Dr. Eva Wolf spent 12 years scaling a 3D printing company in California before starting over. She earned a doctorate in her fifties and moved to Miami.
She runs Big Plans Media out of Coral Gables and teaches marketing at Miami Dade College and Colorado State University. On June 23, she defended her Doctor of Business Administration at Florida International University. Her dissertation tested AI-generated, personality-targeted advertising on 394 B2B decision-makers. It extended earlier consumer research by Sandra Matz into a business market where the results turned out messier.
This is part of a peer-to-peer series of conversations MAIN runs exclusively with Miami's AI founders, investors and leaders.
We asked Wolf five questions by email. Her answers, lightly edited for length, are below.
What actually makes an AI recommend one firm over another, and where should a professional-services firm start?
"It actually depends on the location, on your particular industry and on the AI assistant that is being asked. I've noticed that it's not the same across the board, and I've been testing this for a while… A good place to start is by asking all six AI assistants the same questions, and doing it incognito to establish a baseline. Determine who the recommended agencies are and then do the research for those top five and see why those top five are being recommended across the different AI agents. Is it because they have a Google Business Profile? Is it because they have a certain number of .edu backlinks? Is it because they have a good Trustpilot score? It really depends on the industry that you're in…"
Rankings shift with the training data, she said.
"Keep in mind, however, that this is a moving target. So the variables that help them rank high today may be completely different next month. It really depends on the training data that was used to train the most recent LLM model. And they all train on different data sets. So for example, sometimes they get trained on information from Reddit. So having Reddit backlinks would be beneficial at that time and period, but that may not be the case tomorrow."
Where does AI adoption pay off fastest for expert-led B2B firms, and where are they wasting money right now?
"A lot of companies think that they are adopting AI. There was a recent study done this year by Nicholas Bloom and his co-authors at the National Bureau of Economic Research, a survey of nearly 6,000 senior executives across the US, the UK, Germany and Australia. The survey showed that 69% of firms actively use some form of AI. But then the survey delved deeper into how much these executives actually use it themselves, and it turned out to be about only an hour and a half a week. So the adoption is wide, but it is shallow…"
Real adoption looks different, she said.
"Adoption means actually rolling out an implementation plan, training your staff, helping them build agents, documenting their workflow, evaluating what workflows should be automated, and then which ones need a human overseer in the loop. That is adoption. So where does it pay off? It pays off on workflows that can be automated… Most of the time, there's a higher likelihood that the biggest payoff is in marketing."
The waste shows up elsewhere, too, she said.
"I hear companies complaining about their employees burning through tokens. And so there's a lot of money being wasted on tokens by not following a well thought out, methodical AI implementation plan."
Your doctorate tested AI-generated, personality-targeted ads on 394 B2B decision-makers. What did you find, and does it work as well in B2B as the consumer-market hype suggests?
"I was surprised to see that AI generated personality aligned ads did not work as effectively across the board in the B2B space. A brilliant researcher by the name of Sandra Matz tested the effectiveness of AI generated ads that were aligned and targeted to specific personality types, and they were for the most part effective in the B2C context. But I went ahead and redid a similar study in the B2B space to see if AI generated ads would be as effective. In certain circumstances they were, but in other circumstances, they actually backfired. They were not as effective, particularly with the conscientious personality type. So this tells us that we need to have a human in the loop. There's not ever going to be an AI replacement for good old fashioned human judgment."
Your Research Radar briefing is produced by an automated pipeline you built yourself, published in public as proof of what you build for clients. Why run the engine in public, and what did building it teach you that client work couldn't?
"It started off as a gift for my favorite marketing professor. He started losing his eyesight, and so I built this as a way to automatically generate a podcast on a daily basis and to interpret the leading research that was published most recently, so that he could listen to it. I built it that way as a personal and private thing, but then I decided to do it publicly so that others could see the power of automatically generated podcasts and audio summaries. Up to today, I think so far it's generated around 64 episodes…"
That mistake taught her to verify everything twice, she said.
"One thing that I learned from the AI generated podcast pipeline is that you need to add a second verification process to everything. One of the podcasts that it generated was based on an invented paper. And now I have a fail-safe in there to check every citation, so that when the agent writes down what they say, it's not invented, and it's based on an actual paper that exists."
You scaled up a large 3D printer company in California over twelve years, then earned a doctorate at 53 and set up in Miami. Why here, and what does Miami's AI scene still need to become the place firms like yours thrive?
"I came on vacation to Miami and I fell in love with Miami. I couldn't see myself living anywhere else. I love the energy here, and it reminds me of the energy in Southern California from the late 80s, early 90s, a sense of anything is possible…"
Talent was never the gap, in her telling.
"What does it need to become a place for tech companies to thrive? Definitely not the labor. There's no shortage of skilled talent and labor in South Florida. A community college like Miami Dade College offers four-year degrees now in the computer science fields. So the skill set and the talent is definitely not a shortcoming."
"What Miami still needs, in my opinion, are two things. We need our local firms to show up in the answers. If you ask an AI assistant who to hire, oftentimes they name national agencies, and the local Miami firms are not usually being named. Miami firms need to have experts write about them, which is exactly what this news website does for the public. It's a local outlet and it talks about local firms."
"The second thing we need is more hands-on workshops. We're having AI events, large things happening here in Miami with plenty of AI-related motivational speakers. But what we're missing are sit-down events and workshops that are more hands-on, where owners can build one thing that's going to run for them on Monday."
Wolf's own quote about local coverage describes MAIN's model back to us. A Miami firm gets named by an AI system when a real local outlet writes about it. Spending more on marketing doesn't do that on its own. She sees no shortage of talent behind that gap, only a shortage of coverage.
Most local firms still miss that equation. They add themselves to a relevant, but national directory listing and expect the same result. A directory listing is not the same as a story written about them by a real outlet. Not in an AI system's eyes.
Dr. Eva Wolf, Founder, Big Plans Media, Miami, Florida. LinkedIn | AI Slop episode
