Why Marketers Should Treat AI Like a Frenemy

Introduction

Charlie Gri ell, Co-CEO of RightMetric, brings a refreshing perspective to the AI conversation that every marketer needs to hear. With experience leading marketing at Red Bull and Aritzia, Gri ell now helps brands like Meta and lululemon leverage external data signals to make smarter strategic decisions. His core insight? AI creates more work in the short term before it delivers automation benefits—and most organizations aren't prepared for that reality.

The Hidden Cost of AI Implementation

According to Gri ell, organizations get caught up in AI's "shiny object" appeal without considering the foundational work required. "We have all of these grandiose long-term possibilities and vision, but there's a short-term pain," he explains. Before any automation can happen, companies need proper data infrastructure, documented processes, and cross-functional buy-in. Think of it like going to the moon—everyone's excited about the destination, but someone needs to ask about the launch pad, spaceship, and oxygen supply. These unsexy but mission-critical elements determine whether AI initiatives succeed or fail.

Building AI Maturity Through Incremental Steps

The path from using ChatGPT for emails to having true AI agents follows three distinct stages. First comes manual AI assistance—using chatbots for individual tasks. Next is workflow orchestration, where AI handles entire processes. Finally, organizations reach true automation with feedback loops and iterative improvement. The key to progression? Start by writing detailed SOPs (Standard Operating Procedures) as if explaining tasks to a human assistant. This forces you to extract the context trapped in your head and document it clearly. Gri ell emphasizes breaking down complex processes into micro-tasks rather than attempting massive automation projects.

The Power of External Data

One of Gri ell's most compelling insights challenges the current obsession with first-party data. He shares a story from his time at a fashion brand where they celebrated 20% growth—until the CEO asked how much the category grew. "If the category grew 5%, I should give you a raise. But if the category grew 80%, I should fire you right now," the CEO said. This illustrates why external data matters: without competitive context, you're marking your own homework. Tools like SparkToro, SimilarWeb, and Tubular Labs provide directional accuracy (within 5-20%) that's far more valuable than being completely blind to market dynamics.

Making AI Your Ally Through Better Prompting

The secret to better AI outputs isn't complex prompting courses—it's clear thinking translated into clear writing. Gri ell advocates for always asking AI to "walk me through your thinking and how you did that" to uncover potential hallucinations and improve results. He saves all his AI conversations rather than deleting them, building a valuable context library for future projects. When building agents, he tests ideas in Claude or ChatGPT first, iterating on the approach before implementing automation. This methodical approach transforms AI from an unreliable tool into a trusted partner.

Conclusion

The relationship between marketers and AI will remain complicated, but it doesn't have to be adversarial. By focusing on data organization, starting with clear documentation, and maintaining human judgment in the loop, marketers can harness AI's potential while avoiding its pitfalls. As Gri ell reminds us, the tools will constantly evolve, but the fundamentals—clear thinking, good data, and systematic approaches—remain constant. The marketers who succeed will be those who embrace AI as a powerful but imperfect partner, investing in the unglamorous foundation work that makes true automation possible.