You don't always have to buy a company to get what it has. Sometimes you can borrow it, strike a deal where two firms pool resources, share the upside, and stay separate. The trouble is that these deals are easy to sign and hard to run: by the practitioners' own reckoning, roughly half underperform. This guide is about being in the other half.

The quick version

  • Three forms, one spectrum. A strategic partnership/alliance is a contract to cooperate; a joint venture goes further and creates a jointly-owned company. Pick the lightest structure that protects what's at stake.
  • The odds are sobering. Senior practitioners estimate 40–60% of completed joint ventures underperform their potential; McKinsey's wider JV work puts the share that hit all their original objectives at under a quarter. Alliance failure estimates run 60–70% (Hughes & Weiss, HBR 2007).
  • Three tests every deal must pass. Is there value here that neither side could create alone? Is the deal governed well enough to capture it? Does each partner earn enough to keep showing up? (Gomes-Casseres' three laws.)
  • The failure is usually relational, not legal. Most deals die from how the partners work together day to day, not from a flaw in the contract. Spend your effort on the working relationship and your own internal stakeholders.

The idea in depth

Start with the vocabulary, because it's loose in everyday use and the distinction carries real consequences. A strategic alliance (or strategic partnership) is a cooperative agreement between firms that stay legally independent, a co-marketing deal, a supply arrangement, a shared R&D programme. No new company is born; the relationship lives in a contract. A joint venture is the heavier version: the partners create a separate, jointly-owned entity, fund it, staff it, and split its profits and control. Think of it as a spectrum of commitment, from a handshake-with-a-contract at one end to a co-owned subsidiary at the other.

Why bother with either instead of an outright acquisition? Each form lets you reach a capability, a market, or a technology without paying for, and absorbing, the whole organisation that holds it. You share the cost and the risk. The catch is that you also share control, which is exactly where the difficulty starts.

The evidence: easy to sign, hard to run

The numbers are unusually candid because they come from the people who run these deals. When McKinsey interviewed senior joint-venture practitioners across 30 S&P 500 companies, people with combined experience of more than 300 ventures, they estimated that 40 to 60 percent of their completed JVs had underperformed their potential; some had failed outright ("Avoiding blind spots in your next joint venture," McKinsey, 2014). McKinsey's wider work on joint ventures sharpens the picture: fewer than 25 percent achieve all of their initial objectives, and close to 70 percent hit meaningful friction within the first three years. For lighter alliances the picture is no rosier: Jonathan Hughes and Jeff Weiss, drawing on two decades at Vantage Partners, put the failure rate at 60 to 70 percent ("Simple Rules for Making Alliances Work," HBR, November 2007).

So the move is: treat the signing as the start line, not the finish. McKinsey's diagnosis is that experienced teams, in the rush to close, skip steps and lose discipline in the handoffs between the venture's five stages, business case, structure, deal terms, operating model, and ongoing operations. The cheap insurance is to name who owns each stage and the transition into the next one, so the deal doesn't fall through the gap between the people who negotiated it and the people who have to live with it.

An honest limitation: these are self-reported estimates from practitioners and consultants, not a clean randomised study, "underperform their potential" is a judgement call, and the firms surveyed are large incumbents. Read them as a strong directional warning, not a precise law. The signal that survives across independent sources is consistent: a large share of these deals disappoint, and they disappoint in the running, not the writing.

The three laws: a test before you sign

The most useful frame for deciding whether a deal makes sense comes from Benjamin Gomes-Casseres, who has studied business combinations for three decades. In Remix Strategy (Harvard Business Review Press, 2015) he reduces every alliance, JV or acquisition to three laws. First: the combination must have potential to create more value than either party could alone. Second: it must be designed and managed to realise that value, potential is not capture. Third: each party must earn enough from it to stay motivated to contribute.

Potential value is not captured value. The first law tells you whether to do the deal; the second and third tell you whether you'll actually get anything from it.

What to do with this: run a candidate partnership through all three laws on a single page before you spend money on lawyers. Most teams obsess over the first law (the strategic logic looks great) and barely test the third, will the partner still want to feed this thing in year three, once their priorities have shifted? A deal that passes law one but fails law three is the classic slow-motion failure. It launches, then quietly starves.

flowchart TD
    A(["Spot a capability, market
or technology you lack"]) --> B{"Do you need
to OWN it?"}
    B -->|"Yes, control is core"| C(["Acquire (M&A)"])
    B -->|"No, access is enough"| D{"How much is
at stake / at risk?"}
    D -->|"Lower stakes,
clear scope"| E(["Strategic alliance
(contract only)"])
    D -->|"Higher stakes,
shared operations"| F(["Joint venture
(co-owned entity)"])
    E --> G(["Govern the relationship,
not just the contract"])
    F --> G
Choosing the lightest structure that still protects what's at stake. Leaders Loop

Why the structure choice matters: governance and trust

The alliance-versus-JV decision isn't cosmetic, it's a choice about how you guard against the partner behaving badly. Transaction-cost economics (the tradition running from Oliver Williamson's work on markets and hierarchies) frames it plainly: the more you fear opportunism, a partner walking off with your know-how, or quietly underinvesting, the more you reach for stronger safeguards, and shared equity in a joint venture is one of the strongest. Co-ownership ties each side's money to the outcome, which blunts the incentive to cheat. Where the risk of opportunism is low, a lighter non-equity alliance is cheaper and faster, and the heavier structure is just overhead.

