What can the twin predict?

Revenue, spending, ADP execution and e-GP award counts, each backtested on months it never saw. Where a model does not beat last year’s same month, the forecast is withheld and the reason is shown.