When the estimate shifts, can we clearly trace that back to a specific model revision or a specific user’s change? Not just that the cost changed, but exactly what triggered it and who made that change.