FOUNDATION SERIES FOR FAMILY OFFICES · No. 5 of 5
How to Choose a Venture Manager?
This is our fifth and last installment aimed at educating family offices regarding venture capital as an asset class. You can find the first four articles in order here, here, here and here. This last post focuses on finding the right manager.
The decision the whole series has been leading to — and the questions that separate a real manager from a good pitch.
Pascal Bouvier, MiddleGame Ventures · June 2026
Everything in this series has been leading here. We have walked through how a venture fund works, why the asset class belongs in a family-office portfolio, how much to commit and at what pace, and where we believe the asymmetry sits — European fintech, early. But every one of those decisions is ultimately hostage to a single one: which manager you back. Get the mechanics, the sizing and the geography right, then choose the wrong partner, and you will still lose money — politely, with quarterly reports. This final piece is about not doing that.
Why selection is the investment
In most asset classes, manager choice is a refinement. The gap between a good and a mediocre public-equity fund is a handful of points, which is why so much of that world has gone passive. Venture is the opposite. The spread between top- and bottom-quartile venture funds exceeds thirty percentage points in most vintage years — the widest manager dispersion in private markets, and a different universe from public equities.[1] That is the power law from the first piece reappearing at the level of funds, not just companies: a single fund-returner decides whether a vehicle is top-decile or forgettable. The average venture fund is a poor investment. The good ones are extraordinary. There is not much in between.
Now add a second fact that changes everything: in venture, performance persists. Kaplan and Schoar found the correlation between a firm’s successive funds approaching 0.7 in venture — far higher than in buyout — because the edge is proprietary expertise, not luck, and later work finds that venture persistence has held up even as buyout’s has faded.[2] Persistence cuts both ways, but read it precisely: it is a property of a team’s repeatable judgment, not of its brand or its fund number. The quality worth identifying is the judgment — ideally early, before the brand catches up to the talent — and a weak manager is unlikely to surprise you to the upside. Put dispersion and persistence together and the conclusion is unavoidable: in venture, manager selection is not part of the investment. It is the investment. The rest of this piece is how to do it.
Read the track record honestly
Start with the numbers, using the metrics from the first piece as instruments of diligence rather than decoration. Ask for net DPI by vintage — cash actually returned, after all fees, year by year — not a single blended headline. Insist on the distinction between gross and net: what the deals did is one thing, what you would have kept is another. Treat TVPI with suspicion, because it leans on the manager’s own marks and one aggressive markup can flatter an entire fund; treat IRR with more, because an early exit or a subscription credit line can inflate it handsomely. Ask how concentrated the track record is — how few deals produced the returns — because that tells you whether you are looking at a repeatable process or one fortunate bet. And ask about the losers. A manager who will only walk you through winners is managing your perception, not revealing their judgment. The single most useful sentence you can put to any GP is this: show me net DPI by vintage, and tell me which deals drove it. How they answer tells you most of what you need to know.
Favor the specialist over the generalist
Past the numbers sits the real question: why does this firm win, and will it keep winning? Increasingly the answer turns on one distinction — specialist or generalist. A generalist sells pattern-matching across sectors: “we back great founders in big markets.” That is exactly the layer artificial intelligence now commoditizes, and exactly what every other firm also claims. A specialist sells something a model cannot retrieve and a tourist cannot fake: ground truth in one domain — which buyers actually buy, which regulators actually enforce, which workflows actually resist replacement. The data has caught up with the intuition. Across more than 1,300 venture funds, specialists returned an average IRR of 15% against 11% for generalists, and the specialist edge is sharpest in precisely the sub-$250 million funds where early-stage venture is decided.[3] In an AI world, where analysis is free and plausible companies are infinite, the generalist’s product is deflating and the specialist’s is appreciating. Concentrate your search there.
And within the specialists, look for obsession. The managers worth backing are not dabblers who picked a sector because it was fashionable; they are exceptional talent that obsesses over its area of expertise — investors who have spent years inside one industry, cannot stop thinking about it, and therefore see the signal in it before the crowd does. A real edge survives the question, “what do you see that others don’t, and why does that stay true?” Pedigree does not answer that question; obsession does. A marquee logo on a partner’s résumé is a credential, not a moat — ask a manager to name their territory and prove they own it rather than rent it.
Favor emerging over established
The instinct, having accepted that good judgment persists, is to reach for the famous names. Resist it. The evidence on where that judgment actually pays is strikingly consistent: emerging managers outperform. They reach top-quartile performance about 34% of the time against the 25% you would expect by chance, and an analysis of some 2,500 funds raised between 2000 and 2024 found emerging managers ahead of established ones on DPI, IRR and TVPI alike, by roughly 250 basis points on average.[4] This is not a paradox of persistence; it is the point of it. Persistence belongs to the team’s judgment, and that judgment is most potent before it has been dulled by scale.
There is a structural reason, and an opportunity folded inside it. A firm on its first to fourth fund is hungry, its edge undiluted, its partners’ own money and name fully at stake, and its fund small enough that a single winner still moves the result. The established mega-funds drift the other way — toward gathering assets and fees, and toward a fund so large it can no longer be choosy, which quietly erodes the very edge that built the brand. The opportunity is that most institutional LPs will not commit before Fund IV, citing the lack of a long track record. That caution leaves the best emerging managers under-capitalized, better-priced and more aligned — exactly the inefficiency a family office, answering to no investment committee, is built to exploit. Back the talent before the brand catches up to it.
