reliability in quantitative work, and credibility, transferability and dependability in qualitative work. See the research methods guide for the details.
Ethics, access and feasibility
Ethics approval is a hurdle in most programs, and your proposal should show that you understand the principles: informed consent, voluntary participation, confidentiality or anonymity, data protection and avoiding harm. Say where data will be stored, who will have access and when it will be deleted. If you will study your employer, describe how you will manage the dual role and any pressure on participants.
Feasibility is as important as ethics. Show that you have access to the people or data, enough time and the skills required. A letter or email of support from a gatekeeper is persuasive evidence.
| Risk | Response |
|---|---|
| Low response rate | Multiple reminders; shorter instrument; incentive within ethical limits; backup channel |
| Participants withdraw | Oversample by 15 to 20 percent |
| Access refused | Backup organization identified in advance |
| Sensitive information shared | Anonymize at transcription; secure storage |
Working on this assignment now? Get a price for help with your paper.
Get an instant quoteA model proposal summary
One-paragraph summary (hypothetical)
Regional banks lose about a quarter of new relationship managers within two years, yet most turnover research studies large global banks and focuses on pay. This study asks how onboarding quality and supervisor support relate to intention to stay among relationship managers in regional banks. A cross-sectional survey of about 190 managers, drawn from six banks, will measure both predictors with validated scales and analyze the results with multiple regression. The study will contribute evidence for a context that is rarely studied and practical guidance on early-career retention. Fieldwork is planned for months four to seven, with ethics approval secured by month three.
A summary like this shows the problem, the question, the method, the contribution and the plan in about 120 words. Write yours last, then check that each sentence is supported in the body.
A timeline table reviewers can check
| Task | Start | End | Output |
|---|---|---|---|
| Literature review | Month 1 | Month 3 | Draft chapter and framework |
| Ethics application | Month 2 | Month 3 | Approval letter |
| Pilot survey | Month 3 | Month 4 | Revised instrument |
| Data collection | Month 4 | Month 7 | Cleaned dataset |
| Analysis | Month 7 | Month 9 | Results tables |
| Writing and revision | Month 8 | Month 12 | Final dissertation |
Overlapping tasks are fine, but make sure the dependencies are logical: you cannot start data collection before ethics approval. Add a buffer of a month for the unexpected.
How proposals are usually marked
| Criterion | Strong | Weak |
|---|---|---|
| Problem and significance | Evidence-based and specific | General and unsupported |
| Research question | Focused and answerable | Broad or already answered |
| Use of literature | Shows a gap with sources | Lists sources without a point |
| Method | Fits the question; justified; precise | Vague or mismatched |
| Ethics and feasibility | Realistic access, consent and timeline | Overlooked or optimistic |
| Writing | Clear, concise, well organized | Wordy or poorly structured |
Rewriting a weak question three times
| Draft | Question | Problem |
|---|---|---|
| 1 | How does social media affect business? | Far too broad; no population or outcome |
| 2 | How does social media advertising affect sales for small retailers? | Better, but sales depend on many things; setting unclear |
| 3 | To what extent does the share of marketing budget spent on social video advertising predict repeat purchase rate among independent online retailers in one region? | Specific population, variables and outcome; testable |
Each revision adds a population, a variable or a measure. If you cannot name what you will measure or explore, the question is not ready.
Ethics scenarios reviewers raise
| Scenario | Concern | Response in the proposal |
|---|---|---|
| Surveying your own employees | Pressure to take part; fear of identification | Voluntary, anonymous, collected by a third party, results reported in groups |
| Interviewing customers | Privacy and consent | Consent forms; no personal identifiers in transcripts |
| Using company data | Confidentiality and permission | Written permission; anonymize; agree publication rules |
| Collecting social media data | Expectation of privacy; platform terms | Use public data only, within terms; anonymize quotes |
Timeline and common weaknesses
Present the timeline as a table of tasks, start and end dates and outputs. Include ethics approval, data collection, analysis, writing and revisions, and build in a buffer. Reviewers know that tasks take longer than planned.
- Question too broad The most common reason for rejection.
- Method does not match the question Check that the data can answer it.
- No gap Show what the literature has not answered.
- Unrealistic sample or timeline Test your plan against real constraints.
- Weak writing Short sentences, clear headings, no unexplained jargon.
Use the same structure as your eventual dissertation, so the proposal becomes the first draft of the early chapters. If you want help with a proposal, you can order graduate business research paper help.