7 AI Lessons From CMO Huddles Strategy Labs

Quick summary: Over the past year, CMO Huddles brought five of the most respected AI practitioners in B2B marketing into Strategy Labs across the country. What emerged were seven hard truths about AI adoption, from why most teams overestimate their maturity to why the next six months matter more than the last two years.

---

Most AI conversations in B2B marketing sound the same. Here are the tools and the use cases; here's why you're behind. Then the speaker leaves, and nothing changes.

CMO Huddles ran a different kind of session, bringing six respected AI practitioners in B2B marketing into CMO workshops (Strategy Labs) across a dozen cities to discuss what they were seeing across B2B organizations.

Taken together, here are the seven lessons that emerged:

1. AI Adoption Starts With Mindset, Not Tools
2. Most B2B Marketing Teams Overestimate Their AI Maturity
3. Middle Management Is Where AI Transformation Stalls
4. AI Training Only Works When It's Role-Specific
5. Reddit Is Now B2B Marketing Infrastructure, Not a Social Channel
6. AI Is About to Rewrite the B2B Go-to-Market Cost Structure
7. The Window for AI-Driven Competitive Advantage Is Closing

1. AI Adoption Starts With Mindset, Not Tools

Every expert named this first, and in almost the same words. Teams are using AI, but they are not changing how they work.

Liza Adams, an AI advisor who has worked with marketing teams from 20 to 150 people, describes four mindset shifts she has watched the best teams make.

• Moving from AI as a question-and-answer machine to AI as a thought partner.
• Moving from a tool you chat with to a teammate you train and orchestrate in workflows.
• Moving from using AI to make existing work faster to using it to reimagine work that was not possible before.
• Moving from AI that helps individual functions to AI that drives outcomes across the entire organization.

“If we simply make old work faster and automate it, you can project how the human can be automated out. But if we reimagine the work and do something different, we’re innovating. It grows the business. Then humans are essential.”— Liza Adams, AI advisor and go-to-market strategist

Samantha Stark puts it differently: Most organizations commit "random acts of AI." Pilots here, tools there, a few wins that nobody can replicate. What's missing is not more tools, but a reason to use them. AI in search of a strategy produces burnout. AI in the service of a clear destination produces results.

Nicole Leffer adds the mechanism: Teams will only build using AI to the limits of what they believe AI can do. If the mental model is that AI is fancy search engine, the team will use it like a fancy search engine. The job of the CMO is to show teams what's actually possible so the mental model expands.

Sources: Liza Adams (Dallas, Austin); Samantha Stark (New York); Nicole Leffer (Atlanta)

2. Most B2B Marketing Teams Overestimate Their AI Maturity

Nicole Leffer has trained over 100 B2B companies on AI adoption. The maturity pattern she sees everywhere is the same.

• Level One: Chat. Back-and-forth with an LLM to get outputs, starting from scratch each session.
• Level Two: Workflow. Prompt templates, repeatable processes, GPTs and agents that systematize how the team works.
• Level Three: Agentic. Autonomous AI that receives a task, reasons about how to accomplish it, and executes without a human at each step.

The capacity unlock at each level is not linear. At level one, teams typically see a 25 to 50 percent efficiency gain. At level three, they see 200 to 300 percent. The reason most teams stay at level one is not the technology; it's that they've never been shown what level two and three actually look like in their own work.

“You cannot skip the levels. Every step of the maturity ladder teaches you something you need to know to operate safely and effectively at the next one. The teams that jump straight to agents without the foundation are the ones whose agents go off the rails in ways they cannot diagnose.”— Nicole Leffer, B2B AI adoption strategist

Leffer also names the specific technique that most reliably moves teams from level one to level two: Editing the prompt instead of chatting to refine the output. Most people get a mediocre output and ask for a revision, then another. They are having a conversation.

The unlock is to stop, identify exactly what the mismatch is between what came back and what was needed, edit the original prompt, and rerun it. When you do that, the output improves. More importantly, you now have a prompt you can save, share, and use again. Achieving this will move you from conversation to workflow.

Leffer is equally direct about the risk at the other end: Deploying agentic tools before you understand the failure modes. She tested an agentic tool on a research task across dozens of companies simultaneously. The outputs looked thorough, but when she spot-checked one result, it had missed a major acquisition news story that a standard deep research query caught immediately. The tool was not built for that kind of research. She knew that because she understood its limitations, but most teams deploying agents do not.

Source: Nicole Leffer (Atlanta)

3. Middle Management Is Where AI Transformation Stalls

Front-line teams are largely willing to adopt AI. Senior leadership has mandated movement. The people quietly making every pilot harder than it needs to be are middle managers.

