How do you build a business case for a change?

Quantify the current cost of the problem first, because every other number gets judged against it. Then set out options including doing nothing, with costs, benefits, and the assumptions behind each. Separate cash savings from time saved, since time saved is only real if the hours get redeployed. Be conservative, because someone will check the numbers a year later.

I start with the cost of the problem today, because every other number in the document gets compared to it. That usually means volumes times handling time times a loaded hourly rate, plus whatever the errors cost, and I show the working so anyone can challenge an input rather than the conclusion. Then options, and I always include doing nothing, both because it is often a legitimate choice and because it shows I am not just selling. The distinction I am careful about is cash versus time. If we save four hundred hours a year, that is only a saving if those hours become something else, and on a team of five it is not a headcount reduction, so I present it as capacity rather than money. I would rather be conservative and be believed, because someone will come back in a year and check, and I want the next business case I write to be trusted.