Thirteen months before this conversation, Peter Murphy Lewis had never produced a documentary. By the time he sat down with Ryan Roghaar on EGGS, he had three behind him and more in production, and he was describing documentary work as the part of his business that AI cannot touch. That is the thread Roghaar pulls on for the next forty minutes.
Roghaar hosts EGGS and runs a creative shop of his own, and he had just come back from a London conference full of small agency owners trying to work out what AI does to their model. So the episode is two practitioners comparing notes rather than a founder answering questions. They land on the same worry from different directions: not that AI produces bad work, but that it produces the same work for everyone.
The useful part is how specific Lewis gets about where AI sits in his process now compared with a year ago, and what he moved back into human hands after watching the output. He also makes the case that the businesses best placed to gain from AI are the least fashionable ones: brick and mortar operators whose competitors are doing nothing at all. For more on this, see our article on using AI without sounding generic.
Key Takeaways
1. AI Sits in the Middle of the Work, Not at Either End
A year ago Lewis used AI at every stage, from ideation through execution to analysis. Now the thinking and the finishing are both human. He ideates himself or with a strategist, hands that strategy to AI to develop into content, then brings a person back in to interpret the data. He puts it in American football terms: he calls the play, AI helps decide which gap to run through, and a human handles the contact. “AI is somewhere in between the ideation and the tackle point for me,” he says.
2. Two Car Washes Following the Same AI Advice Will Both Fail
His argument is not that the output is bad but that it is identical. A large share of what these tools return, he believes, was curated from average SEO and blog content published over the previous five years. Imagine two people opening car washes on opposite sides of one town, both taking the same advice: they will do the same good thing at the same moment and neither will stand out. He repeats a line he hears in SaaS: “it’s better to be different today than better.”
3. Documentary Work Now Carries Most of His Revenue
Lewis says his rule is simple: use AI for everything it can do, then lean hard into whatever it cannot do even 80% of. For him that means brand, real content rather than generated content, podcasts and documentaries. On the episode he puts documentaries at roughly 75% of his revenue over the past twelve months and expects that share to rise the following year. His reasoning is demand-side as much as supply-side. As output multiplies, “people also want more human interaction and more human connection.”
4. A Journalist Can Tell When a Pitch Was Written by a Machine
Over the past eighteen months his team has tripled down on earned media, pitching journalists for coverage rather than just chasing backlinks, partly because national mentions now feed AI answers too. The workflow mirrors his wider rule. The quarterly pitch angle is designed by people. AI is used to identify the journalists raising their hands for that kind of story. Then a person writes the actual pitch, “because journalists can pick up on AI.” Machine in the middle, humans on both ends.
5. Brick and Mortar Is the Least Contested Ground for AI
Asked where the real opportunity sits, Lewis names two places. The first is brick and mortar, “mostly because the competitors haven’t used anything with AI.” No automated customer conversations, no identifying who is on the website and reaching them before they walk in. The second is any sector drowning in data. He sits on the board of a bank and points to insurance as well, along with his own YouTube analytics, which he now feeds into AI to generate ideas worth testing.
6. Being the Most Helpful Person in the Room Is a Marketing Strategy
The advice Lewis gives his nine-year-old son is the advice he gives founders. “If you are the most helpful person in the room,” he says, you will always have a job, friends and neighbors. Applied to business, it means learning the fundamentals instead of chasing whichever tool is new this year. Roghaar offers his own version of the same idea, a phrase his agency uses: you have to break bread to make bread. Lewis takes it and runs with it for the rest of the conversation.