Source Checks And Intent Signals Turn Marketing Activity Into Business Intelligence

A B2B marketing team publishes to a calendar of blog posts, case studies, event coverage, and social campaigns. The monthly report shows how each of them performed, and executives read the totals as evidence that the market is paying attention. The counts themselves may be reliable, but reports like these rarely say where the responses came from. Budgets are approved, teams get staffed, and campaigns are renewed against those figures without always knowing who responded.

Ken Kundis is the Chief Marketing Officer at CEI, an enterprise AI engineering firm that builds and runs AI systems inside large companies. There, Kundis leads a three-person marketing team. Before CEI, he was Chief Marketing Officer for North America at a large global systems integrator, and he founded the B2B marketing consultancy B2Brand. He has now run the marketing function at both ends of that range, and the comparison shapes how he reads a performance report.

"So much of the social media that was taking place in my previous role was by and for the people on the team. We would get 600 to 700 likes and 200 to 300 comments. But if you go through them, 90% of them were coming from inside the house," says Kundis. A report that skips the source check presents those reactions as evidence of market interest. Kundis looks at who responded before he looks at how many, and the answer decides whether the number reaches a business review. Engagement from colleagues is marketing activity, and engagement from buyers is business intelligence.

Auditing the content mix

Kundis ran a content analysis when he took over the North America marketing team at the global SI. The company sponsored a major global series of sporting events, and Kundis recalls one event in particular that produced material that performed well by every measure the team reported. The analysis established how much of the calendar that material had taken. "We had the right number of posts, we had the right number of engagements, we had the right number of likes and shares on social media," notes Kundis. "However, 90% of what we were doing was event-related, and we're not selling the event."

Kundis says roughly half the marketing budget went to the sponsorship, which made it the default subject for most of what the team published. Redirecting the calendar toward enterprise services meant arguing against work that was hitting its targets, and he took that argument to the team himself. "The sponsorships can raise all ships and can create a great narrative that we can tell the story around," explains Kundis. "But at the end of the day, we're selling technology services to large enterprises. We're not selling passes to sporting events."

A content calendar fills with whatever earns a response, and Kundis describes that as a gradual process. He has watched it happen inside large organizations, where the volume of output makes the pattern hard to see. Counting the calendar by subject rather than by performance is what brought it into view. "I still very much believe that content is king, that the messages that you're putting out are what you're going to become known for," says Kundis. "Being mindful of who you're talking to, what you're saying to them, and what the point of it all is, is something that surprisingly can get away from some companies."

Output from a smaller team

Kundis judges content by whether it reaches buyers, and he now applies that judgment with a team a fraction of the size of the one he left. He does hands-on production work himself again, which he wanted from the role, and work that used to need more people now gets done with software subscriptions. "Those 70 people in my mind didn't produce as much as I'm producing with three people and a handful of AI subscriptions," adds Kundis.

The conventional route to a launch video would have meant an agency, a scripted shoot, B-roll, stock photography, a videographer, an editor, a voiceover artist, and licensed music. Kundis estimates that package at 15 people and $40,000 at minimum. The version that ran for CEI's brand transition needed two people, a video subscription, and a model for a first draft of the script. "The intention hasn't changed and the creativity hasn't changed," notes Kundis. "The difference is that someone like me can have an idea, and now I have the tools in front of me to execute on that idea."

Kundis recalls a previous employer who handed out AI licenses across the company and blocked the tools from reaching data outside it. The setup covered internal documents well enough, but marketing runs on outside information: what competitors are doing, which prospects are active, and where a market is going. "If I can't query outside of the business, then it doesn't help at all," says Kundis. "Particularly very large organizations that have a tendency to err on the conservative side of data governance and data privacy are still very much struggling with how to maximize these tools."

From summaries to intent signals

A team able to query outside its own systems can act on shorter notice, and CEI moves messaging on a horizon of days and weeks rather than months and years. Kundis sets that pace while noting the company lives with the decisions afterward. A monthly performance summary arrives after the period it covers has ended, changing what a marketing team needs from its data. "There's nothing that's going to replace speed at this point for us," says Kundis.

Summarizing what happened is the application Kundis considers least interesting. He points instead at account-based marketing, where the work runs from an ideal customer profile down to the account, then to named executives inside it. His current priority is widening a contact set that reaches only a narrow slice of decision makers at each client, and each of those names needs evidence of intent attached to it. "If they come to a webinar, if they come by our booth, we did a demo for them, if they're coming to our website, trying to capture all that stuff," explains Kundis.

Personalization built on those signals fails when the underlying facts are wrong. Kundis uses emails naming the wrong employer, taken from a data scrape rather than from his profile, as an example. The recipient of a personalized message is the one person certain to catch a mistake in it. Keeping the human in the loop means someone checks those details before the message goes out. "I can always tell when someone's done a certain amount of research on me, and I always appreciate that, even if that research is done through AI," concludes Kundis.