Sandra Miley, who has helped dozens of companies through AI adoption, is direct about the statistics: Middle management is the primary source of AI resistance in most organizations because they are protecting workflows they own and control. They will not say they are resisting; they will complicate things and raise concerns that sound reasonable until you realize nothing is moving.

She recommends a specific technique for surfacing this: Take the transcript of any meeting where an AI initiative stalled. Put it in Claude or another LLM and ask it to identify the agendas and patterns. What you will see is what you missed in the room. Miley describes getting back an output that said, in effect: Stop. You cannot move forward until you address the change management problem with this specific individual.

“In these meetings, you remember the last thing said and the feeling when you walked out. You seldom remember the first and middle things. When you have a transcript and AI is analyzing all the comments someone makes, it will recognize the pattern.”— Sandra Miley, AI adoption strategist

Samantha Stark adds the cultural dimension. In innovation cultures, people tinker when they hear a clear narrative about why. In risk-averse cultures, they try once, fail, and stop. According to Stark, the number one determinant of adoption success is whether the leader has a clear, repeated story about why the organization is transforming with AI. Not a mandate, a story.

Liza Adams frames the leader’s obligation the same way: Upskilling people is a fiduciary responsibility. The market has shifted. The people you hired to serve a market that no longer exists deserve the chance to build the skills that will let them compete in the one that does. You cannot guarantee their jobs, but you can guarantee you gave them a genuine shot at the next thing.

Sources: Sandra Miley (San Francisco, Palo Alto); Samantha Stark (New York); Liza Adams (Austin, Dallas)

4. AI Training Only Works When It's Role-Specific

Samantha Stark has done research across companies on AI training. She found that 85 percent of organizations offer AI training, but almost none of them are doing it well.

The gap is specificity. Generic training on tools produces generic behavior. A product marketing manager who learns how a chatbot works is not the same as a product marketing manager who learns how AI changes the craft of product marketing, with their actual workflows, deliverables, and tools.

Sandra Miley names the principle directly: Train on tasks, not tools. The goal is never for someone to know how to use Claude. The goal is for them to do their job better. Start there and work backwards to what the tool needs to do.

⚠️  The most commonly overlooked prerequisite for AI agents and automations is an AI knowledge base: A repository of the critical documents that the agents will need to function well.

Teams that build this infrastructure before they build the agents find they can move faster and trust the outputs more. Teams that skip it build sophisticated tools on weak foundations.

Leffer adds a practical unlock that works across almost every organization: Find the people on the team who light up when they see what is possible, give them the skills and time, and let them become the internal case studies. Peer-to-peer adoption is faster and more durable than top-down mandates. The goal isn't to get everyone to the same level simultaneously. Instead, the goal is to find who will lead, give them the conditions to demonstrate results, and let the social proof do the work.

Sources: Samantha Stark (New York); Sandra Miley (San Francisco, Palo Alto); Nicole Leffer (Atlanta)

5. Reddit Is Now B2B Marketing Infrastructure, Not a Social Channel

Eric Eden spent 25 years as a B2B marketer before he became convinced that the playbooks he had used his entire career were broken. Email marketing. LinkedIn. Google SEO. They still work, they just don't work the way they used to. The cost has gone up, the reach has gone down, and the ROI has inverted.

He spent the last year testing a different channel. His conclusion: Reddit is the most important B2B marketing channel of 2026. And his argument is structural, not tactical.

Two years ago, Google and OpenAI each licensed all of Reddit’s content, paying a combined total of roughly $150 million per year. That licensing decision has an immediate consequence for B2B marketers. Reddit content is now prioritized in Google search results. And approximately 40 percent of ChatGPT citations come from Reddit, with another 23 percent from YouTube.

A Reddit and YouTube strategy covers more than two-thirds of the citation landscape in AI-generated responses. Forrester Research puts the share of B2B buyers who now consult AI before any purchasing decision at 95 percent.

If your company does not have a meaningful Reddit presence, you are not in the conversation your buyers are having before they ever contact you.

“You’re flooding the zone with your positive content and your positive perspectives on a wide variety of topics. You can’t do that on other platforms like LinkedIn, because they only want you to do a couple posts a week.”— Eric Eden, multi-time CMO and AI content strategist

Eden’s strategy is counterintuitive to most B2B marketers. Don't try to participate in other communities. That is the path to getting banned. Instead, create your own subreddit. It's free. Once you have your own community, you set the rules, you control the content, and you build an audience that receives everything you post directly in their feed.

The content strategy here is volume combined with quality. Eden posts two to five pieces per day. Each piece takes under 30 minutes to produce using AI tools for research, writing, and visual design. He directs the content as an editor, not a writer. In 13 months, he built an audience of 70,000 subscribers. His content has been read more than 25 million times. He gets 20 times the distribution on Reddit than he gets on LinkedIn for the same content.