But governance on paper is not the same as cooperation in practice, and this is where the research gets interesting. Accumulated trust and reputation can do work that contracts can't, when partners trust each other, the perceived risk of betrayal falls and the marginal value of expensive equity safeguards drops with it. Hughes and Weiss make the operational version of the same point: their evidence says alliances fail less from a bad business plan than from how the partners work together. Their counsel is to focus less on nailing the plan and more on the working relationship; to put the partners' differences to work rather than sand them away; and, the one most teams neglect, to spend as much energy managing your own internal stakeholders as managing the partner.

The practical rule: size the structure to the risk, not the ambition. Ask, "What's the worst a partner could reasonably do to us here, and would shared ownership actually deter it?" If yes, the JV's overhead is buying you something real. If the honest answer is "not much," a tight alliance contract plus a serious relationship cadence will beat a cumbersome co-owned entity. A limitation worth naming: the governance-versus-trust research describes broad tendencies, not a formula, high-trust partners sometimes still need hard safeguards for high-stakes assets, and a strong contract is no substitute for the relationship work either way.

flowchart LR
    A(["Strategic alliance
contract only, lowest commitment"]) --> B(["Equity stake
buy a slice of the partner"])
    B --> C(["Joint venture
co-owned entity, shared P&L"])
    C --> D(["Acquisition
full ownership & control"])
    A -.->|"more control, more cost, harder to exit →"| D
The commitment spectrum: each step buys more control and more protection against opportunism, at the price of flexibility. Leaders Loop

A worked example

The figures below are illustrative, a composite scenario, not a real company's accounts, used to show the mechanics.

A mid-sized European industrial-software firm, "Northwind," wants into South-East Asia. It has a strong product but no local sales force, no regulatory relationships, and no brand. Building all of that alone would cost an estimated €18m and take three years. Acquiring a regional player it likes is priced around €60m, more than it wants to risk on an unproven market.

So Northwind runs the three laws. Law one (potential): pairing its product with a local distributor's market access plausibly unlocks far more revenue than either side gets alone, pass. Law two (capture): this is where it slows down. A loose co-marketing alliance is cheap, but Northwind's source code and customer data would be exposed to a partner that also sells rival products, and there's a real fear the partner could learn the product and then build its own. The risk of opportunism is high. Law three (motivation): the distributor will only invest in a local salesforce if it shares in the long-term upside, not just a referral fee.

Laws two and three both point away from a light alliance and toward a joint venture: a co-owned local entity, 55/45 in Northwind's favour, that licenses the product under tight IP terms and pays the distributor through equity rather than commissions. The shared ownership aligns the partner's incentive to invest (law three) and the equity stake makes walking off with the technology self-defeating (law two). Crucially, Northwind names a single internal sponsor accountable across all five stages, and budgets as much time for managing its own board's nerves as for the partner. Eighteen months in, the venture is behind its revenue plan but the relationship is healthy and the partners are renegotiating scope rather than lawyers. That is what "the other half" usually looks like, not a triumph, but a deal that's still worth running.

Frequently asked questions

What's the actual difference between a strategic alliance and a joint venture?

A strategic alliance is a contract to cooperate while both firms stay fully independent, no new company is created. A joint venture goes further: the partners form a separate, jointly-owned entity, fund it, and share its profits and control. The alliance is lighter and easier to exit; the JV gives more control and stronger protection against a partner behaving opportunistically, at the cost of more overhead and a harder unwind.

Why do so many of these deals fail?

Mostly in the running, not the signing. McKinsey's practitioners reported 40–60% of joint ventures underperforming, and Hughes and Weiss put alliance failure at 60–70%. The common thread across the research is relational: deals collapse from misaligned incentives, neglected day-to-day collaboration, and, frequently, internal stakeholders on one side who were never brought along, rather than from a defective contract.

When should I just acquire instead of partnering?

When control of the asset is core to your strategy and you can't tolerate a partner steering it, or when the value depends on full integration. Partnerships shine when access is enough: a new market, a complementary technology, a one-off R&D push, or a bet too uncertain to justify buying the whole company. If you'd be unwilling to share the upside, that's usually a sign you want ownership, not a partner.

How do I protect our intellectual property in a partnership?

Match the safeguard to the risk. Where the danger of a partner copying or misusing your know-how is high, stronger structures help, tight licensing terms, narrow scope, and shared equity that makes cheating self-defeating. Where the risk is low, heavy safeguards are just friction. Note the general principle here, and have a qualified legal adviser draft the specifics for your jurisdiction; IP and competition law vary widely.

What's the single most important thing to get right?

Governance of the relationship, not just the deal terms. Name one accountable owner who carries the venture across all five stages, agree a regular cadence for surfacing problems early, and spend real time managing your own internal stakeholders, Hughes and Weiss found that's the step teams most often skip and most regret.

Related in the Toolkit

Where to go next