We’ve Forgotten How to Hire a Marketing Leader

Summary

What the case-study assignments, the AI knockout requirement and the questions nobody asks reveal about the companies running these searches, and what CEOs and boards should do instead. Unpaid marketing-plan assignments and vague "AI-savvy" checkboxes are both proxies for judgment that companies use because they've stopped trusting themselves to evaluate it directly, and lays out what to ask instead — of candidates and of the board running the search.

A marketing leader I know recently spent weeks building a full 12-month marketing plan for a final-round interview. Building blocks, value proposition, messaging, integrated campaigns, budget, quarterly allocations, a dashboard, the sales and marketing alignment process. All of it, free.

The company came back and said they didn’t fully understand what they were reading.

Then the recruiter asked her to leave the presentation behind, because the company had a few clarifying questions.

They couldn’t evaluate the plan and they wanted to keep it anyway. All that work, done for free, and for what?

If you can’t assess the work, the work was never the point.

This has been building for a long time

I’ve been hearing versions of this from B2B marketing leaders for years now, and it hasn’t gotten better.

One candidate was asked for a case study, two brand narratives, partnership recommendations, a content strategy, a PR and thought leadership plan and the metrics behind all of it. Three to four days of research, submitted as a PDF. She was a finalist and she didn’t get the role.

Another was handed an assignment covering the competitive picture, an integrated strategy across demand gen, product marketing and communications, next-year targets, headcount investments, program spend, tooling and the operating cadence to manage it. Forty-five minutes of prepared content, please. She found out later that company used her recommendations verbatim after passing on her.

A third was told the assignment was a formality, that the job was effectively hers. Four days of work later she got a stall, then an offer for a role two levels down. She found out afterward that the identical thing had happened to another candidate at the same company six months earlier.

Another CMO candidate lost the role, then saw pieces of her own work show up in that company’s presentation at an industry event.

When I talk to CMOs and hear these stories directly, I almost always feel a sense of exhaustion and outrage.

We are way past the point of diminishing returns

A 30-60-90 plan outline with some initial observations is a reasonable ask. A full marketing plan or a budget is beyond it.

That line is not complicated and it costs a company nothing to hold. Most opportunities at this level involve some kind of exercise, and candidly, that’s fine. There’s a real difference between asking someone how they’d think about your first quarter and asking them to produce the deliverable you’d otherwise pay a consultant $30-$60K to build.

The assignment doesn’t even work anyway!

Set the fairness argument aside for a minute, because there’s a more practical problem. The homework assignment is a bad evaluation instrument.

It’s built on missing inputs. The candidate is constructing a plan without your data, your customer conversations, your sales team’s real objections or any honest account of what’s already been tried. Whatever they hand you can be picked apart, because it was never possible to get right. As one marketing leader put it, no matter what you present, it’ll get holes poked in it. So you learn very little about their judgement and quite a lot about their willingness to comply.

It also quietly selects for the wrong thing. The candidate who can spend four days on your assignment is the one with four days available. That biases the process toward people who aren’t currently employed and away from parents, caregivers and anyone already running a team.

Your best candidate is very often the one who says no to that much work, and you will read that no as disinterest or a lack of ambition/hunger for the role. That’s simply inaccurate and unfair.

Though there are clearly exceptions, I’d argue many companies doing this aren’t trying to steal anything. They’re just scared.

They’ve been burned by a bad senior hire, they don’t trust their own instincts and asking for more work feels like buying more certainty. It isn’t.

A deliverable built without context tells you almost nothing about how someone will operate with context.

Meanwhile, the other half of the job description went missing

A CMO shared recently that whether he’s talking to a prospective client about a fractional role or to a recruiter about a full-time one, he cannot remember the last time anyone asked about his approach to developing and mentoring young marketing talent. Which, as he put it, is the part of the job he enjoys most. That entire section of the job description has been replaced by a requirement for (poorly understood) AI capabilities.

Two marketing job descriptions came into my inbox in the same 48 hours a couple weeks ago. One says “if you aren’t leveraging AI to do this you are not the right fit.” The other lists “AI-savvy” as a must-have, on its own line.

To be clear, I’m not arguing against the AI requirement. Marketers need to build that muscle before it becomes a requirement, and it is now a requirement.

