Every engagement is different, and anyone handing you a laminated 90-day playbook has not done this work recently. But after enough startups, the arc is consistent. Here is what happens, month by month, with the research linked so you can check me against it.
What does a fractional CMO do in the first 90 days?
Three phases. Days 1 to 30 are access and reality: getting into the systems and mapping what data exists, what can be trusted, and what is missing. Days 31 to 60 are truth-testing: pressure-testing the ICP and positioning against real prospect conversations and fact-checking the competitive picture. Days 61 to 90 turn the diagnosis into a motion the team can run: instrumentation, a repositioned message, priorities, and an operating rhythm. Day 90 is a credibility checkpoint, not a finished growth engine.
That last sentence carries research behind it. Egon Zehnder’s study of nearly 600 senior executives found almost 60 percent needed six months or more to reach full impact, and Index Ventures treats 90 days as a review checkpoint, with executive onboarding often running far longer. Anyone promising a transformed pipeline by day 90 is selling you something. What you should expect by day 90 is a diagnosis you can trust, visible corrections, and a plan with resources attached.
Access and reality
The first fight is not strategy. It is access. Before anyone can diagnose anything, I need to get into the systems, and the first real deliverable is a map of what data exists, what can be trusted, and what is missing. Often, most of it is missing.
- The CRM, Google Analytics, and Search Console
- Marketing automation, and any AI or LLM workspaces the team already uses
- Notion or wherever the company’s knowledge lives
- OKRs, prior plans, and every ICP or persona document that exists
- Pipeline reports, win/loss notes, and recorded prospect meetings
Access sounds administrative. It is routinely the first hurdle, and what it uncovers is the first diagnosis. I have seen an entire site review sit blocked for weeks on a single Search Console grant. I have found analytics flowing into an account nobody at the company could see, and offboarded vendors still holding admin keys. Sometimes the discovery is simpler: there are no analytics at all, and month one includes a plan to start capturing what is not there.
The pattern matches what other operators report. Asia Orangio of DemandMaven spent the first two weeks of her own fractional engagement on exactly this systems-and-information download, and Anne Raimondi’s First Round guidance recommends circulating the 90-day plan as a draft and picking early wins that are achievable, measurable, and collaborative. Mine usually come straight out of the access work: fixing broken tracking, stopping spend nobody can measure, unblocking something already in motion.
Truth-testing
Month two pressure-tests what the company believes about itself. Early-stage companies usually have a vague sense of their ICP and personas, and most of their content is busy talking about themselves. So I flip it: the customer becomes the hero of the story, and the product becomes the guide. Then I check the story against reality, analyzing real prospect meetings to confirm the pain points and fact-checking the competitive picture, because the perceived competitors are often wrong too.
Krithika Muthukumar, the first marketer at Stripe and OpenAI, calls this the core skill of early-stage marketing: the job is diagnostician, and the treatment depends entirely on what the evidence says is broken. Her advice matches my experience on one more point: do not just observe. Join something already in flight, because execution exposes how a company works faster than any listening tour.
And here is the part nobody puts in the sales deck. I am usually brought in to increase leads. What the diagnosis keeps finding is drift at the top: founders and executives who do not agree on who the customer is. Organizations do not love hearing that. It has cost me a gig before, and I would say it again, because a lead engine pointed at three different ICPs produces noise, not pipeline.
The research says surfacing this is the job. In Gartner’s survey of 125 CEOs and CFOs, only 22 percent said they had real clarity on marketing’s roles and accountabilities, and Spencer Stuart’s CMO Playbook tells incoming marketing leaders to put expectations and success measures in writing with the CEO precisely because that clarity so rarely exists.
A motion the team can run
Month three converts the diagnosis into operations: instrumentation capturing the data that was missing, the repositioned message live on the assets that matter, explicit stop, fix, and start decisions with budget attached, a first dashboard, and a weekly operating rhythm the team can run without me in the room. Fractional means the system has to outlive my hours.
The day-90 readout to the CEO answers five questions: what did we learn, what changed, what moved, what happens next, and what marketing needs from the rest of leadership. Stopping work belongs in that readout too. A first senior marketer earns trust partly by ending random acts of marketing.
For a worked example of what this produces, the 30 meetings in 60 days campaign dashboard case study walks through a real motion built inside one engagement window, from instrumentation to booked meetings.
How should you judge a fractional CMO at day 90?
On six kinds of traction, not closed revenue alone: a written mandate everyone understands, a ranked diagnosis of what constrains growth, funnel definitions and numbers you can trust, sales and marketing behaving differently toward the same customer, traceable proof of movement, and a resourced plan that makes the next six months less speculative.
A written mandate
Everyone agrees what marketing owns, what it does not, and how success gets measured.
A ranked diagnosis
Market, positioning, funnel, or data problem: named, ordered, and backed by evidence.
Numbers you can trust
Agreed funnel definitions and a baseline, with the measurement gaps stated plainly.
Teams behaving differently
Sales, marketing, and product pointed at the same customer with clear handoffs.
Traceable movement
Leaks fixed and experiments run, with results you can trace rather than take on faith.
A less speculative future
A resourced two-quarter plan, with an explicit list of what will not be done.
Closed revenue belongs on the scorecard only when existing pipeline gave the leader something to close inside the window. The executive-ramp research above is blunt about that, and any promise otherwise deserves your skepticism.
Want your first 90 days mapped?
Thirty minutes. Tell me where your go-to-market motion is stuck and you will leave with my read on what your first 30 days would target, whether or not we end up working together. If the answer is that you are not ready for this yet, I will say so.
Email Crys or check the timing first →