Why Business Partnerships Fail and the Behaviors That Destroy Them

Business partnerships rarely fail over a single bad quarter. They fail when four things drift apart, one partner's effort and the other's, the money each puts in and takes out, who gets to decide, and whose interests come first, and there is nothing written down to pull them back together. The fifth cause is the one that turns a disagreement into a disaster: no agreed way out.

The law does not rescue a partnership that never wrote its terms down. Under the Revised Uniform Partnership Act, which most states have adopted in some form, the default rules split profit equally however hard each partner works, pay no partner a salary, and let any partner in an open-ended partnership leave by giving notice (California's version, sections 16401 and 16601). Those defaults are reasonable for strangers in a textbook. For two owners whose contributions have drifted apart, each one is a grievance waiting for a date.

What follows is how each failure happens, the behaviors that destroy a partnership from inside, the red flags that show up before anything is signed, and what the law lets a partner do once things have gone wrong. It is general information; partnership law varies by state, and a lawyer in yours should read the actual agreement.

Claims that "seven in ten business partnerships fail" circulate widely without a study behind them that can be traced. The better data is narrower. Carta tracked two-founder teams at more than 45,000 US startups and found that fewer than a quarter split up within four years, while more than 60% were still together after eight. Early breakups are becoming more common: the share of two-founder teams that parted in their first year nearly doubled between 2015 and 2024, from 4% to 7.5% (Carta, Founder Ownership Report).

Carta counts only companies still operating, so a partner leaving that killed the business does not appear in it, and its companies are venture-backed rather than the two-owner firms most partnerships are. Read it as a floor. Most of these teams survive their first years; the ones that do not usually end for the reasons below.

1. Effort and reward drift apart

A familiar complaint between partners is that one is doing more. In a partnership with no agreement, that is not a feeling the law can fix: each partner is entitled to an equal share of the profits, and no partner is entitled to pay for services to the business except for winding it up (Cal. Corp. Code 16401(b) and (h)). A partner who works 60 hours a week and one who works 15 split the profit evenly and draw no salary. Every month that continues, resentment compounds.

The fix is written at the start: a salary or guaranteed payment for the partner doing the operating work, paid before profits are split; a defined time commitment; and a rule for what happens to a partner's share if they stop meeting it. For a company with shares, vesting does the same job; the co-founder equity split guide shows the arithmetic.

2. Money going in and coming out

Partners fall out over capital calls (one can afford another $50,000, the other cannot), over draws (one takes cash out to live on, the other leaves it in), and over a partner quietly funding the business from personal savings. The default rule treats a payment beyond a partner's agreed capital as a loan to the partnership that accrues interest (16401(d) and (e)), which surprises the partner who thought it was a gift and the one who thought it bought a bigger share. An agreement that names each partner's capital, how and when more can be required, what happens if a partner cannot meet a call, and how draws are taken removes most of these fights before they start.

3. Decisions made alone

By default, a difference over an ordinary business matter is decided by a majority of the partners, while anything outside the ordinary course, and any change to the partnership agreement, needs every partner's consent (16401(j)). With two partners, majority means both. A partner who signs a lease, hires a manager or takes on debt without the other is often within their power to bind the business to outsiders while breaking the deal with their partner. Partnerships that last write down which decisions either partner can make alone, which need both, and how a deadlock is broken.

4. Loyalty: side deals and taken opportunities

Partners owe each other a duty of loyalty: to hand over any benefit taken from partnership business or information, including a partnership opportunity, not to deal with the partnership on behalf of someone with an adverse interest, and not to compete with it before it dissolves (16404(b)). The agreement can define what does not count as a breach, but it cannot remove the duty (16103(b)(3)).

