Great creative sits at the intersection of art and analysis. The "right-brained" creative instinct carries just as much weight as the "left-brained" analyst — every hook, script, and static should be as emotionally compelling as it is data-informed.
I start with the consumer pain point and the paid social data, then build creative that earns the thumbstop in the first two seconds, communicates a clear value proposition, and gives the audience a reason to click. Insight drives the concept; performance data sharpens it.
Creative is never "done" — it's a living system of hypotheses. The job is execution: ship fast, learn faster, double down on winners, retire what doesn't work, and make every batch stronger than the last. That's what separates ideation from ownership.
Every test begins with a clear hypothesis tied to a consumer insight or a performance gap — e.g. "a problem-first hook will outperform a product-first hook for cold audiences."
Test one variable at a time — hook, format, angle, or offer — so results are attributable. Keep everything else constant to avoid muddy learnings.
Use a clean testing structure (ABO/CBO) with enough budget and runtime to reach statistical significance before making decisions. Define spend and time thresholds up front.
Read creative through the full funnel: Hook Score / thumbstop, Hold (Watch) Score, CTR, landing page CVR, CPA/CPL, and ROAS — not just top-line spend.
Scale winners, iterate on the winning variable to find the next lift, and continuously refresh creative to fight fatigue — feeding learnings back into the next batch.