How to interview a marketing leader as a non-marketing founder
Most non-marketing founders interview marketing leaders using the standard 47-question list. Wrong frame. You cannot evaluate craft you cannot read. Four questions you can actually score, and what to listen for in the answers.
By Stacey Tallitsch | May 14, 2026
You decided you need a head of marketing. You posted the role, ran the funnel, narrowed to four finalists. Now one of them is across the table, and they are competent in a way you cannot verify. They are using vocabulary you half-recognize. They have references that check out. They have managed teams larger than yours. You have 45 minutes to decide whether they are the operator who finally fixes your acquisition problem or the operator who burns 18 months and a $240K base before you figure out it was the wrong call.
You cannot evaluate their craft. That is the problem.
Call it the lookup problem. You ask a candidate how they think about attribution. They answer fluently. You have no way to tell whether the answer is correct, partially correct, or articulate nonsense. The standard interview question list — the 47-question kind that shows up first in every search result — was built for marketing peers evaluating marketing peers. Run it as a non-marketing founder and you are scoring on confidence and likability, not capability.
Stop running that interview.
Re-scope what you are actually evaluating
You are not hiring someone to teach you marketing. You are hiring someone to build the diagnostic loop inside your business that you cannot build yourself. A marketing leader's job is not to know all the channels. It is to figure out which 2 or 3 your business should run, in what sequence, at what cost, and how to tell from the numbers whether each is working.
What you can evaluate in 45 minutes with no marketing background comes down to four things: whether they reason under uncertainty or recite playbooks, whether they kill things or only add them, whether they understand a business they have not yet seen, and what they did the last time they were wrong.
That is the entire list. You do not need more questions. You need better questions.
This is also the wrong interview to run if you have not yet decided whether you need a marketing leader at all. The prior question — whether your next executive hire is marketing or sales — is upstream of this one. So is the question of whether your acquisition problem is actually a pricing problem. Run those diagnostics first. Then run this interview.
The four questions
These work because the answers are hard to fake. A candidate cannot pre-cache an honest answer to a story they do not have. A candidate who is reciting reveals themselves inside two minutes of follow-up. Your job in the interview is not to evaluate the answer. It is to ask three more tell me more questions after the answer, and watch what happens to the texture.
Question 1: a program you killed
Tell me about a marketing program you killed. Walk me through why you killed it and what you replaced it with.
Most candidates have launched a hundred programs and killed five. That ratio is the problem. A marketing leader who has never killed anything has either never had budget pressure or has never been willing to fight for it. You are not hiring someone to add. You are hiring someone to decide what to cut.
What to listen for: did they kill it on data or on politics. How long did they wait before they killed it. What did they replace it with, and was the replacement smaller, larger, or zero. The candidate who proudly killed a paid program and replaced it with nothing is showing you something the candidate who killed-and-replaced is not.
Question 2: the first 30 days
If you started Monday, what would you have to learn from people in this company before you proposed your first plan?
This is the diagnostic-posture question. The wrong answer is a list of marketing things — the brand, the website, the demand-gen stack. The right answer is a list of business things — how the last 10 customers actually found you, why deals stall, what the salespeople say when leads come in cold, what the founder gave up on three years ago.
A marketing leader who treats the first 30 days as learn-the-channel-mix is going to import a generic playbook. A marketing leader who treats the first 30 days as interview the salespeople, sit on three demos, read the last 50 lost-deal notes is going to build something that fits.
Push them. Ask which conversations they would prioritize first. The candidate who has been a real operator has run this play before and has an ordered list. The candidate who has only been an executor will give you a checklist they just invented in the chair across from you.
Question 3: your most expensive mistake
What is the most expensive mistake you have made as a marketing leader, and what changed in how you operate as a result?
This is the failure mode. Three things to listen for. First, do they have a story. A candidate who cannot name an expensive mistake is either lying or has never owned a budget. Second, was the mistake structural or tactical. "I ran a campaign that did not convert" is tactical. "I sat in the head-of-marketing seat for six months without a written quarterly plan and the CEO and I argued about priorities every week" is structural. The structural answer tells you they have built scar tissue around a system, not around a campaign.
Third, what specifically changed in their operating model. The candidate who learned the lesson will name a habit, a document, a meeting, or a metric they now run that they did not run before. The candidate who did not learn it will give you a vibes answer about communication.
Question 4: flat pipeline
If our pipeline went flat for two quarters with no change in spend, what would you suspect first?
This is the live-fire diagnostic. You are not asking them to solve the problem. You are asking them to show you their reasoning order. The wrong answer is "let's look at the dashboard." The right answer starts with a hypothesis. The best answer starts with two or three competing hypotheses and a way to distinguish between them in a week.
Listen for whether the first place they look is upstream or downstream of marketing. If they start with channels, they are reaching for the lookup table. If they start with the deal-flow conversation — what changed in the market, what changed in sales motion, what changed in product positioning — they are reasoning like an operator. You can teach an operator the channels. You cannot teach a channel-mechanic to think like an operator inside one engagement.
What you are listening for, underneath all four
The candidate who passes is not the one with the most fluent answer. It is the one who, when you ask the third tell me more follow-up, gets sharper rather than vaguer. Texture is the signal. Diagnostic reasoning produces more detail under pressure, not less. Rehearsed competence produces less.
Per Spencer Stuart's 2026 CMO tenure research, the average tenure of a chief marketing officer at major US advertisers sits at roughly 4 years, and it runs shorter at consumer-facing businesses. That is a long enough window for a great hire to make a generational dent in your business and a wrong hire to do generational damage. You are not interviewing for a 90-day project. You are interviewing for someone who will spend a meaningful fraction of your company's runway sitting next to you.
The other side of the table is also evaluating you. A senior marketing leader who has run this play before is, in parallel, asking themselves whether your business has the operational baseline to make them successful — clean CRM data, a sales motion that converts, founder willingness to let go, a defined ideal customer. If your content program has been spinning for a year without producing pipeline, the experienced candidate is reading that as upstream dysfunction. They are deciding whether to take your second call.
What to do before the meeting
Write the four questions on a note card. Do not memorize them. The card on the table signals seriousness. It also lets you spend the interview listening instead of trying to remember what to ask next.
Skip the standard 30-minute mutual presentation. Tell the candidate up front: "I am going to ask you four questions. Each one will take 7 to 10 minutes. I will ask follow-ups. I want detail, not summary." That framing turns the interview from a performance into a working session. The candidate built for performance will be uncomfortable. The candidate built for the work will exhale.
Do not score the answers in the room. Score them in the 30 minutes after the candidate leaves, while the texture is still in your head. Write one paragraph per question. If you cannot remember enough texture to write a paragraph, that is the score.
— Stacey Tallitsch, Stronghold CMO
About the Author
Stacey Tallitsch is the President of Stronghold CMO, a Fractional AI CMO service operating under Talisman Capital, Inc. He is a 30-year tech veteran and the author of 21 books on systems thinking, operator-grade decision-making, and personal sovereignty, with more than 30,000 students across his Udemy course catalog.