Over the last decade working with digital marketing, I have sat on both sides of the table. I have been the person executing campaigns while someone else decided the direction, and I have been the person who had to decide the direction while a room full of smart people waited for an answer. The second job is harder, and almost nobody prepares you for it.
What I see today is a market full of excellent marketing operators and a shortage of marketing leaders. We know how to build funnels, read dashboards, structure campaigns and prompt a model. What we struggle with is deciding what deserves attention, holding a direction when results take three months to appear, and keeping a team functional while the tooling underneath it changes every quarter.
This article is my attempt to write down what I have actually seen working. Not leadership theory, not motivational posts. The operating decisions that separate a marketing area that produces revenue from one that produces reports.
Why marketing leadership feels harder in 2026 than it did five years ago
There is a temptation to treat this as a feeling. It is not. The pressure is measurable, and understanding the numbers changes how you respond to them.
The data behind the pressure
Spencer Stuart data reported by Adweek puts the average tenure of a chief marketing officer in the S&P 500 at 4.1 years, against 5.0 years for C suite roles overall and 7.6 years for chief executives. Only the chief operating officer role turns over faster. You can read the breakdown in Adweek’s coverage of the study.
According to the Gartner 2026 CMO Spend Survey of 401 marketing leaders, budgets sit at 7.8% of company revenue, essentially flat since 2022 and roughly 18% below where they were four years ago. In the same study, 70% of CMOs say becoming an AI leader is a critical goal for the year, while only 30% report mature AI readiness. More than half say they do not have the budget to hit their own annual objectives, and 56% expect budget cuts if they miss.
Gallup’s State of the Global Workplace 2026 reports that manager engagement fell from 31% in 2022 to 22% in 2025, with a five point drop between 2024 and 2025 alone. Managers used to be meaningfully more engaged than the people they led. That gap has nearly disappeared.
Put those three findings next to each other and you get a coherent picture. Marketing leaders are being asked to deliver growth, absorb an AI transformation and protect efficiency, all inside a flat budget and a four year window to prove it. The layer that historically carried that translation work, middle management, is running on empty. That is not a motivation problem. It is a design problem.
What the data does not say
Short tenure is not automatically failure. A good share of that turnover is voluntary, with senior marketers moving into chief growth officer, chief commercial officer or general manager roles that carry explicit revenue accountability. The title changes because the mandate changed.
That is the most useful signal in the whole dataset. The market is rewarding marketing leaders who own a number in the P&L and quietly retiring the ones who own a function. If your leadership identity is built on campaigns, brand guidelines and channel expertise alone, the ceiling arrives faster than it used to.
The four jobs nobody else in the company can do for you
When I audit a marketing area that is struggling, the problem is rarely execution quality. It is usually that the leader is doing the team’s work and leaving their own work undone. There are four things that only the person in the chair can do.
1. Direction: deciding what the team will not do
Strategy gets discussed as a document. In practice it is a subtraction exercise. A marketing team with eight priorities has zero priorities, because the moment resources get tight, everyone protects their own initiative and nothing gets killed.
What I do is force a ranked list. Not tiers, not buckets, a strict order from one to ten where two things cannot share a position. Then I draw a line after the third item and tell the team that everything below the line is officially not happening this quarter. The discomfort in that room is the point. If nobody is uncomfortable, you did not actually prioritise.
2. Decision: converting data into a call
Most marketing teams do not suffer from a lack of data. They suffer from a lack of decisions made with the data they already have. I have seen operations with three dashboards, a BI tool and an attribution model that still could not answer whether a channel should get more budget next month.
The fix is boring and it works: attach every recurring report to a decision. If a report does not change a decision, it gets deleted. If a decision happens without a report, that report needs to exist.
| Report | Decision it drives | Cadence | Owner |
|---|---|---|---|
| Channel efficiency by CAC | Reallocate budget between channels | Weekly | Paid media lead |
| Pipeline by source and stage | Adjust lead quality criteria with sales | Biweekly | Marketing lead |
| Content performance by intent cluster | Kill, refresh or scale topics | Monthly | Content lead |
| Revenue attributed to marketing | Defend or renegotiate budget | Quarterly | Marketing leader |
This table is not a template to copy. It is a test. If you cannot fill the second column for every report your team produces, you have found the problem.
3. Development: building people faster than the tools change
This is the job most marketing leaders postpone, because it never appears urgent and the cost of skipping it only shows up eighteen months later when nobody on the team is ready for a bigger role.
The skill map I use
I split the team’s capabilities into two groups, and I treat them very differently.
Craft skills
Media buying, copywriting, technical SEO, analytics implementation, automation. These have a shelf life. They decay, they get partially automated, and they need refreshing every year or so. Training here is cheap, specific and continuous.
Judgment skills
Reading a business model, understanding unit economics, arguing with a sales director without losing the room, knowing when a number is too good to be true. These compound. They take years and they cannot be outsourced to a course.
