How to Interview a Marketing Hire When You’ve Never Done Marketing
You can evaluate an engineering candidate even if you’re rusty: read their code, watch them debug, check whether the thing they built actually runs. For the marketing hire you’re about to make, you have no equivalent. You can’t “read the code” of a campaign that happened two jobs ago at a company you’ve never heard of. And here is the uncomfortable part: interviewing well is a marketing skill. Talking persuasively about work is literally the job. The candidate who dazzles you in the room may be demonstrating the only thing they’re good at, and you have no instrument to tell the difference.
This post is that instrument: what to screen for on paper, the interview questions that separate operators from presenters (with what good and bad answers sound like), the work sample that beats every interview, reference checks that produce real signal, and a way to borrow an expert’s judgment for one hour. If you’re still deciding whether to hire at all, start with when to make your first marketing hire and generalist vs. specialist, then come back here.
Why marketing interviews mislead founders
A marketing interview is a sales conversation in which the product is the candidate. The people best at that conversation are, by definition, good at persuasion - which correlates only loosely with shipping work that produces customers. The distinction you’re trying to draw is between a marketer who ships pipeline (real leads, real revenue, things you can count) and one who ships decks (strategies, frameworks, rebrands, slideware that never touches a customer). Both interview beautifully. Only one is worth hiring at your stage.
The fix is not to become a better interrogator. It is to shift the weight of the decision away from conversation, where they have the advantage, and toward evidence, where you do. Everything below follows from that move.
What to screen for on paper
Before anyone gets an interview, the resume and portfolio can do real filtering. You are looking for three positive signals:
Evidence of shipped work with numbers they can defend. Not “supported brand growth” - actual artifacts: a campaign, a website, an email sequence, a launch, with a number attached. The number doesn’t have to be huge. It has to be specific, and in the interview they have to be able to explain exactly where it came from.
Small-company experience. Someone who spent ten years at a household-name brand ran campaigns with an agency, a budget, and specialists doing the production. At your company they’ll have none of that. Somewhere on the resume there should be a chapter where they marketed something with no budget, no brand recognition, and no team.
Writing samples. Ask for two or three things they wrote themselves - a blog post, a landing page, an email campaign. Most marketing at a small company is writing. You may not be able to judge strategy, but you can judge whether writing is clear, specific, and makes you want to keep reading. Trust that judgment.
And three red flags:
- Only big-brand logos. As above: brand-name experience without any scrappy chapter is a mismatch risk, not a credential.
- “Led” and “oversaw” everywhere, artifacts nowhere. Management verbs with nothing you can look at usually mean the candidate was near the work, not doing it.
- Buzzword density. “Drove synergies across omnichannel brand storytelling to elevate customer journeys.” If you can’t tell what the person actually did after reading the sentence twice, that is the sentence working as intended.
None of these red flags is disqualifying by itself. Each one just moves the burden of proof onto the stages below.
The interview questions that expose real operators
The goal of the interview is not to have a great conversation. It is to force specificity, because specificity is the one thing that can’t be faked by a good talker. Four questions do most of the work.
“Walk me through one campaign or project you ran end to end. What was the goal, what did you personally do, and what happened?” A real operator answers with texture: the goal in numbers, the specific things they did (wrote the emails, built the landing page, set the budget), the result, and - this is the tell - something that went wrong along the way. A presenter answers at team level: “we launched”, “we saw great engagement.” Push once: “what part did you personally do?” If the “we” never becomes “I”, the candidate was in the room, not on the tools.
“Here’s our product. Who would you market it to, and what would you try first?” You’re not grading correctness - they’ve known your company for a week. You’re grading the shape of the answer. Strong candidates ask you questions first (who buys today, why, where did your best customers come from) and then propose something small and cheap to learn from: “I’d talk to your last ten customers, then test the message that comes out of that.” Weak candidates skip the questions and go straight to a big swing: a rebrand, a paid-ads budget, “building out the brand.” A first move that spends a lot before learning anything is exactly the pattern that burns founders.
“Tell me about a marketing effort you ran that failed.” Everyone real has one. The good answer owns it in the first person, explains what they expected versus what happened, and what they changed afterward. The bad answer is a humble-brag (“we set the goal too ambitiously”), a blame shift (“sales didn’t follow up”), or - worst - a blank. A marketer who has never failed has never run anything with a number attached to it.
“How would you decide what NOT to do in your first 90 days?” Marketing has infinite possible activities and your company can afford about three. Strong candidates talk about focus: pick the one or two channels the evidence points to, park everything else, and name what they’d say no to. Weak candidates can’t name anything they’d skip - everything is “important for building the foundation.” A hire who can’t prioritize will spend two quarters being busy instead of effective. Our 90-day plan for a first marketing hire doubles as a grading rubric for this question.
Here is the whole set as a cheat sheet you can bring into the room:
| Question | A strong answer sounds like | Red flag |
|---|---|---|
| ”Walk me through a campaign you ran end to end.” | First person, specific numbers, names what they personally built, includes something that went wrong | Team-level “we” throughout; results described as “great engagement”; no number survives a follow-up |
| ”Who would you market our product to, and what would you try first?” | Asks you questions before answering; proposes a cheap, fast test tied to what your customers actually say | Jumps straight to a rebrand, a big paid budget, or “elevating the brand” with no learning step |
| ”Tell me about a marketing effort that failed.” | Owns it, explains the expectation vs. reality, names what they changed next time | Has never failed; blames another team; the “failure” is secretly a success story |
| ”What would you decide NOT to do in your first 90 days?” | Names specific things they’d park and why; talks about one or two focused bets | Everything is important; wants to “build foundations” across every channel at once |
| (Any answer, any question) | Numbers they can decompose when you ask “how was that measured?” | Numbers that dissolve under one follow-up question |
The one follow-up that works on every answer: “How did you measure that?” You don’t need marketing knowledge to use it. Real numbers have a source and a story; invented or inherited numbers fall apart on the second sentence. Ask it calmly, twice per interview, and watch what happens.
