MARS · Product Sizing

Who Would Actually Pay Us

Sizing the MCP-server client market from what our substrate can see today.

Every fund vehicle registered in the last 12 months is a candidate. Every ADV-registered private-fund advisor is a candidate. Every family office we can identify — a candidate. The quant end of the hedge-fund world is not: they run their own pipelines on raw XBRL and Bloomberg feeds and don't want narrative-shaped data. The interesting question is what remains once you strip them out and stop double-counting fund vehicles versus their management companies. We can count all of it from the substrate.

21,458
Active fund vehicles
Filed a Form D in last 12mo
7,261
ADV fund advisors
Registered fund managers
17,312
Mid-market RIAs
$100M – $10B RAUM
~680
Family offices
Identifiable; real n is 3–5×

01Firms actively raising and buying, by fund type

Grouping SEC Form D filings by fund type + activity. Each row is filings in the last 12 months, tallied as distinct fund vehicles (~1 per Form D) and distinct CIKs (proxy for management company, though many management companies operate several CIKs — real firm count is typically 60–75% of the CIK count).

Fund type Distinct fund vehicles (12mo) Distinct CIKs (12mo) Est. mgmt firms MCP fit
Venture Capital Fund 9,286 9,285 ~6,500 strong — need portfolio & co-investor lookup
Private Equity Fund 5,858 5,858 ~4,000 strong — target sourcing + M&A comp
Other Investment Fund 4,538 4,536 ~3,000 strong — crossover / late-stage crossovers
Hedge Fund 1,776 1,776 ~1,200 mixed — split into fundamental (yes) vs quant (no)

02ADV-registered fund advisors (ground truth for management firms)

Form D counts every fund vehicle. ADV counts the management company. This is the more honest firm-count number: 7,261 total ADV-registered private-fund advisors with a stated top fund type. Also worth noting the aggregate RAUM per type — this is the pool of assets these firms are managing today.

Top fund type RIA firms Aggregate RAUM ($T) Avg firm RAUM MCP demand
Hedge Fund 2,896 $26.2T $9.0B Fundamental-shop half: high. Quant half: low.
Private Equity Fund 2,290 $16.6T $7.3B Target sourcing, comps, exit tracking — high
Other Private Fund 957 $22.2T $23.2B Crossover funds, mostly high
Real Estate Fund 556 $4.1T $7.4B reit vertical serves them; high
Venture Capital Fund 348 $0.95T $2.7B Brand-name VCs; smaller VCs go ERA
Securitized Asset Fund 201 $6.3T $31.3B Mostly structured product shops — low
Why VC count from ADV (348) is so much lower than Form D (9,286)
Sub-$150M-AUM VCs are exempt from full ADV registration (Exempt Reporting Adviser regime — "ERA"). Only marquee VCs and those managing outside pooled vehicles need full SEC registration. The 348 ADV number is the "brand-name" VC universe. The 9,286 Form D number captures the full active-fund-vehicle universe including the ERA-only tail. Both are real; they're just answering different questions.

03The mid-market RIA sweet spot

For the MCP thesis specifically, size matters. Firms managing under $100M can't afford enterprise data. Firms managing over $10B have their own data-engineering teams and mostly build in-house. The $100M – $10B RAUM band is 17,312 RIAs and it's the exact segment PitchBook and Preqin under-serve (both expensive; both priced for enterprise buyers).

RAUM band RIA firms MCP fit
< $100M 3,772 no — priced out; use free tools
$100M – $1B 12,601 sweet spot — can pay, need the data, no eng team
$1B – $10B 4,711 sweet spot — same, at higher price point
> $10B 1,407 maybe — have their own data teams already

04Family offices — 680 identifiable, real market 3–5×

Family offices are the market segment PitchBook and Preqin miss most because family offices actively avoid registration (Family Office Rule exemption). We can identify 444 via ADV name-match ("family office" / "family capital" / "multi family" in the RIA name) and another 235 via Form D name-match. Of the ADV-visible 444, 70 report RAUM > $100M and 24 report RAUM > $1B — those 24 are the mega-single-family-offices (SFOs of Bloomberg / Zuckerberg / Ellison / etc. scale).

Real family-office universe is 3–5× what we can see because most single-family offices choose NOT to register at all (Family Office Rule exempts them if they only serve one family). Industry estimates put the US SFO count at 3,000–5,000. Add multi-family offices (roughly 200 major, several hundred smaller) and the true addressable pool is likely 3,500–6,000 firms.

