Two clocks on the same wall
The most reliable predictor of partnership friction is a horizon mismatch that neither side surfaced. One parent treats the venture as a three-year experiment that must prove itself each budget cycle; the other is building a ten-year platform and prices early losses as investment. A partner owned by a fund has a clock imposed by fund life, whatever its executives say in the room. Exit expectations are the same conversation wearing different clothes: if one side privately expects to sell out and the other privately expects to buy them out, the venture will run smoothly right up until it matters.
Contribution maths gets renegotiated by resentment
Cash is countable. Brand, channel access, IP, data and seconded people are valued in the optimism of negotiation and revisited in the friction of performance reviews. The partner contributing capital tends to conclude, around year two, that it paid real money for assets the other side would have brought anyway; the partner contributing capability concludes its inputs made the venture and were priced as furniture. Neither view needs to be right to be corrosive. Ventures that survive this build revaluation into the structure (milestone-based vesting of equity, contribution audits, pre-agreed adjustments) rather than trusting the founding valuation to age well.
Deadlock will not be solved in the room. Design it on paper.
Shared control means the venture can be stopped by disagreement, and disagreement is certain over any horizon that matters. The instruments are well known: reserved-matter lists kept genuinely short, escalation ladders with named roles and deadlines, cooling-off periods, and buy-sell mechanisms of last resort that make deadlock expensive for whoever provokes it. What distinguishes ventures that endure is not the choice of instrument but the timing: deadlock machinery negotiated at formation, between optimists, is fair; the same machinery negotiated during the first real dispute is a hostage exchange.
Whose salespeople win the overlapping account?
Channel conflict is the failure mode partners most consistently wave away at signing. The venture will, if it works, sell into accounts one or both parents also serve. On the day that happens, two sales teams with different incentive plans will claim the same revenue. Decide the rule now: which opportunities belong to the venture, how the parents' sales forces are compensated when they hand over an account, and what happens when a parent's own product line starts competing with the venture it half-owns. If the answer is left to goodwill, the venture loses, because the parents control the goodwill.
Some partnerships are acquisitions with a waiting period
A venture that gives one side deep access to the other's technology, customers and economics is, functionally, free due diligence. Option-to-buy clauses, staged equity escalators and asymmetric information rights are legitimate structures, provided both sides can see what they are. If you suspect your counterpart is testing before buying, the response is not to refuse but to price it: negotiate the purchase mechanics now, at a moment of mutual dependence, rather than in year three when they have learned your business and you have restructured yours around them.
Run these tests before anything is announced
- Have each side write down, separately, what success looks like in year five and what would trigger exit, then compare the answers before lawyers draft anything.
- Model the venture failing and agree who gets the people, the customers and the IP on the way out.
- Name the accounts where the venture and either parent will compete, and set the rule now.
- Test whether a supply, licensing or reseller agreement would achieve the goal without shared equity.
- Ask which side would buy the other out in year three, and at roughly what kind of price, and watch the reaction as closely as the answer.