The work sample that beats every interview
If you take one thing from this post: give your finalists a paid, time-boxed work sample on your real business. This is the marketing equivalent of reading their code, and it is the single strongest predictor you have access to. Two formats work well; pick one:
The teardown. “Spend 2-4 hours. Look at our homepage and our top competitor’s. Tell us what you’d change about ours and why, and what you’d test first.”
The first-90-days one-pager. “Spend 2-4 hours. One page: what would you focus on in your first 90 days here, what would you explicitly not do, and how would we know it’s working?”
Pay them for the time, keep it genuinely time-boxed so you’re testing judgment rather than free labor or spare time, and say a rough sketch is fine. Then grade the result on things you can judge without marketing expertise:
- Did they get our customer right? You know your customer better than anyone. If the sample describes someone who would never buy from you, that’s decisive.
- Is it specific to us? A strong sample quotes your actual homepage and names your competitor’s actual moves. A weak one could have been written about any company in your category.
- Is the writing clear? Short sentences, plain claims, a reason attached to every recommendation.
- Did they prioritize? A short list of ranked bets beats a long list of everything. Bonus points for a “what I would NOT do” section they included unprompted.
- Is there a cheap first test? The best samples end with “here’s the small thing I’d try in week one to find out if I’m right.”
A candidate who aces the conversation and fumbles the work sample is exactly the hire this framework exists to prevent.
Reference checks that actually work
Standard reference calls produce nothing, because references are pre-selected to be positive and most people won’t volunteer criticism. Two changes fix most of that:
Ask the one question that forces a real answer: “Knowing what you know, would you hire them again for this specific role - first marketer at a small company, doing the work themselves with no team?” Then stop talking. An immediate “yes, in a heartbeat” is signal. A pause, a qualifier (“well, in the right structure…”), or a pivot to personality (“everyone loves working with them”) is also signal - people almost never say no outright, so hesitation is how the truth arrives.
Ask for specifics you can cross-check: “What did they personally produce that you saw with your own eyes?” If the reference describes the same artifacts and numbers the candidate claimed, good. If the reference describes a coordinator of other people’s work while the resume said “led growth,” you’ve learned what you needed.
The calibration trick: borrow one hour of expert judgment
Everything above reduces your risk, but there is a limit to evaluating a discipline you’ve never practiced. So don’t do it entirely alone. Find one experienced marketing operator - an advisor, a marketing leader from your investor’s portfolio network, a fractional CMO you pay for a single session - and bring them in for one focused hour at the finalist stage. Their job is small and specific: review the work sample, sit in on (or watch a recording of) one interview, and answer two questions: “Is this work good?” and “What would you probe that I can’t?”
You’re not outsourcing the decision. Culture, drive, and honesty are yours to judge, and you’re well equipped to judge them. You’re borrowing the one thing you can’t fake: a trained eye on the craft. One hour of that, applied to a real work sample, catches the polished-presenter hire more reliably than five founder interviews in a row.
One caution in the other direction: calibrate your own expectations before grading anyone else’s plan. A first marketer at a small company is a team of one. Modern tooling (Marqeable included) changes how much one strong operator can ship, but it doesn’t change the need to pick two or three bets and ignore the rest. A candidate who tells you “no” to half your wish list in the interview isn’t being difficult. They’re showing you the skill you’re hiring for.
Frequently asked questions
How can I evaluate a marketing hire if I’ve never done marketing?
Move the decision’s weight from conversation to evidence: screen for shipped work with defensible numbers, run a paid 2-4 hour work sample on your real business, ask references “would you hire them again for this exact role,” and have one experienced marketer review the finalist’s work sample. You judge the person; borrowed expertise judges the craft.
What’s the single best interview question for a marketing candidate?
“Walk me through one campaign you ran end to end: the goal, what you personally did, and what happened.” Then follow up with “how did you measure that?” Operators answer in the first person with specifics and at least one failure. Presenters stay at “we” and the numbers dissolve under one follow-up.
Should the work sample be paid?
Yes. Paying for a time-boxed 2-4 hours keeps it fair, works for candidates who are currently employed (usually your best ones), and makes the time-box credible so you’re testing judgment rather than who had a free weekend.
Is big-brand experience a bad sign?
Not by itself. It’s a mismatch risk: at a large company, agencies and specialist teams do the production work. Look for at least one chapter where the candidate marketed something with no budget, no brand, and no team - that’s the job they’ll be doing for you.
The bottom line
You will never out-talk a marketer in an interview, and you don’t need to. Screen for artifacts and numbers on paper. Ask the four questions and listen for “I” versus “we”, for questions before plans, for owned failures, and for what they’d skip. Pay for a small work sample and grade it on the things you already know cold: your customer, clear writing, focus. Ask references the would-you-hire-them-again question and listen for the pause. And spend one hour of a real marketer’s time on your finalist’s work. Do those five things and you can make this hire with the confidence you’d bring to reading an engineer’s code - not because you learned marketing, but because you stopped letting the interview be the test.
Marqeable runs your campaigns, answers every visitor, text, and email in seconds, and turns them into booked jobs and meetings - even at 9pm on a Saturday. We’re in private beta with a small early cohort. Get early access