Family offices are unusually good MCP prospects because: (a) they're generalist investors who need private-market + M&A + investor data + person data + ADV all in one place, (b) they don't have data-eng teams, (c) they can pay real prices ($20-50K+/yr common for boutique data services), (d) they don't buy PitchBook because it's overpriced for their use case.

05Explicitly OUT of scope: quant firms

Kee's point holds: quant / systematic shops don't want narrative-shaped data. They want normalized time series, raw XBRL, and pricing feeds — and they build the pipeline themselves. Named firms to exclude from any prospect list:

Firm RAUM ($B) Type
PGIM Quantitative Solutions111.0Other Private Fund
Quantinno Capital Management60.9Hedge Fund
Blackstone Credit Systematic Strategies35.7PE Fund
Systematica Investments23.8Hedge Fund
Trexquant Investment13.5Hedge Fund
Quantbot Technologies6.7Hedge Fund

Also excluded even though not named "quant" / "systematic": Renaissance, DE Shaw, Two Sigma, Citadel, Millennium, Point72, Jane Street, Susquehanna, Hudson River Trading, Jump Trading, Virtu, Tower Research, Optiver, IMC, Flow Traders. Combined these ~15 shops manage another ~$1T that shows up in ADV under `top_fund_type='Hedge Fund'` but is definitionally NOT the MCP target.

06Sample marquee prospects, Q2 2026 filers

Real firms who filed new fund vehicles in Q2 2026. Each row is a live-active prospect for an MCP-server pitch:

Fund vehicle Type Offering ($B) Filed
Spark Capital Growth Fund VIVC1.95Apr 2
Kleiner Perkins Select IV + XXIIVC2.92Apr 7
Altimeter Premier Growth Expansion FundVC1.50Jun 23
Addition SixVC1.50Jun 16
Benchmark Partners T1VC1.25Jun 30
Diffusion ONEVC (AI)1.00Jun 22
Third Rock Ventures VIIVC0.75Jun 3
Volition Capital Fund VIVC0.88Jun 4

Each of these is a new fund with fresh capital, hunting for deals, needing to understand who's writing what checks and who's exiting where. The MCP-tool pitch to any of them: "your associates can ask MARS a question in plain English — 'show me all AI infrastructure Series B raises where Sequoia was NOT the lead' — and get a structured answer back, without paying $80K/yr for PitchBook enterprise." That's the shape.

07Serviceable addressable market — the honest number

Adding it up with realistic filters (no double-counting fund vehicles vs mgmt firms; excluding quant; adding family-office adjustments; excluding sub-$100M RIAs that can't pay and mega-$10B+ firms that build in-house):

Segment Firms Est. willingness-to-pay ($K/yr) Annualized ceiling ($M)
Fundamental hedge funds ($100M – $10B AUM) ~1,500 10–30 30
PE firms (all sizes, mid-market focus) ~1,800 15–40 50
VC firms (brand + ERA tail combined) ~4,500 5–20 45
Family offices (SFO + MFO, addressable) ~2,500 15–50 75
Other private-fund managers (crossovers etc.) ~800 15–40 20
Total SAM ~11,100 $220M/yr
What "$220M/yr ceiling" actually means
That's the theoretical top if every prospect in these segments bought at the midpoint of their willingness-to-pay range. Real capture is a fraction — even great niche data products land 3–8% market share of their SAM at maturity. Realistic 5-year revenue if MCP is a real product: $5M–$25M ARR range depending on how many enterprise-anchor accounts (funds > $10B AUM, family offices > $500M) sign. Not a $1B business, but a real one.

08What we ship first, and to whom

The MCP catalog is already there (funding-vertical shipped July, mergers vertical live on argos-mcp-server, sedar live for Canadian resource, donor family for wealth research). The 2026-Q2 mega-fund filers above are the highest-signal warm-list because they just closed capital and are about to deploy — they need target sourcing yesterday.

Family offices are the segment where MCP has no incumbent competition worth naming. PitchBook targets institutional buyers; Preqin targets LP-facing GPs; the family office CIO with a small team and generalist mandate is under-served everywhere.

The pipeline sequence to prove the thesis: (1) ship 5-10 mid-market VC/PE trial accounts through 2026 H2 to validate willingness-to-pay in the $10-20K range, (2) close 2-3 marquee family-office accounts at $30-50K each to establish that pricing tier, (3) publish 2-3 case studies. Then scale.