Ask people why a business partnership ended and the answers tend to be vague. Priorities diverged. The fit was not right. Circumstances changed. Press further and something more specific usually emerges, and it is almost never the thing either party would have listed as a risk at the start. The dispute was about how a windfall should be split, or about who was entitled to decide something neither had thought to assign, or about a level of effort that one side considered obvious and the other had never agreed to.
Partnerships fail in the gaps. Not in the areas that were negotiated, where both sides understood they were negotiating, but in the wide territory that was left unstated because raising it early would have felt like distrust. The strongest partnerships are not the ones with the most goodwill at the outset. They are the ones where the awkward conversations happened first.
This applies across every form the relationship takes: a joint venture, a supplier arrangement that both businesses depend on, a co-ownership, a long term client relationship that has become structural for both sides. The mechanics differ. What holds them together does not.
Goodwill Is a Result, Not a Structure
Mutual regard is genuinely valuable and it is the wrong foundation. Goodwill is a stock that depletes under pressure, and pressure is exactly the condition under which a partnership has to work. A relationship carried entirely by how well two people get on will function beautifully for years and then fail during the first period when both are under strain and neither has capacity for generosity.
What survives that period is structure: a shared understanding of who decides what, what each side is responsible for contributing, how money moves, and what happens when something goes wrong. Structure does not replace goodwill. It protects it, by removing the situations where goodwill has to be spent on questions that could have been settled in advance.
The confusion arises because well structured partnerships appear informal from outside. The parties rarely refer to their agreement, decisions get made quickly, and there is little visible process. That ease is the product of the clarity, not evidence that the clarity was unnecessary.
The Conversations Worth Having Before the Agreement
Two conversations do most of the work, and both are more comfortable to have before there is anything at stake.
What is each side actually optimizing for
Partners often assume shared objectives because they share an activity. In practice one party may be building an asset to hold for twenty years while the other is building something to sell in five. Both are legitimate. They imply completely different answers about reinvestment, about how much risk is acceptable, and about whether short term contribution or long term position matters more.
That divergence is manageable when it is known and corrosive when it is not, because every subsequent disagreement will be read as bad faith rather than as the predictable consequence of different horizons. The question is simple to ask and rarely asked directly: what does a good outcome look like for you, and when do you need it.
What happens if this works better or worse than expected
Most agreements are calibrated to the expected case. The expected case is the least likely of the three. A partnership that has not discussed what happens if the venture substantially outperforms is exposed, because unexpected success creates a genuine question about whether the original split still reflects contribution, and that question is far harder to negotiate once the money exists.
The downside conversation is equally important and equally avoided. If additional capital is required, who provides it, on what terms, and what happens to the position of a party who cannot. Left unaddressed, this single question has ended more otherwise sound partnerships than any other.
Incentives That Hold When Conditions Change
An incentive structure is a prediction about behaviour, and predictions decay. Terms set when one party was contributing capital and the other was contributing time can become badly misaligned once the capital is deployed and the time commitment has changed shape.
The failure is rarely dramatic. It shows up as one side becoming slightly less responsive, or as decisions that make sense for one party and are marginally costly for the other. Nobody is behaving badly. The arrangement is simply now rewarding something different from what it was designed to reward.
Partnerships that last tend to have a scheduled point at which the terms are examined, independent of whether anyone is unhappy. Reviewing an arrangement while it is working is a very different conversation from raising it once resentment has formed, and the difference in outcome is substantial. The scheduled review also removes the need for either side to be the one who brings it up, which is frequently the actual barrier.
What makes the review productive is agreeing in advance what would justify a change. A partnership that has said, at the outset, that the split should reflect contribution and that contribution may shift, has given itself permission to revisit terms without either party appearing opportunistic.
Asymmetry Is Normal and Should Be Named
Very few partnerships are balanced. One party usually needs the relationship more, holds more of the operational knowledge, controls the customer relationship, or has better alternatives. Pretending otherwise does not remove the asymmetry, it just prevents anyone from designing around it.
Named asymmetry becomes manageable. If both sides acknowledge that one party would find replacement difficult, that fact can be addressed directly through the terms, through notice periods, or through the more dependent party deliberately reducing the dependency over time. Unnamed, it sits underneath every negotiation and gradually turns a working relationship into one where the weaker party feels obliged to accept things it should not.
There is a related point about concentration. A partnership that becomes a large proportion of one side’s business has changed from a relationship into an exposure, and it should be assessed as one. The party carrying the concentration needs to know that it is carrying it and to decide deliberately whether to reduce it, which connects to the broader practice discussed in How Resilient Enterprises Prepare Before Conditions Become Difficult.
Handled openly, asymmetry does not damage a partnership. Many of the most durable arrangements are markedly unequal in bargaining power and work well for decades, because the stronger party understood that a relationship the other side cannot afford to leave is not the same as one they want to keep.
Contribution That Is Not Money
Financial contribution is easy to record and therefore easy to honour. Everything else gets remembered unevenly. A partner who introduced the relationship that became the largest account, who carried the operation through a bad eighteen months, or whose name is the reason a lender agreed to terms, has contributed something real that appears nowhere in the ledger.
