The AI CMO: How B2B Marketers Use AI and Fix Their Data First
In a 2024 AI Connect webinar, marketing and analytics leaders from Brunswick, OneMagnify and formerly GE Healthcare discussed how B2B teams use AI for customer experience. They said clean data, risk councils and realistic expectations matter as much as the models.
By The AI Enterprise Magazine Team4 min read
Key takeaways
- Use is mostly operational. Panelists said early AI marketing wins were content drafts and efficiency, with customer-facing use still limited.
- Data comes first. Disconnected systems, duplicate contacts and weak agency data feeds block advanced use cases like next best action.
- Govern use cases centrally. Brunswick uses an enterprise risk council of legal, IT and marketing leaders to review AI requests.
- Beware the silver bullet. Beckstedt said leaders must manage expectations that AI instantly solves problems like customer lifetime value.
- Start with a problem. Small projects aimed at a specific issue build support better than deploying AI for its own sake.
In a 2024 AI Connect webinar, three marketing and analytics leaders described how generative AI was changing B2B marketing and why data, governance and expectations were holding it back. The panelists were Lauren Beckstedt, Chief Marketing Officer at Brunswick Corporation; Amy Musolino, former Chief Marketing Officer at GE Healthcare Pharmaceutical Diagnostics (Americas); and Jonathan Prantner, Chief Analytics Officer at OneMagnify. A AI Connect co-founder moderated.
They agreed that adoption was early. Most use was operational, and the harder work was fixing data and building the internal partnerships that AI projects need.
How are B2B marketers using AI to improve customer experience?
Beckstedt said Brunswick sorts use cases into customer-side value (experience, engagement and exchange) and internal value such as strategy and efficiency. On the experience side, she said, use was still limited to things like chatbots. Most of the work was operational, helping "smart people" spend their time better.
Musolino said healthcare and pharmaceutical marketers lag on new technology because the industry is highly regulated. Her teams started with first drafts of social and blog content. Much of their material was technical or based on internal research that compliance would not allow on the internet, so AI helped less there. She cited an agency that used AI on product images to avoid photo shoots around the world.
Prantner said the biggest effect comes when AI is pointed at understanding the customer: segmentation, targeting of high-value B2B prospects and customized content. He also described linking what customers say to each other, to a brand and in call centers and search terms, and using retrieval augmented generation to search internal documents.
How should marketers prepare their data for AI?
Prantner said the techniques resemble existing data hygiene: combine first- and third-party sources, clean them, and add metadata to unstructured text so language models can use it in standardized prompts. He also cited automated cluster generation and labeling, which uses unsupervised methods to create segments people can understand.
Musolino described the basics as the hard part. She said she had seen disparate systems, duplicate contacts entered by different field teams, little visibility once products went through distributors, and agencies that could not deliver timely data in the right format. In that setting, concepts like next best action felt aspirational.
Beckstedt framed it as risk. Brunswick assesses input risk, such as whose intellectual property goes into a model, and output risk, such as a model giving unsafe advice and the question of who is liable. She said companies that own more of their own IP are advantaged.
How do marketing teams work with IT, legal and data teams?
Beckstedt said Brunswick has an enterprise risk council that reviews each AI use case. Legal, IT and she vote, and the chief human resources officer joins when talent data is involved. Requests arrive as short use-case briefs and are reviewed as they come in, because some are as simple as an AI feature inside design software.
Prantner said OneMagnify's technology teams helped set guardrails on cost, hosting and security. He said the company leans toward security and built support by showing small projects that solved specific problems.
The moderator said that as AI use cases multiply, a standing council with standard protocols for data access can keep items from falling through the cracks.
Where is the pressure to adopt AI coming from?
Musolino said her main pressure was justifying marketing itself, since long, complex B2B sales cycles make its impact hard to quantify. She hoped AI-based statistical models could help show long-term value to executives.
Beckstedt called her biggest pain point the "Silver Bullet syndrome," the assumption that having AI means a company already knows its customer lifetime value. She expected AI to arrive as "a means to an end," an enabler like email.
Prantner said marketers were squeezed between boards asking why they weren't using AI and legal and technology teams urging caution. With fewer sales and the loss of third-party cookies, he said, attribution models were being applied to CRM data instead. The moderator added that data leaders hear the same pressure, and that "AI is a tool, it's not an end in itself."
What opportunities did the panel see?
Musolino was most excited about building personas from actual buying patterns, modeling buying groups rather than single buyers, and mapping content to buyer journey stages. Near-term, she named content generation, channel justification and customer service analysis.
Beckstedt wondered whether AI could reduce the premium on customer data as privacy rules tighten, by letting teams make good decisions with less of it. Prantner described research-based persona chatbots that test how content might land with different buyer types.
Hear more from data and AI leaders in 2027
COLLIDE 2027, the largest AI conference in the Southern U.S., brings more than 1,000 enterprise data and AI leaders to Atlanta on Wednesday, October 13, 2027, at the Sandy Springs Performing Arts Center. AI Connect also offers 24 free live webinars on its 2027 schedule. Details are at /webinars/ and /events/collide/.
Frequently asked questions
How are B2B marketers using generative AI?
Panelists described content drafting, customer segmentation and targeting, voice-of-customer analysis, retrieval over internal documents and persona-based testing of content.
Why does data quality matter for AI marketing?
Amy Musolino said inconsistent systems, duplicate contacts and poor agency data made even manual decisions hard, so advanced AI uses were aspirational.
What is an AI risk council?
Lauren Beckstedt described Brunswick's council of legal, IT and marketing leaders that reviews each AI use case and decides whether to proceed or assess further.
Why is measuring B2B marketing impact hard?
Long, complex sales cycles and fewer transactions limit the data available, the panel said, which is why teams are applying attribution to CRM data.
This article is based on remarks by Lauren Beckstedt, Chief Marketing Officer, Brunswick; Amy Musolino, former Chief Marketing Officer, GE Healthcare; Jonathan Prantner, Chief Analytics Officer, OneMagnify at AI Connect webinar, 2024 on May 23, 2024. More coverage: Webinars.
Filed under: AI Use Cases & Applications, AI Governance & Trust, People, Culture & Leadership
