As buyers increasingly turn to large language models for product research, B2B marketing strategies must evolve to accommodate for this shift, according to a panel of CMOs speaking at HubSpot’s INBOUND conference earlier this month.
Chief marketers from G2, Notion and Deloitte—companies at the forefront of the AI marketing revolution—emphasized the importance of brand marketing for B2B organizations, rethinking owned communications and the benefits of AI experimentation.
The Resurgence of Brand in the AI Era
Prior to the ubiquity of LLMs—whether it’s OpenAI’s ChatGPT, Anthropic’s Claude, Google’s Gemini or another popular model—B2B companies viewed brand marketing as a less critical part of the marketing mix, according to the panelists.
But with the majority of buyers now researching products and services on these AI platforms—four out of five people, according to a recent G2 study—the need for consistent brand messaging to stand out amid AI search responses has become more urgent.
“I think a lot of us took advantage of the fact that we could buy demand without investing in brand for a long time—and now you can’t,” said Sydney Sloan, CMO at software review platform G2. “Your brand has to play a larger role because you still want people to think about the brand, if the LLM is not indexing on it yet,” she said.
And that requires an uptick in brand marketing to generate interest and awareness. “[It’s] thinking about where people go and the strategies associated with that,” she said. “They’re going to the LLM. Where else do they live? I think that’s where we need to start getting creative about understanding where our personas and our buyers live.”
Though it’s not always perceived as such, buyer decision-making is an emotional process rather than a rational one, according to Lena Waters, CMO at AI workspace tool Notion. “The fallacy in B2B marketing is that we assume that people make decisions via logic, and really we know they make them with emotion—and then they use the logic and our comparison charts to backwards justify whatever their decision was,” she said. “That doesn’t go away with B2B. I think we accept it for consumer, so why not through business? … That’s where the brand is going to come in and play that huge part.”
Rethinking Content Creation for AI Indexing
With consumers educating themselves about products and services using LLMs, the role that owned content plays in representing brands is evolving within the buyer decision journey, panelists agreed.
Companies’ owned websites, for example, will evolve and serve a different function for users, said Suzanne Kounkel, Global CMO at Deloitte. “It’s all about making sure your content is discoverable and attributable. And that causes all of us to rethink the way we create content, and certainly what the role of our website is,” she said.
Rather than being the first point at which a consumer encounters your brand after a Google search, the interaction is further down the consideration funnel. “You need to make sure that your website going forward has a strategy around first-party data and the true benefit it can offer to people that are coming, because again, people won’t be coming to a website for the same reasons they historically have,” Kounkel said.
Yet the traffic headed to your website might actually have greater value, which means your content strategy should reflect that, according to Waters. “Do we have better intent there to be focused more on conversion and less about education?” she noted. “If buyers can really educate themselves and convince themselves and become aware through multiple other channels, is the role of your website still to be that standard brochure that takes the audience all the way from awareness and interest and education, engagement and conversion? Or is it really more at the bottom of the funnel?”
All this points back to the rising importance of brand for B2B, Waters said. “You need to be able to go out into the world and make sure that your brand is living in all the places where your customers are, and you’re not necessarily relying on the website. Websites still important, but it’s just one tool now in the portfolio and you have to adapt to creating those ground-level experiences elsewhere.”
Sloan added that a brand’s website should function as a continuation of the conversation that potential buyers are having with LLMs. So, just as answer engines are responding to questions, website content should as well—albeit at a different stage.
“We’re thinking first, how do you make it a prompt-oriented or a conversation-oriented experience where you have this mix of, ask it a question, but you still might have your agent there for support?” Sloan said. “You still have to educate the LLMs, so how do you think about the content structure of your website in order to do that?”
Sloan shared that some brands she’s spoken to are even building “shadow sites” to fuel LLM answer engines while keeping a frontend that continues the “conversation engine” they’re building through brand experiences. “I think we’ll all have very different website experiences in the next 12 months,” she said.
Internal Integration: Favor a ‘Fast Experimenter’ over ‘First Mover’ Mindset
When asked about implementing AI tools within their own organizations, panelists said a key step to integrating AI into business workflows with success is acclimating teams to these new enterprise tools. But Kounkel warns that adapting too fast within a market that changes so quickly could backfire.
“In this market, because it’s moving so fast and the capabilities are changing so dramatically, don’t necessarily favor a mindset where it’s first mover advantage … it for sure will punish you if you think you’re going to be a fast follower,” she said. Instead, Kounkel asks her teams to take on a “fast experimenter” approach. “In that world, it’s really about getting your teams comfortable with that and making sure that you sit down immediately with whatever you call the risk part of your organization and design that together,” she said. And that entails getting your data strategy in line. “Because that will either help or hurt you for many years to come,” she said.
For AI Use Cases, Think Automation and Scale
To determine which use cases work best for your marketing organization, Kounkel recommends that companies identify tasks that humans “don’t do that well and don’t do that quickly.” “Those are easy ways to get gen AI up and running and faster,” she said. For Deloitte specifically, the company has found AI useful for creating derivative content from original content and quickly creating pieces for social. “That’s a table stakes, no regrets move,” she said.
The next step is moving on to the work that your team hasn’t been able to either scale to a satisfactory degree or fortify. “As marketing organizations … [it’s] figuring out some places where you just haven’t been able to do what you want to do … and thinking about how that might look, whether that’s lead nurturing … we’re doing some things on that where it’s sort of another team member, if you will, that we can call on to help us with that.”
How Marketers Can Think About Using AI Agents
Not every marketing team is equipped to begin using AI agents, of course, which HubSpot defines as “an AI system that can act independently to set goals and accomplish tasks.” A new global marketing study from SAS and Coleman-Parkes shows that 51% of marketing decision-makers say their organizations have plans to invest in agentic AI in the next year—but only 21% are actively testing it in live environments.
For one, your data must be highly accurate in order for systems to properly communicate and deliver desired outcomes. A sophisticated example of this in action, Sloan explained, is the self-driving car brand Waymo. “You are trusting that somebody has built this system of agents that sit on the data and know how to take you from point A to point B [and] they can trust it,” Sloan said.
But that’s the top end of the spectrum, Sloan noted. For marketers looking to implement AI agents, it’s not a one-size-fits-all scenario. “As you’re thinking about your strategies around where do agents fit, it’s a gradient, and you will use the appropriate level for the task or the process at hand. And it can be something simple as a copilot or it could be something as complex as a system of agents,” she said.
Waters envisions AI agents as “teammates” that assist the marketing team. “We can start to think about putting those agents together and really creating workflows to create outcomes,” she said. “It hasn’t quite happened yet. We’re sort of on the cusp. You might be using it to find your trip, but you’re not giving it your credit card number and it’s not booking it for you and setting your out of office—but it’ll be soon.”
Network
