The first 90 days as a new CMO are not for launching campaigns. They are for mapping what nobody will describe to you: a marketing technology stack running largely out of control, and an invisible operational power structure that decides who really owns data and reporting. Getting that mapping done first is what separates a solid start from one that collapses by month four.
- A new CMO does not inherit a department. They inherit a system.
- Onboarding plans assume conditions that no longer exist.
- The Shadow Org controls more than the official org chart.
- The first 90 days are for diagnosis, not execution.
- Mapping systems and operational power before any campaign is the real priority.
What you find when you arrive
You open the dashboard in your first week. You see four different metrics measuring the same thing in four different ways. You ask the team which is the source of truth. Nobody knows for certain. That same gap, a number that looks reliable but answers a different question, is what makes attributed pipeline a weak proxy for revenue contribution.
The report you were shown during the hiring process does not match what you see in HubSpot. HubSpot does not match Salesforce. Salesforce does not match the summary Finance sends the CEO every week.
Between 50% and 70% of executives fail within the first 18 months.[2] The documented cause is not lack of vision or capability: it is lack of role clarity and poor onboarding.
And they do so across systems that nobody fully controls.
It is like arriving at a building where every floor has its own electrical system and nobody has the full plan.
Onboarding plans for chief marketing officers have been built on the same assumption for decades: that the new leader arrives in an environment with reliable data, coherent systems and clear ownership. That environment is no longer the default.
What a new CMO typically finds are the residues of three or four years of accelerated AI and automation adoption without any underlying architecture. Specifically:
- Tools procured by different departments that do not communicate with one another.
- Automations that nobody knows who configured or what triggers they run on.
- Platform vendors with active access to accounts that should be under internal control.
- A reporting layer that mixes data from multiple sources into a dashboard that looks coherent but is not.
This is AI sprawl: multiple automation tools adopted without coherence or central ownership. It is not minor technical debt. It is ownership debt. And on top of that debt sits something harder to see: an operational power structure that does not appear on any org chart.
The Shadow Org nobody introduces you to
RevOps owns the reporting. Sales owns the CRM. Finance owns attribution. IT owns integrations.
The Shadow Org is the operational power structure that does not appear on the org chart: who really controls the systems, the data and the reporting, regardless of who is named as responsible on paper. In many organisations, marketing has the budget, the headcount and the mandate. But real control over the systems that determine what marketing does, how it is measured and what can be changed is distributed across five teams with their own objectives, metrics and incentives.
The list is more concrete than it sounds:
| Team | What they control in your stack |
|---|---|
| RevOps | Attribution model. Changing it requires their approval, their time and their buy-in. |
| Sales | CRM and undocumented data conventions that marketing cannot override. |
| Finance | Their own reading of your numbers, shared directly with the CEO before you can contextualise it. |
| IT | Platform integrations, on a priority queue where marketing does not always come first. |
| Product | Customer behaviour data, shared at the granularity they decide. |
Most CMOs think they are inheriting a team. In practice they are inheriting political agreements that have been codified in software. Reyes Brusola, CMO
The problem is not that these teams are obstructive. The problem is that the new CMO arrives believing they have authority over decisions that in practice require coalition. That gap between formal mandate and real operational capacity is what destroys well-crafted 90-day plans.
The role that nobody defines properly
Many organisations hire an ‘AI-growth CMO’ without defining what that actually means in practice. It is one of the archetypes replacing the classic CMO, and the least defined of them.
There is a pattern that repeats. The company wants to modernise its marketing, adopt AI, improve the pipeline. They hire someone with exactly that experience. They give them a title, a budget and a team.
What they do not do is define the operational mandate. Questions that should be answered before day one and rarely are:
- Does the CMO have authority over the CRM, or can they only submit requests to Sales?
- Can they modify the attribution model, or does that need to be negotiated with RevOps?
- Do they have access to product data, or is that Product’s territory?
- Can they make decisions about the marketing technology stack, or does every change go through a committee with IT?
These are not details. They are the work.
Week three tends to be when this becomes concrete. The CMO proposes a change to the attribution model, something reasonable and data-driven. RevOps says they will look into it. Sales says it needs to align with their CRM conventions. IT says a ticket needs to be raised. Three weeks later the change is still in the queue. And the CMO understands that ‘having authority over marketing’ and ‘being able to execute in marketing’ are two different things.
First 30 days: listen. Next 30: define strategy. Final 30: execute.
30 days understanding which data is reliable. 30 days negotiating access. 30 days executing with half the resources.
The result is that the CMO arrives with a plan built on assumptions that nobody has validated. In the best case, those assumptions are broadly correct and the plan survives. In the more frequent case, the CMO discovers in month two or three that the systems they need to execute are under someone else’s control, with different priorities.
