Home > Business

How a Successful Marketing Agency Is Transforming Its Business With AI

Artificial intelligence has a lot of marketers worried it’ll upend their business, but Erik Huberman sees an opportunity for exponential growth.

Photo via Pexels

In 2017, Erik Huberman, the founder and CEO of marketing agency Hawke Media, was serving on the advisory board for XPRIZE, the celebrity-heavy organization known for its multi-million-dollar competitions to reward breakthrough technological development. There, the organization’s then-CEO, Marcus Shingles, asked him a provocative question: “What’s going to disrupt your company?” Shingles thought it would be the rise of the gig economy, in which freelancers would displace traditional professional service companies like Hawke.

Instead, Erik answered “AI.” 

These days, the arrival of ChatGPT and other sophisticated bots would make that answer highly sensible and even expected. Not long ago, though, the idea that artificial intelligence could replace a marketing agency seemed pretty far-fetched. But Erik had given it a lot of thought and realized that even before AI could write ad copy, it could automate a lot of what marketers do—especially if it had the right data to work with. 

Related ChatGPT is the Deepfake of Thought

“Right, great,” Shingles countered. “So why aren’t you building it?”

Erik must have recalled that conversation in the fall of 2021, when he attended the Birthing of Giants Fellowship Week, which I run with entrepreneur whisperer Norm Brodsky and a team of business experts. There, one of the most important questions we ask the gathered entrepreneurs is, “How will you achieve exponential growth?” Erik spent the week trying to answer that question for himself, and he kept coming up with one answer: software. 

Now, Erik is a smart, energetic, and driven entrepreneur, so this isn’t the first time the idea occurred to him. In fact, he tried to bring a software solution for marketing agencies to market three times previously over the past several years, spending hundreds of thousands of dollars in development costs with little to show for it. But in the intervening time, AI and machine learning technology advanced considerably and—even more importantly—Hawke Media had been collecting anonymized data from its clients’ marketing campaigns (with their agreement, of course). In fact, by 2021, it had nearly 10 years’ of data from 4,000 companies. You see, Erik knew something that a lot of business leaders are only beginning to grasp: His most valuable asset was his company’s data. 

Back at Fellowship Week, Erik knew it was time to pull the trigger. Right there during the session, he messaged his partner to advertise a job opening for an AI developer. 

Related Building on Black America’s Legacy of Resilience: A FinTech Blueprint

As it turns out, he got something much, much better. Serendipitously, a Canadian company called Morphio sent out a cold email pitching its latest product: an agency dashboard built to monitor multiple marketing channels. Erik realized that this was the ideal interface for his own software project, but it lacked the most important ingredients: first, historical data that would train a machine-learning algorithm to benchmark performance, and second, a flow of real-time data that the algorithm could use to identify weak spots that needed addressing. “This is stuff we do manually already,” he told me recently. “The question is, how do I automate that?”

As is his habit, Erik moved quickly. By February 2022, he had a deal to acquire Morphio, and just a few months later, the company launched the first version of its new software-as-a-service (SaaS) product, dubbed Hawke AI, with 2,000 subscribers paying $99 per month each. That might seem like a drop in the bucket for an agency whose typical engagements reap about $7,000, but there are a few key reasons why this revenue stream could quickly outstrip the agency’s more traditional service offerings. First, SaaS products are cloud-based, which makes them highly scalable to serve millions of customers concurrently. Second, Hawke AI’s total addressable market is multitudes larger than that of the agency. Hawke Media prided itself on offering affordable marketing to smaller companies, but at $99 per month, the entry barrier fell even lower. Third, while more agency clients mean higher costs in payroll and other expenses, the math for SaaS is far more attractive. Adding a new paying customer costs just pennies in additional cloud computing and storage fees, so the more customers you have, the higher your profits. And unlike adding headcount, these fees scale up and down smoothly, so you’re never paying for capacity (or benefits, vacation, and overtime) you’re not using. 

Related Leading Voices: Financial Literacy Produces Both Capital and Courage

For Erik, this launch is just the beginning of the journey. Hawke AI has proved highly accurate in its ability to predict marketing and advertising performance across multiple channels, but he knows it can do so much more—like offering users specific advice on optimizing their marketing spend, and eventually, doing that optimization by itself automatically. 

Is your company leveraging the huge volume of data it pulls in from sources such as sales, purchases, and advertising performance? Is it correlating that with outside data—such as weather or other demand signals to improve accuracy? If not, you may be leaving money on the table. And if you are making the most of this data internally, how can you turn that around with a product or service that could become a moonshot? Let me know in the comments!

Lewis Schiff runs the Birthing of Giants Fellowship Program, a one-week guided strategic planning process that’s attended by the owners of the fastest-growing companies in the world. He also runs Moonshots & Moneymakers: The Oxford Innovation Conference for American Entrepreneurs

Want the inside scoop on how Erik Huberman and other entrepreneurs built their companies from moneymakers to moonshots? Sign up for Birthing of Giants’ free monthly video series, How I Did It.

Related Articles