The difficulty is that these contributions are front loaded while capital contributions continue. A party who provided the original access may find, five years later, that their role is described entirely in terms of what they do now, and the introduction that made the business possible has quietly stopped counting. The resentment this produces is usually attributed to personality when the cause is structural.
Two things help. The first is writing down, at the outset, what each party is understood to be bringing, including the parts that are not financial. It costs an afternoon and it fixes a shared record before anyone has reason to remember it differently. The second is distinguishing between contributions that were made once and those that must continue. A relationship introduced ten years ago is a permanent contribution to the enterprise’s existence. Ongoing account management is not the same thing, and conflating the two in either direction causes trouble.
Where the imbalance is genuinely large and continuing, the honest response is to adjust the arrangement rather than to rely on one party’s tolerance. Partnerships that depend on someone accepting less than they are owed are stable only while that person’s circumstances permit generosity.
Disagreement as a Working Process
The common belief is that good partnerships have fewer disagreements. Observation suggests the opposite. Partners who work closely on things that matter disagree regularly. What distinguishes durable relationships is that disagreement has somewhere to go.
Having somewhere to go means several concrete things. It means both parties can raise a problem early without it being treated as an escalation, which requires a habit of routine contact where nothing is wrong. It means there is an agreed way to settle a question when discussion does not resolve it, whether that is a designated decision maker for particular categories, a third party both respect, or a defined process. And it means disagreements get addressed on the specific matter rather than accumulating into a general grievance.
The accumulation problem deserves attention because it is the usual mechanism of failure. Small unresolved frictions do not stay small and separate. They combine into a narrative about the other party’s character, and once that narrative forms, every subsequent event is read through it. A partner who is three days late with information becomes evidence of a pattern rather than a person who had a difficult week.
Partnerships that handle this well tend to have made it normal to say something at the time, in proportion to the matter. That is a cultural property more than a contractual one, and it is established early or not at all.
Reliability in the Small Commitments
Trust between businesses is built almost entirely through the accumulation of minor kept commitments. Information arriving when it was promised. A figure that turns out to be accurate. A concern raised before it becomes a problem for the other side.
This matters more than the large gestures because it is continuous and because it is a genuine signal. Anyone can perform well at a signing. Sending the reconciliation on the day it was due, for four years, indicates something about how the business runs, and it is the kind of information that cannot be faked over any meaningful period.
The reverse holds too. A partner who is consistently slightly unreliable in small matters is providing accurate information about what will happen in a large one, and businesses routinely ignore that signal because each individual instance is too minor to raise. The instances are minor. The pattern is not.
Reliability of this kind is also the mechanism by which a professional reputation is built across a career. Norman Ebenstein’s long involvement in commercial real estate reflects an environment where relationships frequently outlast individual transactions, and where the ability to be relied upon over decades carries more weight than the terms achieved in any single negotiation.
Designing the Ending While Everyone Is Content
Discussing how a partnership would end feels like doubting it, which is why it is usually skipped. The absence of an exit path is what turns an ordinary divergence into a damaging one, because parties who cannot leave cleanly are forced to fight instead.
A workable exit provision does a specific job. It establishes how a departing party’s interest is valued, over what period it is paid, what happens to shared customers, assets, and staff, and what either side may do afterward. None of this is complicated to agree while both parties are content and neither knows which of them will want out.
Agreed early, it also improves the partnership while it is running. Knowing that leaving is possible and would be handled fairly means neither party is participating out of entrapment, and both can raise difficult subjects without the implicit threat that the whole arrangement might collapse. Partnerships with clean exits tend to last longer than those without, which is the opposite of what most people expect.
Where a partnership is intended to outlast the individuals who formed it, the design question extends further into how the relationship itself transfers, which is part of the broader subject of What Allows an Enterprise to Create Value Across Generations.
Decision Rights Between Two Businesses
Within a single company, unclear authority causes delay. Between two businesses it causes disputes, because each party fills the gap with its own assumption and both assumptions feel obviously correct to the party holding them.
The practical fix is to sort decisions into categories before they arise. Matters one party decides alone and simply reports. Matters requiring consultation but where one party still decides. Matters requiring genuine agreement, kept deliberately short. Most partnerships that struggle with authority have made the third category too large, which produces a relationship where nothing moves quickly and both sides feel constrained by the other.
Getting this right inside each partner organization matters as much as getting it right between them, since a partner who cannot obtain a decision from their own side is functionally unreliable regardless of intent. The internal version of this problem is examined in The Role of Governance in Keeping Enterprise Decisions Clear.
What Durable Partnerships Look Like From Outside
A partnership that has been working for fifteen years tends to look unremarkable. The parties talk often and without ceremony. Problems get mentioned early and in ordinary terms. Neither side appears to be monitoring the other, and the formal agreement is rarely consulted.
All of that ease was purchased with specific work done early: expectations stated plainly, incentives designed to survive changes in conditions, asymmetry acknowledged, a route for disagreement, and an ending that both sides would consider fair. The relationship feels informal because the structure underneath it is sound enough that nobody has to think about it.
The alternative, a partnership carried by mutual regard and left otherwise undefined, feels equally comfortable at the start and considerably less so in year six. Selecting the right counterparties in the first place is its own discipline, related to the broader question of commitment examined in How Durable Enterprises Decide Which Opportunities to Pursue, but selection alone does not carry a relationship. The design does.