The conversation about the mandate should have happened before accepting the role.
Before you accept a CMO role, ask for a two-hour session with RevOps, Sales and IT. Not to impress them. To find out whether the mandate you are being offered has operational backing. If access is refused or delayed, that answer tells you more than any offer letter.
What to do in the first 90 days
Do not launch campaigns. Do not change the positioning. Do not reorganise the team. Map first.
The most useful instruction a new chief marketing officer can receive in their first 90 days is not about strategy. It is about sequence.
Before defining anything, there are three mappings that determine whether the rest of the work has a foundation or not:
| Mapping | What it reveals |
|---|---|
| Systems | Real state of the stack: which tools are active, who configured them, where data lives. |
| Operational power | Who actually decides: CRM, attribution, reporting, integrations, formally or informally. |
| Definition of marketing | Distance between what the CEO, Sales and CFO each mean by ‘marketing’. |
Mapping 1: the systems and their real state.
Which tools are active? Who has access to each? Which automations are running and who configured them? Where does customer data live and in what format? Which integrations work and which are broken or partially broken?
This mapping is not done by asking the team for a presentation. It is done by looking at the systems directly, or by asking someone technical to audit them. A common pattern: the CMO discovers in this process that there are three different instances of the same platform running in parallel, each procured by a different department in different years, none deprecated, all sending data to different destinations. Nobody had documented it because nobody had asked. The signal that the mapping is complete: you can explain in one page which tool produces which piece of data and who has the authority to change it.
Mapping 2: real operational power.
Who makes decisions about the CRM, attribution, reporting, integrations? Are those decisions formally assigned or informally acquired territory? What coalitions does each change require?
This mapping is not political in a pejorative sense. It is a reading of the company’s real decision architecture. The signal that it is complete: you can predict who will block what before you propose it.
Mapping 3: the implicit definition of marketing.
What does the CEO mean by marketing? What does the VP of Sales mean by it? What does the CFO mean? The three answers will differ. The distance between those three definitions is the actual margin in which the CMO can operate without constant friction.
If the CEO thinks marketing means brand awareness and the VP of Sales thinks marketing should deliver qualified leads for the coming month, no strategy will close that gap without an explicit conversation first.
- Launching campaigns before validating the stack. The most frequent result: outcomes that cannot be attributed correctly, discrepancies with Sales that damage the relationship in the first quarter, and a CMO defending metrics without solid methodological support.
- Changing the positioning in the first 60 days. The existing positioning exists for reasons, even if they are not good reasons. Changing it before understanding those reasons generates internal resistance that outlasts any improvement.
- Reorganising the team without a dependency map. Every member of the team has relationships with external stakeholders that do not appear on the org chart. Moving people without that map breaks relationships that are difficult to rebuild.
If you are joining a company with fewer than 15 people, a single CRM and no existing marketing automation, the mapping phase is a two-week exercise, not 90 days. The framework also changes when the company is in an acute crisis: revenue in freefall or a major customer at risk. In those cases the sequence compresses and some decisions cannot wait for a complete diagnosis. Know which situation you are walking into before you arrive.
There is one final thing the first 90 days reveal: whether the organisation has clarity about what it is actually asking marketing to do. In companies where that clarity exists, the 90 days are hard but have direction.
Before launching the first campaign, complete the three mappings and have a direct conversation with the owners of each system. It is considerably easier to have that conversation before you have committed to anything.
Frequently asked
What separates a solid CMO start from one that collapses by month four
The difference is almost always about sequence, not the quality of the plan. CMOs who start well complete the three mappings first: systems, real operational power and the implicit definition of marketing. Those who begin executing without those mappings discover in month three that their commitments were built on assumptions nobody validated.
Why 90-day plans fail when the organisation has AI sprawl
AI sprawl is not just technical debt. It is ownership debt. When nobody knows who configured a given automation or what data it is processing, the CMO cannot measure the real impact of their actions. Executing on top of invisible automations produces results that cannot be attributed, defended or systematically improved.
How to identify the Shadow Org before you begin executing
Four questions reveal the real operational power structure within the first 15 days: Who approves changes to the CRM? Who built the attribution model and who has the authority to modify it? Who receives the marketing report directly without going through the CMO? Which team has to agree before marketing can launch a new integration?
Sources: [1] Gallup, ‘State of the Global Workplace Report’, cited in Chronus, ‘Executive Onboarding Plan’, 2025. [2] Chronus, ‘Executive Onboarding Plan’, 2025. Compilation of studies on executive failure in the first 18 months. [3] Spencer Stuart, sector reference on average CMO tenure at Fortune 500, spencerstuart.com.