On the paid side: Reddit ads ran 89 percent cheaper than LinkedIn in his testing, at under a dollar per click for targeted B2B traffic. If you are spending $40 or more per click on Google and seeing diminishing returns, the economics are worth a test.

Sources: Eric Eden (Chicago, Seattle, Boston, Raleigh-Durham)

6. AI Is About to Rewrite the B2B Go-to-Market Cost Structure

Amanda Kahlow has been making a prediction that reliably gets a laugh: Within a year or two, sales will report to marketing. She says it not to provoke but because she believes the economics make it inevitable once AI reshapes the cost structure of go-to-market.

The argument starts with a number. It currently costs between three and five dollars to acquire one dollar of ARR using traditional B2B GTM motions. Human SDRs and AEs are expensive and capped by time and attention. The question Kahlow has been working on is what happens when you replace significant portions of that motion with AI that never sleeps, never has a bad day, and scales without adding headcount.

She calls these AI superhumans, not bots. The superhumans are trained AI representatives that carry the brand voice, know the product deeply, qualify buyers in real conversations, and hand off to human AEs at the exact moment human judgment is required.

Her pitch is not replacement, it's that humans become dramatically more effective when they are no longer spending their time on everything that does not require a human.

“The companies that figure out how to deploy AI across the customer journey are not just going to be more efficient. They are going to have a fundamentally different cost structure than their competitors.”— Amanda Kahlow, founder and CEO, 1mind

The results she has seen at companies deploying this model show conversion rates from AI-assisted interactions running three to four times higher than traditional chatbots. Buyers find the AI conversations more useful than waiting for a human, because the AI is always available, always prepared, and never pushes to close before the buyer is ready.

What she pushes back on hardest is the efficiency framing that most AI conversations default to. Efficiency is a floor. The companies that will win are not the ones that use AI to do their current GTM 30 percent cheaper, they are the ones reimagining what GTM is capable of. The goal: Serve buyer segments they could never have served before, and build a motion that competitors still running the old playbook cannot match on cost or coverage.

Source: Amanda Kahlow (CMO Huddles Strategy Labs, multiple cities)

7. The Window for AI-Driven Competitive Advantage Is Closing

Every expert said some version of this. The teams that move in the next six to twelve months will be very hard to catch.

On Reddit, Eden is direct: The category authority plays are filling up. The time to build a subreddit with topical authority is before your competitors do. Early mover advantage in LLM citations is real and compounding. B2B brands flooding the zone now are training the models that buyers will query for years.

On agentic workflows, Leffer and Adams converge on the same observation: Smaller companies are advancing fastest because they have no choice. They cannot afford to staff up, so they have to reimagine work. Enterprise CMOs with 100-person teams are watching smaller competitors build AI-native GTM motions that large teams, weighed down by existing processes and headcount, cannot easily replicate. The innovator’s dilemma is happening in real time inside B2B marketing.

On GTM, Kahlow’s point about cost structure is a competitive argument. When the companies that crack AI-native GTM can acquire a dollar of ARR for significantly less than it costs their competitors, they can reinvest that margin into growth, pricing flexibility, or product. The gap compounds. It becomes very hard to close.

“What I fear is that the top five to ten percent will create market differentiation. They’re going to have competitive advantage. And then by the time the rest are able to react, it is as a result of the market telling them that it’s not going to cut it. By that time, it’s going to be too late.”— Liza Adams, AI advisor and go-to-market strategist

The consistent counsel across all five experts is not panic, but sequencing. Pick the two or three initiatives tied to strategic priorities that already have executive attention and budget. Get a win. Prove the model in a contained way. Then expand. The teams that tried to transform everything at once are the ones with burnout and sprawl. The teams that started small, succeeded visibly, and earned the mandate for more are the ones now running ahead.

The moment is real. So is the time available to act on it, if you start now.

Sources: Eric Eden; Nicole Leffer; Liza Adams; Amanda Kahlow

---

This article draws on insights from CMO Huddles Strategy Labs held across Austin, Dallas, Chicago, Seattle, Boston, San Francisco, Palo Alto, Raleigh-Durham, New York, and Atlanta in 2026.

Want More?

• For the full AI maturity model framework, read: Nicole Leffer’s AI Maturity Model for Marketing Teams.
• For what is actually stalling AI adoption inside B2B marketing teams, read: 9 Reasons AI Adoption Stalls in B2B Teams.
• For the practical case on why Reddit belongs in your B2B marketing strategy, read: Reddit Is Now the #1 B2B Marketing Channel.
• For the full AI-native GTM argument, read: The GTM Team of the Future Might Not Be Human.
• CMO Huddles brings together senior B2B marketing leaders for candid, peer-to-peer conversations on the challenges that matter most.

Frequently Asked Questions About AI and the Future of B2B Marketing