Multi-Family or Single-Family?

Families almost always open this conversation as a cost comparison, and the cost numbers are knowable enough that the discussion feels settled quickly. It is the wrong frame. The durable difference between multi-family and single-family advisory is not price. It is who carries the judgment when a decision is genuinely close, and how that judgment gets tested before capital moves.

The structural difference: who the advisor actually works for

A single-family office serves one family. Every hire, system, reporting cycle and investment policy is built around that family's assets, tax position, time horizon and tolerance for illiquidity. Nothing is a compromise between competing principals, because there are none.

A multi-family office serves several families on shared infrastructure. That sharing is the entire proposition: a family gains access to a professional team, institutional-grade reporting, manager relationships and deal flow that would be uneconomic to build alone. What is shared alongside the cost, though, is capacity, attention and, in the cases that matter most, allocation.

That distinction is abstract until a specific moment arrives. A co-investment comes in with limited capacity. Two families on the same platform want it. The allocation policy, which read as boilerplate at onboarding, is suddenly the most important document in the relationship. Single-family structures never face that test. Multi-family structures face it regularly, and the good ones have a written, defensible answer ready before it happens.

What each model actually costs

The economics are well documented and worth stating plainly, because they set the boundaries of the decision.

A single-family office is generally discussed as viable from around $100 million in investable assets. In practice, many practitioners now point to roughly $250 million as the level where the economics genuinely work, on the reasoning that fixed operating costs should stay under 1% of assets. Below that, the family is paying institutional overhead on a base too small to absorb it. Average running costs sit near $3 million a year, rising to roughly $6.6 million for offices above $1 billion in assets.

Multi-family offices typically engage from $10 million to $25 million, and are most cost-efficient in the $25 million to $100 million range. Fees usually run 0.50% to 1.00% of assets, sometimes higher at smaller sizes, often with a retainer for services outside investment management. A frequently cited crossover point sits somewhere in the high $40 millions, which is a useful anchor provided it is treated as an indication rather than a rule.

Two adjustments matter more than the headline numbers. First, complexity is not proportional to assets. A family with $80 million spread across an operating business, three jurisdictions and a philanthropic vehicle consumes more governance than a family with $300 million in marketable securities. Second, the cost of the structure is not the cost of the decisions it produces. A single avoidable direct investment loss will exceed a decade of fee differential.

Decision vetting: where the two models diverge most

This is the part of the comparison that gets least attention and determines the most. Decision vetting is the process by which a recommendation is tested before capital commits, and the two structures fail in opposite directions.

The single-family failure mode is insufficient challenge. A small team, loyal to one principal, working on opportunities the principal often sourced personally, is structurally poor at saying no. There is no second client whose different answer creates a natural comparison, and frequently no investment committee with genuine independence. Conviction and familiarity get mistaken for diligence.

The multi-family failure mode is generic challenge. A platform running a standard process across many families produces consistency, which is valuable, and standardisation, which is not always. The recommendation that emerges may be entirely defensible in general terms and wrong for this family's tax position, liquidity needs or concentration in the same sector through an operating business.

Whichever structure a family uses, the same three questions do most of the work on any recommendation:

1. What would have to be true for this to work? Force the thesis into explicit assumptions. If they cannot be listed, the analysis is not finished.
2. Who verified this independently of the person recommending it? Not reviewed. Verified, against source documents, by someone whose compensation does not depend on the deal proceeding.
3. What does the downside case cost us, and can we absorb it? Not the probability of loss, which is usually unknowable, but the size of it against total family capital.

Process discipline is testable in the same way. Direct deals in the lower middle market normally require four to eight weeks of diligence from signed letter of intent to close. Timelines compressed below three weeks raise risk materially, and a compressed timeline is almost always a symptom of something else: competitive pressure, a seller-driven deadline, or a sponsor who does not want the file examined closely.

Return profile analysis and the allocation question

Return profile analysis means understanding not just what an investment is expected to return, but the shape of that return: when cash actually arrives, what has to be committed and held, how the position behaves in a drawdown, and how it correlates with everything else the family owns, including the operating business that produced the wealth.

