Session Recaps

June 1, 2026Brand Marketing AI

How Top CMOs Are Using AI for B2B Marketing Strategy

Brand is back. Not because CMOs decided it mattered more, but because AI changed where buyers start, and for most B2B buyers today, that's no longer your website. A G2 study shared at INBOUND 2025 found that 51% of buyers say they go to an LLM first.

In that context, three CMOs, Suzanne Kounkel of Deloitte, Lena Waters of Notion, and Sydney Sloan of G2, covered what B2B AI marketing strategy actually looks like in practice: the brand questions, the agent questions, and what happens when you start thinking of AI as a collaborator.

The through-line: the CMOs generating real results right now are the ones running many experiments instead of one big bet, building brand presence outside the channels they've historically owned, and using AI to think faster, not just produce faster.

Why B2B Brand Strategy Can't Wait for AI to Stabilize

For years, B2B marketing ran on a serviceable formula—buy demand, drive traffic, convert on site. Brand was a nice-to-have. That's over now, because LLMs have taken over the first stage of the buyer journey, and what they surface is shaped by which brands have the deepest, clearest presence across the web.

As Sloan explained, buyers start in LLMs, which means brand has to live in those answers. That requires a two-part approach. You need to optimize for how LLMs represent your brand in their outputs, and showing up in the physical and digital environments where your buyers actually spend time. Her example: watching a tennis tournament and seeing Deloitte ads throughout. Your buyers are somewhere. The question is whether your brand is there with them.

"You're asking the buyer for their trust, and that's what the brand has to do." — Lena Waters, CMO, Notion

Do this: Search your product category in ChatGPT, Perplexity, and Google AI Overviews today. Note whether your brand appears in the answers. And if not, what competitors do. That gap is your brand's current visibility problem.

Your Website Has a New Job

The traffic that historically came to your site for education and awareness is declining. What's replacing it is smaller in volume and higher in intent. By the time a buyer reaches your website now, they've often already formed a view; they've asked an LLM, seen your brand mentioned, maybe validated socially. They're there to confirm, not browse.

"It's an answer engine. They ask a question. You can't break that experience, then, when they do come to your website. It needs to be a continuation of the conversation." — Sydney Sloan, CMO, G2

Navigation designed for exploration doesn't serve someone who has already done their research. Teams are already rebuilding around this reality, moving away from awareness-stage architecture and toward conversion-focused design, first-party data capture, and experiences that add something the LLM couldn't give them.

Kounkel added that the website remains an important asset, but it's just one asset now, not the whole strategy. Brand-level experiences have to exist in the places where customers actually are.

Do this: Look at your homepage through the eyes of someone who already knows what you do. Does it give them a reason to take the next step, or does it make them start over?

Skip Fast Follower. Become a Fast Experimenter.

The instinct for marketing leaders watching AI evolve this fast is to pick a direction and go hard. Suzanne Kounkel's advice is to resist that. The tools are changing fast enough that any single bet could be obsolete before it scales.

"It for sure will punish you if you think you're going to be a fast follower. So what I've asked our teams to do... is having a mindset of being a fast experimenter." — Suzanne Kounkel, CMO, Deloitte

A fast experimenter runs many lightweight tests across tools rather than going deep on any one. At Deloitte (a firm with roughly 450,000 people) Kounkel's team embedded AI into existing internal tools so that brand voice guidance reaches employees without a separate training program.

The second piece of her approach: encourage personal use. Ask team members to use AI tools in their actual lives—planning a trip, writing something for themselves—because that helps fluency develop and where they'll see how fast things are actually moving.

Agents Are a Spectrum. Match the Level to the Task.

The word "agent" covers significant ground. At one end, there's a copilot, something that assists you in completing a task faster. At the other end, full autonomy with a system of agents orchestrating complex workflows without human involvement at each step.

Sloan described it as a gradient, using the analogy of Waze to Waymo. Waze made it possible for a less experienced driver to get somewhere confidently. Waymo removes the driver entirely. Both are useful; they require different levels of trust, preparation, and infrastructure. The same logic applies to agents. A copilot that helps you write better prospecting emails is a Waze. A system of agents orchestrating your entire outbound motion end-to-end is a Waymo.

Waters added that as agents move up the spectrum, the composition of a team starts to change too. Go-to-market has typically run in waterfall; engineering and product runs agile — introduce agents operating at a third speed and the relationships between roles, functions, and humans start to look different.

What AI as a Thought Partner Actually Looks Like

So how do these CMOs use AI to think?

Sloan shared how she spent an hour in conversation with her ChatGPT instance on a set of leadership-level strategic initiatives. After some back and forth, she had a complete strategic paper.

"We went back and forth, and then in an afternoon I had a strategic paper. That changed the game for me." — Sydney Sloan, CMO, G2

Waters described how she uses AI to read the room at Notion. Rather than waiting for a survey or relying on a direct report's interpretation, she'd ask Notion AI to assess whether a product launch felt on track, based on everything surfacing across company channels. The AI synthesized sentiment from Slack, docs, and meeting notes and gave her a human-sounding read that was faster than any traditional feedback mechanism. The information came from the same places a well-connected colleague would have consulted; it just got there first.

The common thread: these are thinking use cases.

Do this: Before your next strategic planning session, spend 30 minutes in dialogue with an AI. Give it context on the problem, ask it to challenge your assumptions, and use the output as a starting point

The Takeaways

The CMOs in this conversation came with real use cases, real teams, and honest accounts of what's working. Brand carries more weight now because AI compressed the buyer journey. The teams ahead are the ones treating experimentation as an operating principle rather than a phase.

AI as a thought partner is already producing results:

a strategic paper in an afternoon
sentiment reads without a survey
brand voice enforced across 450,000 people without a training program.

And they're all early evidence of what a well-integrated AI practice delivers to a go-to-market team. Want access to more sessions and content like this? Join us in Boston this September.

FAQs

How are top CMOs using AI in their B2B marketing strategy right now?

The CMOs at INBOUND 2025 described three primary uses: making their brand visible in LLM outputs, embedding AI tools into existing workflows for team-wide adoption (including brand voice guardrails), and using AI as a strategic thought partner to have substantive conversations that produce better starting points for decisions, campaigns, and plans.

What should B2B marketers do if buyers are starting their research in LLMs?

Develop a two-part approach. First, optimize for LLM visibility through structured content, consistent brand positioning, and clear, authoritative sourcing. Second, ensure your brand appears in the environments where your buyers spend time outside of AI tools. LLMs don't surface every brand equally, and presence in the physical world still drives awareness that LLMs later reflect.

What's the difference between an AI copilot and an AI agent?

A copilot assists a human in completing a task but the person still drives the outcome. An agent operates with more autonomy, executing a sequence of actions toward a defined goal. A system of agents can orchestrate complex marketing workflows end-to-end with minimal step-by-step human involvement. Most teams benefit from starting with copilot functionality and adding agent complexity deliberately.

How should a marketing team get started with AI without overcommitting?

Suzanne Kounkel recommends a "fast experimenter" approach: run many lightweight tests across tools rather than going deep on a single solution early. Encourage personal use among team members so they build natural fluency with how the tools actually behave. Start with tasks that are high-volume and low-judgment like derivative content, data synthesis, and first drafts.

How does a B2B website need to change in the AI era?

Websites designed for awareness and education are losing relevance as LLMs handle that earlier stage of the buyer journey. The visitors who reach your site now are closer to a decision. That means prioritizing conversion-oriented design, conversational entry points, and first-party data capture over broad navigation structures built for exploration.