How Madhive Is Powering Local Advertising With Automation

AI might still feel like a buzzword to some, but at Madhive, it’s baked into the foundation.

In this Q&A, part of a panel series filmed in partnership with Marketecture, Madhive CEO Spencer Potts sits down with Marketecture Media CEO Ari Paparo to talk about how Madhive is using AI to simplify the chaos of local digital advertising. From helping field sales teams close deals to generating creative and optimizing campaigns, Madhive’s AI assistant—Maverick—is designed to make life easier for sellers, advertisers, and media companies alike.

Spencer also shares how Madhive is expanding beyond its local roots, why they acquired Frequence, and how AI is uncovering audience insights that humans might miss.

Ari Paparo: AI, right? It's a fad, right? Next year we’ll be back to normal?

Spencer Potts: I wish it were a fad. When people start talking about AI, the first question is always, “Where are you using it?” The truth is, AI can touch so many parts of an organization. For us, it’s foundational—it underpins our entire platform. We’re building a series of products where AI drives different elements of our offering.

Ari Paparo: What does that look like in practice? What’s actually in production and being used?

Spencer Potts: Fortunately, Madhive had an early start with machine learning and AI, especially around local advertising. We figured out early on how to survey most of the bidstream through a distributed server network—basically, how to get reach and scale for our clients.

When we thought about building an operating system for local advertising, we asked: What can AI help optimize? We started with field sales—thousands of sellers at broadcasters and media groups that white-label our platform. They needed a sales coach, assistant, and ad ops partner in their pocket.

Ari Paparo: Because local sellers are used to selling a $50,000 TV spot on the evening news. Now it’s digital and fragmented—they need help.

Spencer Potts: Exactly. So we built Maverick as a sort of AI sales buddy. It helps sellers run comparative analyses, access performance data from past campaigns, and identify target audiences. The idea is to make the process less intimidating and more turnkey—especially for people new to digital advertising.

Ari Paparo: And what about the advertisers themselves? What do they want?

Spencer Potts: They want outcomes. That could mean more foot traffic, website visits, or lowering their customer acquisition cost. We’ve built a system where advertisers can choose key KPIs, and Maverick suggests strategies to hit those outcomes. It can even recommend ongoing optimizations during a campaign. Of course, they can override it—but often, the robots are right.

Ari Paparo: So what exactly is Maverick?

Spencer Potts: It has three main components. First, the sales assistant—similar to Gemini or ChatGPT—helps sellers build proposals and generate creatives. Second, the audience studio pulls in data from Madhive performance outcomes, enterprise-level learning systems, and client first-party data to build custom audiences. Third, we’re rolling out an executive coach tool. That’ll give business leaders visibility into their performance across markets—what’s working, what’s not.

Ari Paparo: Prompting media planners can be tricky though. They don’t always know what to ask, and the models can hallucinate.

Spencer Potts: Totally. We’ve been thoughtful about guardrails. Our product team looked at the biggest pain points and figured out where we could automate without introducing risk. And we’ve built in manual overrides for when things go off track. It’s not generalized AI—it’s purpose-built for the jobs our users need to do.

Ari Paparo: What about creative? That’s always a bottleneck.

Spencer Potts: We acquired Frequence, which has a creative studio. So now we’re generating thousands of display and video ads. Clients can upload their website or brand assets, and we’ll build creatives that are even included in proposals.

One fun example: we ran a campaign for organic dog food. Humans picked “dog lovers” as the target audience. The AI, on the other hand, recommended people who do yoga, eat organic, and shop at certain stores—five times the conversion rate. The machine found the behavioral signal we didn’t.

Ari Paparo: That's great. I also wanted to ask about Vica?

Spencer Potts: Vica’s a client—an AI creative studio. They generate an ad for a brand and then white-label Madhive to activate it. It’s a great model. We originally grew by white-labeling for broadcasters, and now we’re expanding into AI creative shops too.

Ari Paparo: So you’re not just doing local anymore?

Spencer Potts: We still specialize in local, but we enable national campaigns with localized execution. You can pick and choose markets—down to the zip code—and we’ll help you reach those audiences efficiently. That’s our competitive edge.

Ari Paparo: What’s next for Maverick and Madhive?

Spencer Potts: We’re rolling out Maverick in phases throughout the year. It's about helping sellers close deals, advertisers hit outcomes, and broadcasters understand what’s working in their markets. Long term, we’re even pointing Maverick inward—to improve internal enterprise intelligence. It's a bit Big Brother-ish, but it helps with operational alignment.