A father reads a bedtime story.
Stories are powerful things. And for 18 years, Spencer Stuart, the blue-chip executive search giant, has released its CMO Tenure Study, which has become a de facto State of the Union of chief marketing officer job security and, thus implicitly, a story about CMO performance.
And stories can beget more stories. In this case, because bad news travels fast and the survey’s annual results almost always are, they generate a rash of earned-media headlines like:
· Average Tenure of CMOs Falls Again (WSJ)
· CMOs’ Average Tenure Drops to Lowest In Decade (MediaPost)
· Gap between average CMO and CEO tenure widens (Marketing Week)
· CMO Tenure Slips to Just 40 Months—The Lowest Average (The Drum)
· Why CMOs Are Failing, And Who Can Save Them (Forbes)
While Spencer Stuart is not alone in seeking to define the current state of the CMO each year, its report has become the common reference, and its bleak data has become a default narrative for the CMO position—and many of us in marketing and media help perpetuate it as the sorry state of affairs. Me included.
But the thing is, it may very well be wrong or at least a flawed narrative based on barely sufficient data, offered without context, analysis or insight. We don’t know, for instance, how the tenures in question ended. Who left for another CMO job? Who got promoted? Did anyone retire? Instead, each change in the CMO’s chair is left to be interpreted as another “Oops, there we go again” patch in a uniform quilt of failure. The CMO’s job is hard enough without this.
It was Andrea Brimmer, CMO of Ally Financial for the past seven years, who first pointed this out to Forbes, sharing that “for almost two decades, CMOs—and importantly, the rest of the C-suite—have taken this (CMO tenure) stat to heart, internalizing the notion that they’re practically working on borrowed time from day one. It’s time for a different conversation about the role of the CMO and the real impact of tenure across our industry.”
Brimmer is right. Their most recent report, released in April, was my first since joining Forbes in February, and the first time I looked beyond the headlines, to the survey’s methodology—to the extent one was even presented. It turns out that their conclusions were based on a sample size of only the Top 100 U.S. advertisers—barely, as I commented on my LinkedIn page at the time, a statistically relevant number and perhaps more saliently absolutely not a representative sample of an industry as diffuse as marketing.
Data samples are supposed to be representative of the bigger universe, an accurate if miniature replica of something larger. While the Top 100 advertisers may represent the universe from which Spencer Stuart wants to source its clients, it does not reflect the broader CMO/marketing community—and yet we’ve allowed it to do exactly this, becoming a fundamental part of the CMO story and marketing’s zeitgeist.
Inspired by Brimmer’s observation, I called Greg Welch, who is a core member of the CMO practice at Spencer Stuart, and coauthor of the most recent report, to better understand the methodology behind their longstanding tenure story. It is, to quote 1992 Presidential Candidate Ross Perot, fuzzy math. Let’s break it down.
Beyond the sample size being something other than a representative one—quantitatively or qualitatively—the survey is also methodologically flawed. As Welch told Forbes, Spencer Stuart takes “a snapshot of how long each person has been in their chair on December 31” and then divides this cumulative number by 100. There are some inherent biases and flaws in this. Consider:
1. Empty CMO chairs, and this past year Welch told us there were four (and at least another one or two among the top U.S. advertisers who don’t have CMOs) are scored as zeros, and rather than being excluded from the data set they are averaged into it at nil. In other words, in 2021, there were at least four zeros amongst the universe of 100. That’s 4% of the total scored as zeros when, again, they should more reasonably have been removed from the consideration set.
2. Welch also told Forbes that the numbers in this year’s report reflect 17 CMO “transitions.” So, when the report suggests that “external hires increased to 45% in 2021 from 37% in 2020,” what that means is about eight CMOs hired in 2021 were hired from outside the organization. Eight.
By taking eight and presenting it as 45%, the survey does what John Gerzema, CEO of The Harris Poll, a global market research firm, described to Forbes as “quanting the qualitative”—framing something that is at best directional (qualitative) as if it were demonstrable—explaining that “once you start slicing up a sample of this size, your statistical accuracy starts to become unreliable.”
3. While acknowledging that measurements are always as of a certain moment, by taking “a snapshot as of December 31” the firm is more accurately calculating tenure in progress and motion, not actually tenure, even among the 100.
Because the CMO of Acme Widget Co. has been in her position for 28 months as of 12/31, that’s the number that gets defined as “tenure”—irrespective of the fact that she’s going to stay in that chair for another three years, which is essentially like calling tonight’s NBA Finals game as over in the second or third quarter. Putting aside the absence of context about why the 17 moves took place, a more useful and insightful “snapshot” might be one that considers the tenure of those who left their positions and, separately, the tenure of those who did not. But 17 (as was the case this past year) isn’t much of a sample size.
Brimmer reinforced this macro point when she told Forbes, “Some of the world’s most admired companies, P&G, Google and Adobe have long-tenured CMOs who have been driving shareholder value and paving the path forward for their companies. If you look at the 2021 Forbes’ World’s Most Influential CMOs List more than 40% have been with the same company for more than four years, above the benchmark stat. And among the world’s most admired companies? Their CMOs have an average tenure of nearly 50 months and counting.
Marketers know that data is frequently manipulated to tell the story we want it to if not exactly the story it might more accurately represent, but this “snapshot” offered by Spencer Stuart seems more than a bit like a face-tuned profile picture on a dating-app. While there is some truth to it, the picture neither fully nor accurately represents reality.
As Brimmer added, “We’ve allowed the narrative to define the outcome and we need to stop; as an industry, we need to change the conversation.” And she’s right. Presented as a data-driven verdict on CMO tenure from a leading and authoritative industry voice, it turns out this particular bit of research isn’t really representative of the population it purports to reflect, and that it might best be seen as a creative interpretation of data that tells one story when there is more than one to be told.
But that’s just marketing, isn’t it?
