Marketing teams are drowning in shiny new AI tools, and honestly? Most are making the same mistake.

“I think that in three to five years, companies who are over-investing in AI right now, and specifically marketing teams who are over-investing in AI, are going to see the same implications of people who over-hired during the pandemic,” says Drue Stinnett Maybell, Senior Demand Generation Manager at Bungii. “Obviously, AI is here. You should learn to use it. But find ways to use it sustainably.”

Her blunt take cuts through the hype: just because AI exists doesn't mean you need to use it for everything, everywhere, all at once.

The real problem with AI adoption in marketing

While Harvard Business Review research shows AI adoption is accelerating sales and marketing decisions, most marketing teams are jumping into AI without establishing the foundational work first.

“I think we're over-investing,” Maybell explains. “There are use cases where you should use AI, but you don't need to use it for everything.”

She predicts that teams that rush to scale without proper foundations will end up with bloated or misconfigured operations that don't deliver results.

What teams are missing before diving into AI

Before your marketing team starts building agents and automations, Maybell recommends focusing on these fundamentals:

• Data cleanup and organization: “Get your house in order before you start bringing in complex AI workflows.”
• Legal parameters and brand ethics: “Understand what your company can and cannot do with AI.”
• Clear processes for repetitive tasks: “Identify which ‘intern tasks’ actually need automation.”
• Time for experimentation: “Your creative teams need time to do nothing.”

Where AI agents actually add value for marketing teams

Not all marketing tasks need AI intervention. But when used strategically, AI agents can handle the mundane work that frees up your team for higher-level thinking.

The easiest starting points for marketing automation

According to Maybell, these are the most practical applications for marketing teams just getting started:

• Brand tracking and sentiment analysis: “Brand tracking and brand analysis is one of the easiest things to start with,” Maybell notes. Try tools that can automatically monitor social media mentions, news alerts, and overall brand sentiment without requiring significant setup.
• Automated reporting: “What I hope falls off from manual marketing tasks is actually scheduling social media content. I would trade my left foot to never have to do that again...In general, I’d like to cut everything with reporting.”
• Industry news curation: “Use AI to automatically pull and summarize relevant industry updates, similar to how newsletter tools can scrape specific websites and deliver customized digests.”

Building agents that actually work

When Maybell's team builds AI agents for clients (using tools like Pipedream), they focus on tasks that are:

• Manual but important
• Repetitive and time-consuming
• Don't require human creativity or judgment
• Have clear, defined parameters

“We're helping people identify those repeatable, junior-level tasks,” she explains. “For example, right now we're helping a client on a small marketing team build out an agent to automatically pull all of their reporting.”

The team dynamics of AI implementation

Successfully implementing AI isn't just about the technology — it's about having the right people involved and giving them enough time to experiment.

Who should be involved in AI projects

For the reporting automation project Maybell described, the team includes:

• Someone focused on marketing operations (in a full-time role)
• An external IT team for technical implementation
• Consultants who understand both the business needs and technical requirements

Notably missing from this list? Engineering teams. “The cool thing about these agents is, at least with the tools that we're using, it's really not requiring engineering,” Maybellexplains.

The importance of giving teams time to experiment

One of Maybell’s most important insights centers on time: “Your creative teams need time to do nothing and experiment. If you’re having your team work at 100% capacity, they won’t have time to learn and figure out what's even possible.”

This aligns with recent research from Boston Consulting Group showing that 70% of AI implementation challenges stem from people and process issues, not technology problems. The study emphasizes that successful AI leaders “follow the rule of putting 10% of their resources into algorithms, 20% into technology and data, and 70% in people and processes.”

Metrics that matter: focusing on revenue over vanity numbers

Throughout the conversation, Maybell returns to a key theme: focusing on meaningful metrics rather than vanity numbers.

“I could care less about traffic or clicks. MRR is my key focus now.”

This perspective is important when evaluating AI tools. Instead of getting excited about efficiency gains or time savings, marketing teams should ask: will this directly contribute to revenue growth?

The path forward: sustainable AI adoption

As marketing teams plan their AI strategies for the coming year, Maybell’s advice is refreshingly practical: slow down, clean up your existing processes, and then selectively add AI where it truly makes sense.

The companies that will succeed with AI aren't the ones implementing the most tools — they're the ones implementing the right tools, with proper foundations, focused on the metrics that actually matter.

Follow along with our team as we build Flowerplot — a practical AI-powered data analysis and visualization that helps marketing teams get more out of their survey data.