HeroApply
Auto Apply to hundreds of jobs in a few hours
English (US)EspañolItalianoPortuguêsFrançaisDeutsch日本語한국어
Sign up

Market Research Analyst interview questions and how to answer them

A market research analyst interview checks two things at once: whether your methods are sound, and whether you can explain them to someone who doesn't care about methods. Expect to critique a badly worded survey, pick a research approach for a vague business question, and walk through a study you ran from brief to readout. Bring one study you're proud of and one where the answer disappointed somebody.

12 questions8 minute read

Part 1

1.Walk me through a research project you ran from start to finish.

Why they ask

This is the backbone of the interview. They want to see the whole chain: the business question, your method, the sample, the analysis and what changed because of it.

How to answer

  • Open with the decision the business was trying to make, not the method
  • Explain why you picked that method and who you surveyed or interviewed
  • Name one problem mid-study, such as a quota that wouldn't fill, and how you handled it
  • Finish with the recommendation and what the team actually did next

2.Here's a survey question from our last study. What's wrong with it?

Why they ask

Question writing is the core craft of the job. Interviewers often hand you a double-barreled or leading question on purpose to see if you spot it quickly.

How to answer

  • Name the specific flaw: leading wording, two questions in one, a missing option or unbalanced scale
  • Say how that flaw would bias the answers, and in which direction
  • Rewrite it on the spot in plain words
  • Mention you'd pretest the rewrite with a few people before fielding

3.A product manager asks, 'Will people buy this?' How do you decide what research to do?

Why they ask

Real requests arrive vague. They want to see you turn a loose question into a study that can actually answer something, within time and budget.

How to answer

  • Ask what decision hangs on the answer and what would change their mind
  • Check what's already known from sales data, past studies or reviews before buying new research
  • Pick an approach, such as a concept test for appeal or interviews if the idea is still early
  • Be clear about what the study can't promise, like predicting actual sales
4.How would you explain statistical significance to a brand manager who has no stats background?

Why they ask

You'll be doing this every week. They want to hear whether you can keep the idea honest without burying people in terms.

How to answer

  • Frame it as how confident you are that a difference isn't just luck of the draw
  • Use a concrete example from a test, like two package designs scoring close together
  • Point out that significant isn't the same as big enough to matter
  • Show how you'd mark it on a chart so nobody has to ask
5.When would you use qualitative research instead of a survey?

Why they ask

Good analysts pick the method that fits the question. Teams want someone who won't force every problem into a questionnaire.

How to answer

  • Use interviews or focus groups when you don't yet know the right questions or answer options
  • Use them to understand why people behave a certain way, in their own words
  • Use a survey once you need to size something or compare groups reliably
  • Give an example where qual shaped the survey that followed

Part 2

6.A subgroup in your crosstabs shows a dramatic result, but the base is small. What do you do?

Why they ask

Thin bases are where research gets misused. They want to know you'll protect the team from a headline the data can't hold.

How to answer

  • Check the base size and whether the difference holds up statistically
  • Report it as directional at most, with the caveat on the slide itself
  • Suggest a boost sample or a follow-up study if the finding matters to a decision
  • Keep it out of the executive summary if it can't carry the weight
7.How do you make sure the data you get back from a panel is trustworthy?

Why they ask

Online panel data can be full of bots, fraud and careless answers. They want a researcher who checks, not one who assumes.

How to answer

  • Build in attention checks and a consistency question during programming
  • Review completion times, repeated answer patterns and nonsense open ends
  • Agree with the vendor up front on how bad completes get removed and replaced
  • Compare key results against a known benchmark or past wave for sanity
8.Tell me about a time your findings contradicted what a stakeholder wanted to hear.

Why they ask

This happens often in the role. They're checking whether you hold your ground on the data while keeping the relationship intact.

How to answer

  • Set up the stakes: what the stakeholder believed and what the study showed
  • Explain how you shared it early and privately before the big meeting
  • Show how you separated what the data proved from what was still open
  • Say what happened after and what you learned about delivering bad news
9.What's the difference between MaxDiff and conjoint, and when would you use each?

Why they ask

Trade-off methods come up constantly in pricing and product work. They're checking you understand them well enough to choose, not just name them.

How to answer

  • MaxDiff ranks a list of items, like claims, features or names, by forcing best and worst choices
  • Conjoint tests combinations of features and price to see what people trade off
  • Use MaxDiff to prioritize a long list, conjoint to model product and pricing choices
  • Mention the tool you've run them in, such as Sawtooth or Qualtrics
10.How would you estimate the size of a market with only public data and a few days?

Why they ask

Not every question needs a survey. Some roles lean on secondary research, and they want to see you reason it out from sources you can defend.

How to answer

  • Pick a top-down or bottom-up approach and say why
  • Name the kinds of sources you'd use, like census data, trade association reports and company filings
  • Write down every assumption so someone can challenge it
  • Give a range rather than a single number and say which assumption moves it most

Part 3

11.Which tools do you use for analysis and reporting, and how deep does your use go?

Why they ask

Postings name specific tools, and they want to know whether you've driven them yourself or watched someone else do it.

How to answer

  • Name the survey platform, stats package and charting tool you've used most
  • Give a specific task for each, like building a tab banner in SPSS or a tracker in Tableau
  • Be honest about where you're a beginner and how you'd close the gap
  • Mention any automation you've built for repeat reports
12.What's a survey or research finding you've seen used badly, and what would you have done differently?

Why they ask

This shows judgment and whether you think critically about the research you consume, not only the research you produce.

How to answer

  • Pick a real example, like a customer satisfaction score driving decisions it can't support
  • Explain the specific flaw: sample, wording, timing or how it was summarized
  • Say what the better study or better framing would have been
  • Keep it professional and avoid bashing a past employer

Questions to ask them

Ask at least two. It shows you're picking them too.

  • How much of your research is run in-house versus bought from agencies and panel vendors?
  • Can you tell me about a recent study that changed a decision, and one that got ignored?
  • Who are the main internal clients for this role, and how do requests reach the team?
  • What tools and data sources would I be expected to use in my first months?
  • How does the team handle it when a stakeholder pushes back on a finding?

Related interview guides

Data AnalystBrand Manager