Decades and billions spent measuring a known problem we’ve yet to fix. The highest-leverage variable remaining in dairy is a sneaky one because we’ve known about it for generations — these data are visible through milk yield and quality metrics, short-term health measures like blood metabolites, conversion metrics like rumen-pH, long-term measures like body condition — all captured by advanced wearables that add their cow-diagnostic data to the decades of global herd data we now constantly analyse with artificial intelligence. An entire modern toolset exists to describe the problem beautifully, and each year billions more are spent managing its symptoms —ration adjustments, pharmaceutical and surgical interventions, early culls, breeding positive heritable traits into replacements — yet almost none of it addresses the cause at its source. For all our progress to optimise our cows and everything around them, the one variable left almost untouched is what she does herself — in the 20+ hours no one is watching. This was not left alone through neglect, but because it seemed impossible to reach. More importantly, what matters most is what cows do in a specific, short, predictable, weeks-long window that repeats every year, for every productive cow, on every farm in the world.