Marcus has a fixed $100k change-management budget and four levers — support, training, comms, incentives — and has to say how much goes to each. The change budget is a portfolio, not a checklist: every lever has diminishing returns, and incentives are an inverted-U that backfires once you over-pay (crowding out intrinsic buy-in). Allocate the ten $10k chunks to the highest marginal return — the equimarginal move reaches 84% adoption. The workhorse-only, the broadcast-everything, the even-split, and the just-pay-them instincts all underperform, and over-incentivising drives adoption DOWN. Where does the budget go?