Findem Puts CMO Tenure at 2.6 Years. Marketing Has to Measure AI First

On September 29, Findem and CMO Huddles released what they call the largest CMO tenure study of its kind. The headline number is a 35% drop. CMOs who started in 2010 lasted a median 4.0 years in the role. The 2022 class, the most recent one with enough completed tenures to read, is at 2.6 years. The study draws on employment records for more than 13,000 sitting US CMOs at companies with 100 or more employees. Records are harder to wave off than the usual self-reported tenure poll. Sixty percent of the CMOs appointed so far in 2026 have never held the title before, and only 19% were promoted from inside.

The next day, BearingPoint published a survey of 1,050 C-suite executives and senior leaders across 13 countries, fielded in August. Only 13% of organizations with AI in place have scaled it fully in line with the original business case. Yet nearly three-quarters report a measurable top-line or bottom-line effect. The same day as Findem's study, Kantar's Media Reactions 2026 found that 57% of senior marketers feel confident they have the right media mix, down from 64% a year earlier.

Together, the three releases describe a squeeze on the person who signs the marketing budget. CMOs get less time to prove a program works, and their tools for proving anything got weaker in the same year. A first-time CMO with about 30 months can't wait for a brand program to pay back or for an AI pilot to show up in revenue. The CMOs who last will be the ones who agree on the scoreboard with finance before they spend.

Thirty months is shorter than most marketing proof cycles

The two numbers in the study are worth separating. The 2.6 years is a completed-tenure median for CMOs who started in 2022. The snapshot of sitting CMOs puts median time in the current role at 36 months against 51 for CEOs. That second figure measures time served so far, and it says nothing about when anyone leaves. The study's own methodology note calls the 2026 appointment data partial-year and directional. It also treats company-context patterns as associations only.

For this audience, the ownership split matters most. Per the Findem release, public-company CMOs average 4.0 years, PE-backed CMOs 3.1 and VC-backed CMOs 2.6. Most B2B software companies fall into the last two groups. Drew Neisser, who runs CMO Huddles, told Digiday that too many CMOs are handed "transformation-sized mandates on trial-period timelines."

Brand programs and AI rollouts both take time to show up in pipeline, and reading the result takes longer still. A CMO on a 30-month clock can run out of tenure before the program can be defended. On AI, BearingPoint found that almost three-quarters of companies either cut the original scope or got less scale than they planned. Reuters reported that fewer than a third have moved beyond pilot projects. A CMO who spends year one in pilots presents results in year two to a CFO who has heard "early signs" before.

AI's provable wins sit on the cost side of the ledger

BearingPoint’s numbers complicate the pitch that AI pays for itself. In the release BearingPoint issued through Business Wire, 24% of companies with AI in place report cost reductions of at least 10%. Only 4% report revenue or service-delivery gains of the same size. Around four in ten report some revenue growth and some cost reduction at once, so revenue gains exist on a smaller scale. For a marketing leader that's uncomfortable because a CMO is usually hired to grow the top line.

a16z sees the same gap in public companies, based on its own read of what they disclose. In the firm's State of Markets II deck, David George writes that nearly 30% of S&P 500 companies report some quantifiable AI impact. Only about 2% report any tracked metric. a16z partner Sarah Wang summed up the deck on X as 69% of the index runs a live AI deployment, and 2% disclose a metric they track over time. The deck doesn't say how either figure was compiled. Treat them as counts of what companies disclose, which may differ from what AI delivered. The gap between 69 and 2 suggests evidence is now the bottleneck, with adoption no longer the hard part.

McKinsey senior partner David Pralong made a related argument in Fortune the day BearingPoint's numbers hit the wire. In Pralong’s account, the AI wins CEOs cite most, in contact centers and coding, came from work that was already standardized and measured. Support teams tracked issues resolved per hour long before AI arrived, and engineering had test suites. Pralong cites McKinsey research showing that 37% of organizations report some positive effect on earnings from AI, while the share of high performers stayed flat at 6%. For marketing, the lesson is about sequence. Returns show up first where measurement was already in place, and most marketing workflows don't have it yet.

Kantar's data suggests marketers know this about themselves. Of the 806 senior marketers it surveyed between May and July, only half feel confident in their balance of brand and performance spending, down from 60%. Just 51% say they're getting actionable intelligence from their data. Yet a net 75% still plan to raise investment in AI assistants in 2027. Meanwhile, only 23% of the 23,452 consumers Kantar polled across 33 markets trust AI recommendations. The full Media Reactions study is a mix of marketer opinion and consumer receptivity, and it doesn’t test what any channel returned. Confidence is falling while spending rises, and neither number is tied to a result.

Tenure ticked up from its low, and some CMOs still get a full cycle

The tenure trend hasn’t fallen in a straight line. CMOs who started in 2019 hit a low of 2.3 years, and the 2022 class recovered to 2.6. MediaPost headlined the same study as CMO tenure inching up. A third of sitting CMOs have been in the role five years or more. The public-company average of 4.0 years is long enough for a full brand cycle.

That weakens the case for reading the study as a crisis, though it leaves the operating problem intact. The CMOs who now get the least time work at venture-backed companies. The same companies are most likely to be asked for an AI story in the next board deck. A cohort that gained four months from its low point still sits well inside the window where a slow program becomes someone else's result.

The BearingPoint figure needs the same care. Its "nearly three-quarters" is self-reported by executives, many of whom approved the projects. The survey covers AI across the whole business, beyond marketing. The 4% revenue figure counts companies above a 10% threshold, so it doesn’t show that AI never grows revenue. A narrower point still holds. Executives can name cost savings far more easily than revenue gains, and a CMO walks into that asymmetry at every budget review.

What this means for CMOs and the vendors who sell to them

Agree on the scoreboard before the program starts. Before signing an AI or creator budget, settle with finance the one metric that will judge it. Agree on how it'll be tracked and when it will be read. BearingPoint's most mature group, which it calls Leaders, shows why. Of the Leaders, 70% tie more than half their AI projects to financial KPIs, against 34% of everyone else. And 47% of them scale as planned, compared with 6% of the rest.

Negotiate a first scorecard inside 12 months. A first-time CMO facing a 30-month median shouldn't accept success criteria that can't be read until month 24. Split long programs into a leading indicator the board will accept at month nine, then the revenue result that follows.

Start AI where a baseline already exists. Pralong's argument transfers directly. Paid-media bidding and content production already track numbers like cost per qualified meeting and cost per asset. Brand and AI-search visibility programs mostly don't. Put the first AI dollars where you can show a before-and-after. Then use those wins to pay for the instrumentation the harder programs need.

Vendors selling to marketing should plan around the same math. The median sitting CMO has three years in the seat and most new appointees are new to the title. So the champion who signs a two-year contract may not be there for the renewal. A vendor that hands over a measurement plan with the contract gives the next CMO a reason to keep it.

My bet is that Kantar's media-mix confidence number falls again in its next study, even as AI-assistant spending rises. If those lines keep diverging, the first cuts in the following budget review will hit AI and creator programs that launched without an agreed metric. A new CMO will be the one explaining why.

CMOs get less time to prove a program works, and their tools for proving anything got weaker in the same year.

Before signing an AI or creator budget, settle with finance the one metric that will judge it, how it will be tracked, and when it will be read, with a first scorecard due inside 12 months.

By Business of GTM