Probe the team, not the star
Many venture track records belong to one person. Find out whether you are buying a franchise or a founder. Who actually makes the investment decision; how the partnership reaches it; what happens to the strategy if the rainmaker leaves — the key-person clause exists precisely because this risk is real. For a relationship you intend to renew across funds, and venture rewards exactly that kind of loyalty, continuity is everything. Watch, too, how the firm treats founders, because that reputation is what sources the next generation of winners. A manager founders quietly avoid will lose access long before it shows up in the returns — and by the time you see it in the numbers, you are two funds too late.
Test the discipline
Edge without discipline burns out. Ask how the firm constructs a portfolio: its ownership targets, and the reserves it holds to follow its winners — a manager with no follow-on reserve is leaving its best outcomes on the table, and has misunderstood its own business. Ask whether it stays in its lane or chased the last fashionable thing; style drift is a warning light. And ask the 2021-versus-2026 question directly: what did you change when the market turned? A firm whose reflexes never updated through that cycle is telling you something important about how it will behave in the next one. A disciplined strategy you can state in a single sentence beats a flexible one that needs a paragraph to explain itself.
Scrutinize alignment, fees and terms
Alignment reveals itself in the documents, not the pitch. Confirm that the GP has meaningful skin in the game — its own capital in the fund, ideally well beyond the one-to-two-percent minimum — so that it loses alongside you. Check that fees and carry sit at market, and that any transaction or board fees are offset against the management fee rather than pocketed on the side. Confirm a whole-fund waterfall with a clawback, so that carry is earned only once you are made whole. And look at how the firm treats its limited partners: clear reporting, fair side letters, real co-investment rights for those who bring value. A manager that is generous with information and disciplined on terms is showing you an alignment no glossy deck can fake.
The questions to ask
If you take nothing else from this piece, take these seven questions into your next meeting with a prospective manager:
- Show me net DPI by vintage — and tell me which deals drove it. (For mature managers; see the caveat below.)
- What do you see that others don’t, and why does that stay true?
- How much of the fund is reserved for follow-on, and how do you decide who gets it?
- How much of your own money is in the fund?
- Who makes the investment decision, and what happens if you leave?
- Walk me through your three worst investments and what you changed afterward.
- What will you not invest in?
One caveat on the first question. Net DPI by vintage is the right test for a mature manager with realized funds behind it. For an emerging manager — the kind this piece argues you should favor — there is little realized DPI to show yet, by definition, and demanding it would screen out precisely the talent you want. There, the weight shifts to questions two through seven: the edge, the reserves, the alignment, the honesty about losses. Track record describes the past; for a newer GP, those other six questions are what tell you about the future.
You are not testing whether they can sell — they can. You are testing whether they think clearly, own a real edge, hold their discipline, and will treat your capital the way they treat their own. The good ones welcome the interrogation. The ones who bristle are answering the question for you.
Where this leaves you
This is where the series has been pointing all along. A family office that understands the mechanics, commits to the asset class, sizes and paces with discipline, and aims its capital at the European fintech asymmetry has done the serious intellectual work. The return still comes down to the partner it chooses — so choose for edge, discipline, alignment and honesty, not for brand, comfort, or the warmth of the room. The right manager hands you the winners and the insight while sparing you the heartache of the write-offs; the wrong one hands you the fees and a story. That difference is worth more than any other decision you will make in this asset class.
When you are ready to put a manager through these seven questions — including us — we would welcome it. [Placeholder: link to the MiddleGame Ventures one-pager and materials once finalized.]
This concludes the Foundation Series: how a venture fund works · funds, direct, or both · how much and at what pace · why European fintech · how to choose a manager.
[1]The spread between top- and bottom-quartile venture funds exceeds 30 percentage points in most vintage years — the widest manager dispersion of any private-markets asset class, and far wider than public equities; buyout dispersion is materially narrower. It is why LPs spend disproportionate time on VC manager selection. See Cambridge Associates, “Venture Capital Positively Disrupts Intergenerational Investing.”
[2]Kaplan & Schoar (2005) found returns persist across a firm’s successive funds, with venture correlations approaching 0.7 — far higher than buyout — attributed to proprietary GP expertise; later work finds VC persistence has held up even as buyout persistence weakened. See Kaplan & Schoar, “Private Equity Performance” (MIT), and Harvard Law on VC persistence.
[3]A PitchBook study of 1,306 VC funds (vintages 2000–2020) found specialist funds averaged a 15% IRR against 11% for generalists; the specialist advantage is strongest in funds under $250m and narrows for mega-funds or where a generalist fields a team of specialists. See PitchBook, “Do VC specialists outperform generalists?”
[4]Emerging managers reach top-quartile performance about 34% of the time, against the 25% expected by chance; an analysis of roughly 2,500 funds (2000–2024) found emerging managers ahead of established ones on DPI, IRR and TVPI, by around 250 basis points on average — yet most institutional LPs wait until Fund IV. See Venture Capital Journal and Velocity Ventures.