The conversation with the person who built it

Everything up to here is technique and can be learned in a weekend. This module is the one that decides whether any of it happens, and it is difficult because the conversation is not really about the business. To the person who built it, the business is a record of their judgement over thirty years, and a proposal to change it is heard as a verdict on that judgement, however carefully you phrase it.

What works, reliably. Bring a number rather than an opinion — "we are waiting fifty-two days to get paid" is a fact you can stand next to, while "our systems are outdated" is a criticism of a person. Ask about history before proposing anything, and mean it, because you will genuinely be wrong about a quarter of the time and finding out early is cheaper than being corrected publicly later. Propose it as a trial with an end date, so agreeing costs them nothing permanent. Name what you are not changing, because the loudest unspoken fear is that this is the first of a hundred changes. And give the credit away when it works — you will get more of what you want and it costs you nothing you needed.

What does not work: comparing them to another business, invoking what you were taught, changing something first and explaining afterwards, or making the case in front of staff. And accept that some things you will not win now. A person who has run something for three decades is allowed to keep a few decisions that are theirs. Winning ninety percent of what matters, slowly, beats being right about all of it and being shut out.