Darin McOwen

How I operate

Twenty years running growth for consumer and enterprise businesses, most of it in the band between $30M and $300M — the stage where the engine that got a company here will not get it to the next number.

My work sits where marketing meets machinery. At Vitalpax I built an enterprise-grade ML predictive-bidding platform from thirty data sources, deployed demand forecasting against it, and took inbound leads up sixfold — enough to create a six-month production backlog. Before that I ran an affiliate and paid engine doing 1–2K direct-to-consumer orders a day, and won shelf at Sprouts, Whole Foods, TJ Maxx, Macy's, HSN, and Wayfair.

I have also been the one signing the front of the cheque. I bootstrapped Auto Parts Place to eight figures across 50 warehouses and 45 sub-brands, then sold it to a NASDAQ buyer. That is the difference between advising a P&L and owning one.

What I look for now: a consumer or B2B software business where AI is not a slide but an operating change, and where the marketing function is expected to carry a number.

What I'm known for

Beyond the numbers

Three things a list of roles never shows.

I started as a founder, not a marketer. My first company was bootstrapped — no budget, no team, no brand to borrow credibility from. That is where I learned to read a P&L, and why I'm sceptical of any plan that only works if everything goes right.

I build rather than brief. Claude, OpenAI, and Gemini are in my day-to-day, not in a pilot deck — agentic workflows orchestrated over MCP, retrieval running against our own data, and eval suites that catch a model drifting before a customer does. Underneath that: Python, BigQuery, Vertex AI, TensorFlow. The predictive-bidding platform exists because I could sit inside the data instead of writing a requirements document about it — and it's why I can tell when a vendor's AI story is a wrapper.

I publish. Most weeks I write about what agentic AI actually changes for marketing leaders — the uncomfortable parts, not the exciting ones. It's a useful discipline. A claim made in public is a claim somebody can check.

Recognition

• Utah Business Executive Excellence Honoree, 2026
• Microsoft Certified: AI Transformation Leader (in progress)
• B.A. Business Management, cum laude — California State University, Fullerton, with business law coursework at the J. Reuben Clark Law School, Brigham Young University

Writing

Notes on what agentic AI actually changes for marketing leaders — the uncomfortable parts, not the exciting ones.

• Agentic AI makes brand falsifiable for the first timeWhen the buyer is an agent, every soft brand claim becomes a testable input with a pass or fail.
• LinkedIn isn't penalizing AI contentThe platform demotes generic content. AI is incidental — it just made generic cheap enough to flood the feed.
• Your brand guidelines are not a brand codeBrand guidelines tell humans how to decide. A brand code tells an agent what to do when no human is in the loop.
• Publicis bought the yardstickLiveRamp's RampID was the neutral referee grading your agency bakeoffs. Switzerland just signed with one of the teams.
• Your agent is negotiating with a black box that gets paid when you spend moreEvery major ad platform now ships an MCP server. The bottleneck was never the clicking — it was the judgment.