Ownership, governance, IP, pricing and the exit. Get these on paper early and the rest of the deal gets easier.
Most cross-border joint ventures that fail did not fail on the commercial logic. They failed on a question that both sides assumed had an obvious answer, and which turned out to have two obvious answers.
Five are worth settling before anything is signed.
Who owns what, precisely. Not the percentage — that part is always agreed — but what happens to it. Whether shares can be transferred, to whom, at what price, and what occurs if one side wants to bring in a third party. A split that is fair on day one is worth little without a mechanism for day one thousand.
How decisions get made when the two sides disagree. Equal shareholdings feel fair and deadlock beautifully. Either someone has a casting vote on defined matters, or there is a route out that both sides can live with. "We will work it out" is not a governance provision.
Who owns the intellectual property, including the intellectual property that does not exist yet. The process improvements made inside the venture, the tooling designed for it, the customer data it generates. This is the question most often left to the end and most often fought over later.
How related-party pricing is set. If one partner also supplies the venture, or buys from it, the transfer price is where the economics actually live. Agree the method — not this year's number — and agree who can audit it.
How it ends. Every venture ends: in a sale, a buyout, a wind-down or a fight. Naming the exit early is not pessimism, it is the single clause most likely to keep the relationship civil, because both sides negotiate it while they still like each other.
None of this replaces good lawyers, and none of it is exotic. What it does is force the conversations that reveal whether the two businesses actually want the same thing — which is cheaper to discover across a table than across a tribunal.