But somehow (perhaps unintentionally) it’s replaced “how do you build people” as a leadership priority. The leaders who were good at developing people before AI are the same ones who’ll get real value out of AI on a team, because both jobs run on the same underlying skill.

What both moves have in common

The case-study assignment and the AI checkbox look like opposites, one asking for far too much and the other for far too little.

They’re the same failure. Both are proxies for judgement, adopted by companies that have quietly stopped believing they can evaluate judgement directly.

One hands you a deliverable you can hold and the other a capability you can verify on a résumé. Neither measures whether this person can walk into an ambiguous situation, figure out which three things matter, say no to the other ninety-seven and bring a team along with them. That’s the actual job, and it’s the one thing neither proxy touches.

We’ve hired a lot of marketers at Heinz Marketing over the years and gotten it wrong too. Every miss I can remember came from the same place – I evaluated what someone had produced instead of how they decided what to produce. I overindexed on their past vs their capacity and potential.

Back to basics

Someone suggested a while back that marketing needed an interview bill of rights. Nobody ever wrote it. So here’s my attempt at the hiring side, aimed squarely at the CEOs and founders running these searches, as well as the recruiters helping them.

Decide what you’re hiring for before you open the search. If your executive team can’t agree, in one sentence, on what marketing has to fix in the next year, no candidate is going to fix it for you. A whole lot of case-study theater is a company outsourcing its own strategy homework to applicants and calling it diligence. Do that work first, then hire against it.

Use the evidence that already exists. Before you ask anyone to build something new, ask them to walk you through something they already built. Pick a number they owned and go deep: where it started, where it ended, what they personally changed and what they’d do differently. Ask for the artifacts from that work, redacted as needed. This costs the candidate nothing new and it tells you far more than a plan constructed without your data.

If you need to watch them think, do it live and do it together. A 30-60-90 with initial observations, presented in a working session where you push back and they push back and you both change your minds a little. Ninety minutes. You’ll learn how they handle being challenged, whether they ask better questions than you expected and whether they can say “I don’t know yet, here’s how I’d find out.” That’s the take you actually want.

Ask about people, and then check the answer. “Tell me about someone you developed. Where are they now?” It’s a simple question and it’s completely verifiable. Then call that person. References from people who reported to a candidate beat references from people the candidate reported to, nearly every time, and almost nobody does the work to get them.

Test for the negative. “What would you say no to in your first ninety days?” and “What’s the last thing you were badly wrong about, and what changed your mind?” Marketing leaders fail far more often on what they allowed than on what they lacked. Anyone who can’t name something they’d kill hasn’t led a function under real constraints.

Make the AI question about evidence, not fluency. Don’t ask whether they use AI. Ask them to show you something they built with it and walk you through where the output was wrong and how they caught it. One marketing leader I know does exactly this with every hire on her team, and it’s a much better filter than the knockout bullet – it surfaces judgement and tool fluency in the same answer.

Watch what the candidate asks you. The questions a senior marketer asks about your pipeline math, your sales relationship, your last two failed launches and what happened to the person in the seat before them are a better signal than anything they’ll present on a slide. If they don’t ask hard questions, they aren’t going to ask hard questions internally either.

What the board owns

Three things, and none of them happen in the interview room.

Set the runway before the hire and write it down. Marketing leadership tenures can be short, often because expectations were never made explicit and quarter two got uncomfortable. Agree in advance on what year one success looks like, which leading indicators you’ll watch on the way there and what you’ll do when an early number disappoints.

Own the scorecard, and make it a real one. Pipeline contribution, cost per qualified opportunity, sales cycle movement, retention on the marketing team. If the board can’t name the measures before the search, the new leader will spend their first two quarters inventing them and defending them instead of building.

And take a look at the process itself. Boards review sales comp plans and customer churn without blinking, then never once ask how the executive search was run. The process is the first honest look a candidate gets at how your company operates, and it’s the one part of the hire you can inspect before the money is spent.

The candidates you most want are the ones with enough standing to turn down a four-day assignment. Your process is quietly selecting against them, and you’ll never see it happen, because that no looks exactly like a candidate who wasn’t that interested.

Run the process you’d be willing to sit through yourself.

This post originally appeared on Matt Heinz’s Substack.