The two structures approach this differently, and both approaches have a bias worth naming.

A single-family office can build the analysis around the family's actual balance sheet, including assets an outside advisor may not see or price. That is a real advantage. The bias is concentration: families tend to invest in what they know, and the sector that generated the wealth is often already the largest exposure. A portfolio can look diversified by line item while being a single correlated bet on one industry.

A multi-family office brings comparative data across families, manager access, and a discipline of benchmarking that is genuinely hard to replicate alone. The bias is structural and worth stating directly: a fee earned on assets under management creates an incentive to grow and retain assets, which tilts gently toward illiquid private strategies where redemption is not straightforward. That is not misconduct. It is an incentive, and incentives should be visible rather than assumed away.

The practical test on any recommendation is the same in both models: model the cash, not the return. A 2.2x gross multiple that returns nothing for six years is a different instrument from a 1.7x that distributes from year two, and for a family with a live liquidity requirement it may be the wrong one regardless of which shows better on a returns page.

Conflicts, and how to test for them

Every structure has conflicts. The question is whether they are disclosed, priced and managed, or simply unmentioned. The SEC has signalled that its 2026 examination cycle emphasises fiduciary adherence, conflict-of-interest management and documentation of investment decisions, which makes this a good moment to ask the questions directly.

• How are you paid, in total? Every component: assets under management, retainers, transaction fees, referral arrangements, revenue from affiliated products or vehicles. A reluctance to itemise is itself the answer.
• What is your allocation policy when capacity is limited? Ask for it in writing. Ask when it was last applied and what happened.
• Are you a fiduciary at all times, or only for some services? Partial fiduciary status is common and legitimate, but the boundary should be explicit.
• Who verifies the analysis you present to us? An internal committee that reports to the same person who owns the client relationship is not independent verification.
• What have you recommended against in the last twelve months? An advisor who cannot name declined opportunities is either not seeing enough or not filtering.

The single-family equivalent of this exercise is uncomfortable but more important, because there is no counterparty to interrogate. It means asking whether the team has the standing to disagree with the principal, and whether anyone has done so recently.

Choosing, and the hybrid most families actually run

Stated as a binary, the choice is straightforward at the extremes. Below roughly $50 million, a multi-family office or a well-constructed advisory relationship is almost always right. Above roughly $250 million, with material complexity and an intention to invest directly, a single-family office usually earns its cost.

Between those points, which is where most of the market sits, the honest answer is that the binary is false. In practice, the majority of families operate a hybrid: a small internal team of one to three people handling governance, consolidated reporting, family coordination and relationship ownership, supported by external specialists engaged for the work that arrives episodically. Transaction diligence, valuation, capital structure design and independent decision vetting are all consumed in bursts and are expensive to keep on the payroll waiting.

That model tends to work because it separates two things the pure structures conflate. Continuity, judgment about the family and ownership of the relationship sit inside. Specialist capability and independent challenge come from outside, where they are easier to hold to a standard and easier to replace when they underperform.

Whichever structure a family lands on, the question to revisit annually is not what it costs. It is whether the last five significant decisions were tested by someone with the expertise to find the flaw and the independence to say so.

Questions we get asked

What is the main difference between a single-family and a multi-family office?

Who the advisor works for. A single-family office serves one family exclusively, so priorities and reporting are built around that family alone. A multi-family office serves several families on shared infrastructure, which lowers cost and broadens access, but means capacity, attention and deal allocation are shared.

At what level of assets does a single-family office make sense?

Around $100 million is the commonly cited entry point, though many practitioners now point to roughly $250 million as the level where the economics genuinely work, since that is where fixed costs stay under 1% of assets.

How should a family vet an investment decision made by its advisor?

Ask what would have to be true for it to work, who verified that independently of the person recommending it, and what the downside case costs. Then check the process: lower middle market direct deals normally need four to eight weeks of diligence, and anything under three weeks warrants an explanation.