A new report says median tenure has fallen 35% since 2010. The data support a narrower, more useful conclusion about mandates.
The headline number in a new study of chief marketing officers is designed to alarm: median tenure has fallen by 35% since 2010, from 4.0 years to 2.6 years. Findem, a talent-data company, and CMO Huddles, a peer network for B2B marketers, say it is the largest such study ever done. Drew Neisser, the CEO of CMO Huddles, called the findings “a fire alarm” for companies.
The number is real. What it means is more complicated, and the more complicated the reading, the more useful it is.
What the 35% does and doesn’t say
The figure comes from a cohort analysis that groups CMOs by year of start and calculates each group’s median stay. It is the study’s cleanest result, and the authors were careful to withhold 2023 and 2024 medians because too many of those executives are still in their jobs.
Read the accompanying chart closely, though, and the story is not one of steady decline. Median tenure appears to fall sharply in the first few cohorts, to about three years by 2012, and then drifts in a narrow band of roughly 2.3 to 3.2 years through 2022. The 35% drop is real, but it looks more like an early reset than a continuing slide. Companies hoping to justify patience with a “crisis” narrative should notice that the crisis, if it is one, is more than a decade old.
There is also a measurement question that the report does not address. Career records for people who started jobs in 2010 are typically thinner and less complete than records for recent hires. The study does not say how it defined a CMO, how it handled fractional and interim roles, or whether the quality of the underlying data changed over time. None of that invalidates the finding. All of it is worth asking about before repeating the number.
Two clocks, not one
The study’s second analysis is a snapshot of more than 13,000 CMOs now in the job at U.S. companies with 100 or more employees. Their median stay so far is about 36 months. That is a different measure from the cohort medians, and the two should not be compared directly. A snapshot of sitting executives counts tenure to date, not completed stays.
The distribution is the more interesting part. About 30% of current CMOs have been in the role for under 18 months, while 33% have served five years or more. The job is not uniformly short. It is polarized, with large groups of newcomers and long-timers.
The report also notes that CEOs, with a median of 51 months, stay about 30% longer than CMOs. That is a useful yardstick, but a modest one. A four-year median for the top job is hardly a picture of stability.
Also Read: In Agentic Advertising, the Bottleneck Is Data, Not AI
Context beats individual performance
The study’s strongest contribution is demonstrating that tenure varies across companies. Average current tenure is 2.6 years at venture-backed companies, 3.1 years at private equity-backed companies, and 4.0 years at public companies. Median tenure rises with size, from 31 months at companies with 501 to 1,000 employees to 41 months at those with more than 10,000. Software and professional services sit at the short end of the industry. Telecommunications, energy, health care, financial services, and manufacturing sit at the long end.
Caution is warranted here, too. Ownership, size and industry overlap: large, public, regulated companies differ from young, venture-backed software firms in many ways beyond who owns them. The authors call the industry gap “a pattern” rather than a causal claim, and the same restraint applies to ownership. It is also notable that the ownership figures are averages, while the other comparisons are medians. The report does find that the business model barely matters, with B2B and B2C medians of roughly 34 and 36 months.
The pattern suggests that tenure tracks how much a company’s expectations shift. A venture-backed company that raises money, changes strategy or misses a growth target may rewrite the CMO’s job midstream. As one investor quoted in the report put it, “growth moves the goalpost.”
A job with a widening brief
The most telling data may be about who is being hired. Among CMOs appointed in 2026, 60% are first-time CMOs, and only 19% were promoted from within. Fractional and interim appointments rose from 2.9% of new appointments in 2020 to 7.8% in partial-year 2026 data. Recent hires show more brand, demand generation and strategy experience than 2022 hires did, and less sales experience. AI-related skills appear in 26% of 2026 appointees, though the study does not define the term.
Taken together, those figures describe companies hiring first-timers for an expanding job on short timelines. That is the plausible mechanism behind a short tenure, but the study cannot prove it. It measures how long CMOs stay, not why they leave.
The report also flags two pipeline concerns. Women hold 55% of CMO seats, against 62% of the marketing workforce. Asian, Black and Latino representation among new CMOs was 16% in partial-year 2026 data, down from 22% in 2020. The authors say the 2026 figure should be read directionally, and it should.
Also Read: Delight.ai’s John Kim on Building an AI That Remembers
What to do with it
The report’s recommendation is sensible and largely independent of the statistics: define the mandate before hiring, agree on what success looks like in the first 18 months, and give strategy time to compound. For CMOs, it advises testing that mandate before accepting an offer.
Readers should weigh one more fact. Findem sells talent software, and CMO Huddles serves the executives in question. Both are interested in a story that says the job is precarious and that better hiring and better support are the remedy. That does not make the findings wrong. It does mean the sponsors’ framing deserves the same scrutiny as any other.
The fairest conclusion is a modest one. The CMO seat turns over faster than the CEO seat. Turnover varies widely by company. The people entering it are less experienced and are asked to do more. Whether that is a crisis or simply a hard job with unclear expectations is a question this data raises but cannot settle.
