100 Directors, VPs, and Heads of Marketing at mid-market and enterprise SaaS companies on what they actually do with AI, what's working, what isn't, and what nobody's writing about.
Conducted via Wynter's panel of 80,000+ verified B2B professionals. Results in under 48 hours.
Run a study like this →We surveyed 100 Directors, VPs, and Heads of Marketing at B2B SaaS companies to find out what they actually do with AI, beyond conference keynotes and LinkedIn hot takes. The findings cut against both the doomer and booster narratives. The cuts have already happened, but most don't look like cuts. Senior marketers feel safe while quietly cutting the seat below them. Shadow AI is the default operating model. Most companies bought licenses and called it a strategy. And leadership is bimodal: half thinks AI is magic, the other half is moving too slow. The common thread: leaders aren't using AI themselves.
47% of B2B companies have eliminated, reduced, or stopped backfilling marketing roles due to AI in the last 12 months. Only 7% had visible team shrinkage. The other 40% happened through attrition and unfilled reqs. As one respondent put it: most leaders don't even call this an AI decision. They call it "we got more efficient." Same outcome, fewer headlines. Mid-market (52%) is cutting harder than enterprise (40%).
94% of marketing leaders think their own role survives the next 24 months. 60% name content and copywriting as the role getting cut. 20% explicitly name junior and entry-level positions. One Sr Director of PMM at an enterprise SaaS company: "My PMM team are nearly all Directors. The execution is much easier with AI, so I just need senior folks who can build trust." The job isn't getting eliminated. The path to the job is.
41% of marketers openly admit using personal accounts to bypass corporate AI restrictions. 90% of their companies have at least partial restrictions in place. One Director of PMM at a mid-market SaaS company: "We are imposed to use Copilot, and because it's crap we pretty much all use the paid version of Claude. We get what we need from it and export our work." You can't police this. Marketers will pay for Claude personally before they wait six months for IT to evaluate it.
53% of companies say AI hasn't delivered ROI. The 23% who got ROI share a consistent playbook: strategy first, tight procurement, start with the stuck backlog, actually build the workflows. Claude at $25 per seat is essentially free for any B2B company. The tool isn't the bottleneck. The bottleneck is the implementation work. As one Director of PMM put it: "AI is just an accelerant of your strategy. Garbage in, garbage out."
22% of marketers say leadership thinks AI is magic. 15% say leadership is too slow. Only 14% say leadership has a realistic read. The unifying pattern across both camps: leaders aren't using AI themselves. One Head of PMM at an enterprise SaaS company: "Everyone is using AI and we need to move way faster. Our competitive moats will wither away." If you lead marketing and you're not personally building with AI, you're guessing. Four hours shipping something real beats any vendor pitch.
26-page PDF with all 10 questions, charts, verbatim quotes, plus three unreported workplace shifts: surveillance-based AI mandates, the erosion of job satisfaction among heavy AI users, and the normalization of a two-tier AI stack.
PDF delivered instantly. No spam.
All responses were collected via Wynter's Target Customer Survey product. We asked 10 questions designed to surface what marketing leaders actually do with AI rather than what they say in conference keynotes. Four were structured multiple-choice; six were open-ended. The focus was behavior, not opinion: "Has your company eliminated roles?" not "Do you think AI will eliminate roles?"
Sample: 100 verified B2B marketing leaders from Wynter's professional research panel, all Directors, VPs, or Heads of Marketing at mid-market (200-999 employees, n=58) or enterprise (1,000+ employees, n=42) SaaS companies. Job titles skewed toward Product Marketing, with a meaningful subset of Brand, Demand Gen, and CMO-track leaders. Survey ran May 18-22, 2026. All data is self-reported. Thematic coding of open-ended responses with explicit attention to disconfirming evidence.
Wynter's panel of 80,000+ verified B2B professionals delivers research like this in under 48 hours. Message tests, ICP surveys, brand tracking, preference tests.
Each stat has its own link. Please cite Wynter and link to this page when you use them.
Every question from the study, free to read and cite. Charts show the share of respondents (n=100).
"Do you believe your role exists in its current form in 24 months?"
"Which marketing roles or functions at your company do you believe are most at risk of being eliminated or reduced in the next 24 months?" (Open-ended, coded.)
"Has your company eliminated, reduced, or stopped backfilling marketing roles due to AI in the last 12 months?"
"How does your company's external messaging about AI compare to what's actually deployed internally?"
"Which AI marketing tool has most disappointed you or your team in the last 18 months, and why?" (Share of respondents naming each tool.)
"Has IT or security restricted which AI tools your marketing team can use?"
48% of enterprise companies are heavily restricted; 62% of mid-market companies are partially restricted. Companies with heavy, partial and no restrictions report negative ROI at similar rates, so tighter policy doesn't fix ROI.
"Describe how you or your peers actually get AI work done when official tools are restricted or unavailable." (Open-ended, coded.)
"What does your marketing leadership most misunderstand about AI today?" (Open-ended, coded.)
"What's the single biggest reason your company's AI marketing investments have or haven't delivered ROI?"
Top blockers named by the no-ROI group:
What the 23% with positive ROI do: they set strategy first and treat AI as an accelerant; they keep procurement tight with short contracts and value tests; they start with the stuck backlog of work that was never going to happen; and they build specific, repeatable workflows instead of "the team uses Claude sometimes."
"How has your marketing team's size and structure changed over the last 18 months?"
Get new Wynter studies by email.