The leading case is a century old and still taught. In Meinhard v. Salmon (1928), Walter Salmon managed a 20-year lease on the Hotel Bristol in Manhattan, financed by Morton Meinhard. As it expired, the landlord offered Salmon a much larger redevelopment lease; Salmon took it in his own name without telling Meinhard. New York's highest court held that Salmon owed his partner "the punctilio of an honor the most sensitive" and awarded Meinhard a 49% interest in the new lease (Meinhard v. Salmon). Most loyalty disputes are smaller: a partner steering a client to a side business, or buying from a company a relative owns. They end partnerships for the same reason.

5. No way out

Every partnership ends, by sale, retirement, death, a falling-out or a better offer. The ones that end badly are the ones that never agreed how. In a partnership with no fixed term, a partner can leave by giving notice (16601(1)), and the business must then buy out the departing partner's interest at the greater of its liquidation value or its value as a going concern without that partner, with interest (16701(b)). Nobody agreed what that value is, so it becomes the next dispute. A buy-sell clause with a valuation method, a payment schedule and the triggers (death, disability, retirement, deadlock) settles it in advance; how to buy out a business partner covers the mechanics.

The causes above are structural. Day to day, they show up as behaviors, and the partner on the other side usually notices months before saying anything. Each of these is also something the default law addresses, which is useful to know when the conversation finally happens.

Sources: California Corporations Code sections 16401, 16403, 16404 and 16701. Other states' versions of RUPA use similar wording; an LLC is governed by the state's LLC act and its operating agreement instead.

Two behaviors do not appear in any statute and end as many partnerships as the rest: mixing personal and business money (paying personal bills from the business account, or the reverse), and avoiding the hard conversation until it arrives as a lawyer's letter. The first makes every later number disputable. The second turns a fixable problem into a dissolution.

Most of the five causes are visible in the weeks of talking before an agreement. The warning signs worth stopping for:

• Contributions that stay vague. "I'll bring the clients" or "I can put some money in" that never becomes a figure, a date or a number of hours. Ask each side to write down, separately, what they will contribute in the first year, then compare the two lists.
• Resistance to writing it down. "We trust each other, we don't need lawyers" is a preference for leaving the terms open, and open terms get interpreted later by whoever wants something.
• Pressure to sign now. A deal that cannot survive a week of checking references and reading the agreement is a deal built on not being checked.
• Every former partner was the problem. One failed partnership can be bad luck. A pattern, told the same way each time, is information about the person telling it.
• No interest in how it ends. A prospective partner who will not discuss death, disability, a buyout price or a deadlock has not thought about the commitment, or would rather not be bound by an answer.
• Different appetites for risk and pace. One wants to borrow to grow; the other wants no debt. Neither is wrong. Unspoken, the difference resurfaces in every decision.

None of these is an accusation. A partner worth having expects to be checked and will check back.

A partnership in trouble has more options than a lawsuit, and they are best tried in order.

Read what was signed. Most disputes turn on a clause someone forgot: a buy-sell provision, a deadlock procedure, a mediation requirement. If nothing was signed, the state's partnership or LLC act is the agreement.

Get the numbers in front of both partners. A partner has a statutory right to the books and to information about the business (16403). Many arguments about who is carrying whom end once both people are looking at the same figures.

Use a neutral before a court. Mediation is cheap next to litigation, and a mediator can propose a buyout price neither partner would accept from the other.

Buy out or be bought out. If one partner wants to keep the business, a negotiated buyout is usually the best outcome for both. How to buy out a business partner covers valuation and payment terms.

Ask a court as a last resort. A court can order a partnership wound up when another partner's conduct makes it "not reasonably practicable to carry on the business in partnership with that partner," or can expel a partner for a willful or persistent material breach (16801(5); 16601(5)). It is slow and expensive, and the business rarely survives it intact.

For when to stop trying, see when to walk away from a partnership; for how this played out at well-known companies, from Snapchat's $157.5 million settlement with an early participant (Forbes) to others, see famous failed business partnerships and what broke them.

The common thread is that every one of these failures is cheaper to prevent than to repair. The agreement that names contributions, pay, capital, decision rights, loyalty and the exit takes a few weeks and a lawyer's fee to write, and is worth most on the day nobody wants to read it.