Marketing Interview Questions That Reveal Judgement, and the Ones to Drop

The one test every question has to pass

Does the question force a choice between two good options, and make the candidate defend what they gave up?

That is the whole ranking. A question passes if a smart, well-prepared candidate could still get it wrong by choosing badly or defending the choice weakly. It fails if there is a correct answer they could have looked up on the way in.

Recall used to be a fair proxy for experience. If someone could explain attribution models cleanly, you assumed they had worked with them. I would not assume that any more. In a Gartner survey of 3,290 job candidates in the last quarter of 2024, 39% said they had used AI during the application process. Of those, 29% used it to generate answers to assessment questions. Gartner also predicts that by 2028, one in four candidate profiles worldwide will be fake. That is a forecast, so I hold it loosely, but it points the same way: the conversation in the room has to carry more of the weight.

So when a question has an answer that fits in a search result, I assume it is testing the candidate’s phone.

If you are the candidate reading this, each question below comes with what the interviewer is listening for. Use that to work out the trade-off you would actually make, and why. A memorised answer to a trade-off question is easy to hear.

Six questions worth keeping, ranked

1. “Tell me about a campaign you stopped while it was still hitting its numbers. Why?”

This tests whether the candidate can see past the dashboard. Hitting the number and doing the right thing usually line up. The cases that teach you something are the ones where they split: leads that hit target and never closed, reach that grew while the brand got cheaper.

A strong answer names the number they were hitting and what it was hiding, then tells you what stopping it cost them. A rehearsed answer describes a campaign that was already failing and calls stopping it a tough call. If they have never stopped a campaign that was on target, ask what they would have needed to see before they did. That answer tells you nearly as much.

2. “Your budget is cut by a third halfway through the quarter. What do you stop first, and who do you tell?”

Here you find out whether the candidate knows which of their activities produce revenue and which produce activity, and whether they think about the people affected before they open the spreadsheet.

Listen for an order. A strong candidate cuts something specific, protects something specific, gives the reason, and names who hears first, sales or the founder. A weak answer spreads the cut evenly across everything. An even cut feels fair, and it skips the decision entirely.

3. “Which part of your work would you hand to an AI tool tomorrow, and which part would you refuse to, even if the tool were faster?”

Marketing teams are making this call right now, often without saying it out loud. The question tests whether the candidate has a view on where their own judgement adds value, which is the same as knowing what they are being paid for.

The trade-off is speed against something they have to name. A strong answer is concrete on both sides. First drafts of ad variations might go to the tool, while the call on which customer complaint becomes next quarter’s message stays with them. A rehearsed answer says AI is a co-pilot and humans bring the creativity. That sentence fits every job on earth, so it tells you nothing about this one.

4. “Sales says your leads are bad. Your dashboard says they are fine. Who is right, and what do you check first?”

If a company has a sales team and a marketing team, it has had this argument. There is no answer that keeps both sides happy, which is why I like it. It shows you whether the candidate defends their own numbers by reflex or goes looking for the place where the two definitions split.

Strong candidates ask what “good” means to each side and start from the closed deals, working backwards. They are willing to say marketing might be the one that is wrong. Rehearsed candidates talk about alignment meetings and shared SLAs and never say whose number they would trust.

5. “If you could only fund brand or performance for the next six months here, which would you pick, and what would you accept getting worse?”

The honest answer is both, and every candidate knows it. That is exactly why the question takes “both” away: it tests whether they can reason from the company’s stage to a choice, and whether they will say the cost of that choice out loud.

The pick matters less than the second half. A candidate who chooses performance and tells you what they expect to weaken on the brand side, and how they would watch it, has thought it through. A candidate who chooses one and claims nothing gets worse has not.

6. “What would you deliberately not do in your first 90 days here, even if I asked you to?”

This one is uncomfortable to ask. It tests whether the candidate will hold a line with you, in the interview, where holding it costs them something.

Strong answers are specific and polite: no rebrand before speaking to customers, no new channel before the current one is measured properly. Weak answers find nothing they would refuse. For a role where you will need them to push back on you later, I would treat that as a disqualifying answer.

Five questions to drop

Each of these once earned its place. None of them forces a choice, so none of them tells you how the candidate will decide anything.

• Definition questions. “What are the 4 Ps of marketing?” “What is the difference between SEO and SEM?” A phone answers both in seconds, so a correct answer only tells you the candidate prepared.
• “What is your greatest weakness?” The candidate has had their whole career to rehearse this one, and you will hear the rehearsal.
• “Where do you see yourself in five years?” You learn how well they prepared for interviews in general. You learn nothing about how they think about marketing.
• “Which marketing tools have you worked with?” That belongs on the CV, and the list ages quickly.
• “How do you stay updated on marketing trends?” You get a reading list. A better question is what they changed their mind about this year, and what changed it.

If this is your first marketing hire and you are still unsure whether you need one, settle the timing question first. And if you are the candidate wondering what to bring into the room besides good answers, I have covered that in the piece on whether digital marketing is still a good career.

How to score the answers

A fixed set of questions does half the job. The other half is writing down, before the first interview, what a good answer to each one sounds like.

Google’s re:Work guide to structured interviewing, updated March 2026, uses four levels: poor, borderline, solid and outstanding. Use the same four, with one line for each level, per question. For question 2, “outstanding” might read: cuts one named activity, protects one, gives the revenue reason, says who hears first. “Poor” might read: cuts evenly across the board. Write those lines before you meet anyone, while you have no favourite yet.

Three habits make the rubric hold up:

• Score each answer on your own, straight after the interview, before you talk to anyone else on the panel.
• Score what the candidate decided and why. If delivery matters for the role, note it separately so it does not quietly inflate the score.
• Ask the same questions in the same order to everyone. Follow-up questions are fine. Extra questions for some candidates and not others break the comparison.

Google’s guide says pre-made questions, guides and rubrics save an average of 40 minutes per interview, and that rejected candidates who had a structured interview were 35% happier than those who did not. Those are Google’s internal figures with no sample sizes published, so they do not count as research. The research case is the Sackett number from the top of this piece: among the methods they re-examined, structured interviews had the strongest link to later job performance.

If you also want to know which skills I think deserve the most weight once the interviews are done, that list is here.

For this week, write the rubric: six questions, four levels, one line per level. Then everyone on the panel is scoring the same thing. If one interview question has ever caught a bad hire for you, leave it in the comments along with the answer that gave the candidate away.