One caveat about certifications
Certifications prove exposure, not judgment. I have hired people with thin portfolios and sharp thinking, and I have interviewed people with a wall of badges who could not explain why a campaign worked. Use certifications as a filter for craft skills and interviews as a filter for judgment.
4. Defense: protecting the team from noise
Every marketing team I have worked with absorbs a constant flow of unplanned requests: a last minute banner, a deck for a board meeting, a competitor campaign that someone saw on LinkedIn. Individually each request is small. Collectively they destroy the quarter.
How the defense layer works in practice
I run everything unplanned through three questions before it reaches the team. Which of the three priorities does this serve? What comes off the list to make room for it? Who is the decision maker if the answer is no? Roughly seven out of ten requests do not survive that conversation, and the person who asked usually agrees once the trade off is explicit.
The operating cadence that makes leadership visible
Leadership that only exists in one to one conversations is invisible to the rest of the company. A cadence makes it legible. This is the rhythm I use and recommend.
Weekly: execution and blockers
Forty five minutes, fixed agenda, no status theatre. What moved, what is stuck, what decision is needed from me today. Anything that requires more than ten minutes of debate gets scheduled separately. The weekly meeting exists to unblock, not to inform.
Monthly: performance and reallocation
Ninety minutes reviewing the numbers that matter against targets, with an explicit budget reallocation decision at the end. If no money moved and no priority changed, the meeting produced nothing.
Quarterly: direction and people
Half a day. The first part revisits the ranked priority list and the assumptions behind it. The second part is dedicated to people: who is growing, who is stuck, who is at risk of leaving, what the team will need six months from now. Most leaders do the first part and skip the second, and then act surprised when their best analyst resigns.
A scorecard for the leader, not the team
We measure campaigns obsessively and measure leadership not at all. These are the indicators I track on myself:
- Decision latency. Average time between a decision being requested and made. Anything above five working days is a bottleneck wearing a leadership title.
- Priority stability. How many times the top three priorities changed inside a quarter. More than once usually means the direction was never real.
- Internal promotion rate. Share of senior openings filled from inside. This is the honest measure of whether development is happening.
- Revenue influenced per person. Not a productivity whip, a capacity signal. It tells you when the team is under resourced before people burn out.
- Voluntary regretted attrition. The people you did not want to lose. One is a data point, three is a verdict on the environment.
Leading creative and analytical people in the same room
This is where most marketing leadership fails quietly. A modern marketing team mixes two populations with genuinely different operating systems, and the leader is the only translation layer between them.
The translation problem
Analytical people trust what can be measured and get uncomfortable with ambiguity. Creative people trust intuition built on pattern recognition and get defensive when a number contradicts them. Left alone, the analytical side starts calling creative work subjective and the creative side starts calling data reductive. Both are partly right, which is why the argument never resolves on its own.
What has worked for me is separating the conversation into two moments. In the divergence moment, ideas are not evaluated by projected performance. In the convergence moment, ideas are evaluated by nothing else. Mixing the two kills good ideas early and protects bad ones late.
The brief is the contract
Almost every conflict I have mediated between creative and analytical teams traces back to a vague brief. A brief that says we need something impactful for Black Friday is an invitation to disagree later.
A brief that works contains the business objective, the single metric that defines success, the audience and their current belief, the one thing the piece must communicate, the explicit restrictions and the decision maker’s name. Six fields. It takes twenty minutes to write and saves two weeks of rework.
AI changed the job description, not the job
McKinsey’s State of AI research found that high performing organisations are far more likely to have fundamentally redesigned workflows around AI rather than inserting AI into existing ones, with nearly three quarters of high performers reporting that redesign against about one quarter of everyone else. The same research points to senior leadership ownership as a distinguishing factor.
The ANA, citing BCG data, reports that 96% of CMOs say AI is driving an end to end transformation of their function while only around a third have done the structural work to support it. The ANA summary of current CMO trends covers the gap in detail.
The differentiator is not access to models. Everyone has access. The differentiator is whether a leader is willing to rebuild the workflow, which means rewriting job descriptions, changing approval paths and accepting a temporary drop in output while the team relearns how to work. That is a leadership decision with political cost, and it is exactly why most organisations stop at pilots.
What I delegate to AI and what I never delegate
My rule is simple. I delegate volume and I keep judgment.
Research synthesis, first drafts, variation generation, data cleaning, competitive monitoring, transcript summarisation and structural analysis all go to the model. Positioning decisions, budget reallocation above a certain threshold, anything that touches a client relationship, performance feedback and hiring calls stay with people. Not for romantic reasons. Because those decisions carry consequences that the person signing off has to be able to defend.
The governance layer most teams skip
If your team is using AI daily and you have no written rules, you do not have an AI strategy. You have exposure. At minimum a marketing area needs a defined list of what data can never be pasted into an external tool, a human review checkpoint for anything published under the brand, source verification for every statistic a model produces, and a named owner for each AI assisted workflow.
I learned this the uncomfortable way, reviewing a piece of content that cited a study which did not exist. Nobody acted in bad faith. The process simply had no verification step, and a plausible sentence passed through four people without anyone checking the source.
