Back to Newsletters
B2B CMOs are moving fast with AI, but many are still struggling to prove ROI. We cover the shift toward marketing-owned AI infrastructure, stronger governance, changing team skills, board pressure, AI content risks, and the hidden costs of bloated MarTech stacks.
Penguins are one of the few creatures built for both land and sea, and the difference between the two is almost comic. On land, they waddle: Awkward, unstable, tipping over their own feet. In water, the same animal turns graceful and fast, some hitting 22 miles an hour. This month, in Lunch Huddles in Seattle and Palo Alto, one day apart, that image kept resurfacing. Most CMOs are still on land. A few have found the water. Almost nobody has figured out how to move between the two without looking ridiculous at least some of the time. Here's what that messy middle actually looks like right now.
[Special reader request: If you’re a senior marketer, please take our AI Maturity Survey (click here) by Sunday, August 23rd. You’ll help provide benchmarks for the whole B2B industry.]
This is the part of the shoreline where the waddling is loudest. Teams are adopting AI with a vengeance, no question, but the downstream impact is unclear. One CMO I spoke with noted, “AI helped us make up for staff lost in a layoff and helped us hit our numbers, but the AI costs pretty much equal the payroll cuts.” In sum, they're treading water. This was reinforced at another recent event with roughly 100 CMOs in the room; only two hands stayed up when I asked who could tie AI directly to a business metric. Everyone's flapping to some extent.
Some teams have decided the fastest way into the water is to stop waiting for someone else to build the diving board. Faced with engineering orgs that have other priorities, a few CMOs have started building their own infrastructure. "I have to own my own [AI infrastructure]. Marketing needs its own version of that... even though I have a very large engineering team that is really good at this stuff, I can't count on it. I need my own center of excellence." That center of excellence now includes marketing-run GitHub repos, homegrown agents, and increasingly a data lake of its own, pooling CRM and campaign data into one place. As one CMO put it, "for the first time, I can actually see the funnel on a per-event and per-campaign basis." That's a team that's found its footing.
Getting into the water fast is one thing. Not getting swept out is another. The thrill of ungoverned experimentation is devolving into something more sober, as teams realize speed without structure is its own kind of undertow. One CMO watched a near-miss play out firsthand: a customer-facing tool built and quietly put into production without review, and summed up the lesson bluntly: "This is the chaos that can happen if you don't have governance." The fix already underway at several companies is equally blunt: "We put in guardrails by department or by role for how much you can spend. Not everybody gets unlimited." Governance, in other words, is the wetsuit nobody wanted to buy until they got cold.
There's disagreement over whether today’s climate rewards generalists or specialists, and honestly, both camps have a case. One argues marketing has spent years over-sub-functioning itself, and AI is finally forcing everyone back into the same tide: "Marketing, I think to the fault of all of us, has gotten over-sub-functioned." The other insists specialists aren't disappearing: "They're not shedding their creativity, their marketing understanding, their creative energy. None of that stuff is going away. They just now are builders too." What both sides agreed on is that the real predictor isn't a job title: "The employees who jump in and actually discover that moment where they flip the bit... that's the moment where I discover I've got one of those people."
Boards, watching from the shore, keep asking CMOs to prove they're in the water rather than asking where the water is taking them. "Everyone here has had a CEO that just saw something on Twitter from their favorite person, and you get asked about it. They don't do that with their CFO. They don't do that with their engineering leader." The underlying ask, several CMOs agreed, is really just whether marketing looks modern, not a strategy question at all. Most have learned to answer it before it's asked: "I have a slide where I walk through, and I always end with ‘here's our efficiency slide.’"
Even the water itself is starting to shift underfoot. As brands lean harder into AI-generated content, AI labs are starting to watermark their output so platforms can identify, and possibly penalize, it. "If Google somehow decides that your post is 100% AI-generated by its own tool, it might somehow penalize you, as if you weren't already penalized enough with SEO." One content-heavy team, having built an entire owned-media operation on AI-assisted production, is already asking the follow-up question out loud: "Do we need to start investing in human editorials again?" [More on this topic soon.]
Underneath the governance scramble, the DIY infrastructure, the board pressure to prove the water was worth diving into, sits a cost most CMOs haven't fully measured yet. New research from Super Huddle founding sponsor DemandScience, based on a survey of 750+ senior marketing leaders, found that most run at least 11 tools in their stack, with roughly a quarter of that spend going to tools that overlap, sit underused, or don't talk to each other.
DemandScience's Head of Demand Gen, Blaine Prince, described the CMO's version of that pain plainly: "My dashboards and everything I look at on a daily basis in these platforms may look great, and then when it comes down to me going to my pipeline meetings and my revenue teams, I have little to show, and it's always a pain." Blaine's team calls that gap “the platform tax.”
This is just one of the conversations we’ll be having at the CMO Super Huddle, October 22–23.