Most founders open a fund’s portfolio page, scan for logos they recognize, and close the tab feeling either encouraged or dismissed. Neither reaction is informed by anything useful, because a portfolio page is not a brag sheet. It is a constraint map, and read correctly it tells you which partner has room for a new board seat, which of your competitors already rules the fund out, and whether this fund is early enough in its current cycle to write you a check at all.
Why Founders Read Portfolios Wrong (And What the Book Actually Tells You)
The habit starts early. A founder pulls up a fund’s site, scrolls the logo wall, spots two or three companies they admire, and mentally files the fund as “a fit.” That’s the logo-scan habit, and it wastes outreach because recognition is not evidence. A famous portfolio company from five years ago, under a different partner, in a different fund vintage, tells you almost nothing about whether this fund will write your check in 2026.
A portfolio, read properly, answers exactly three questions: fit (does this firm actually fund companies like yours, at your stage, in your category), conflict (is a competitor already in the book, and does that block you), and capacity (does the fund still have money and partner attention to deploy). Everything else, whether a partner will personally champion you, what they care about beyond the thesis, how they behave in a term sheet negotiation, is a different layer of research entirely.
That’s why portfolio analysis and partner research are a paired workflow, not substitutes for each other. This guide covers the fund-level questions. For the partner-level questions, the companion piece is how to research VC partners before a pitch, which walks through background, public writing, and deal history once you already know which partner to target. Skipping straight to partner research without doing the portfolio pass first is how founders end up pitching the wrong person at the right firm.
Signal 1: Stage Reality
A fund’s “stated stage” (what its website says) and its actual entry stage (what its last ten deals show) drift apart constantly, and the gap is one of the most reliable tells in the entire analysis.
Pull the fund’s last 8 to 12 announced deals and record the round each one entered at: pre-seed, seed, Series A, or later. Don’t guess from a headline. Read the actual announcement (a press release, the company’s own blog post, or a filing) to see whether the fund led, co-led, or merely participated.
Signal 2: Check-Size Drift
Seed funds that raise a larger new vehicle often keep calling themselves “seed investors” while their actual entries have moved to seed-extension or Series A rounds, because a bigger fund needs bigger ownership per check to make the math work. This is stage drift, and it is usually quiet, unannounced, and easy to miss if you only read the “About” page.
The pattern is visible even among well known names. Bessemer Venture Partners spent decades as a traditional early and growth stage venture firm, then launched BVP Forge in 2022 as a separately branded growth-equity vehicle built specifically to write larger, later checks, an explicit, public acknowledgment that its core fund’s check size no longer covered the later rounds it wanted into. Andreessen Horowitz sits at the other end: it has described itself as a seed-through-growth, multi-stage investor from the start, so a seed check followed much later by a growth-round follow-on isn’t drift, it’s the stated strategy working as designed. Union Square Ventures is a useful control case, having stayed concentrated on seed and Series A for most of its history, with little evidence of chasing the larger late-stage checks Bessemer added through a separate fund.
| Firm | States itself as | Observed pattern | What it signals |
|---|---|---|---|
| Bessemer Venture Partners | Early and growth stage venture | Added BVP Forge (2022) as a dedicated growth-equity vehicle for larger, later checks | Check-size drift formalized into a separate fund, worth confirming which vehicle would actually lead your round |
| Andreessen Horowitz | Seed through growth, multi-stage by design | Writes checks across the full range under one umbrella | Stage claim matches behavior, little drift to detect because the stated scope is already broad |
| Union Square Ventures | Early stage, thesis-driven | Concentrated on seed and Series A across most of its history | Consistent, narrow stage claim matches observed focus |
Record entry stage per deal from the firm’s own portfolio page and the company’s announcement, never from an estimated check size. Dollar figures are rarely disclosed accurately in press coverage, but the round label almost always is.
Signal 3: Thesis Density: Is Your Category a Bet or a Tourist Visit?
A single portfolio company in your category is an anecdote. Five is a thesis. Counting is the entire method here, and it is the fastest way to separate a fund that has actually committed to your space from one that made one opportunistic bet and moved on.
Start by pulling every company in your category from the firm’s own portfolio list, not from a third-party database that may be stale. Then classify what you find:
- Cluster: four or more companies in your category, spread across multiple fund vintages, often with a named partner attached to most of them.
- One-off: a single company, frequently a spray-and-pray check with no follow-on pattern and no partner clearly attached.
- Stale cluster: several companies in your category, but every one of them entered before 2023. That’s a real signal too: it can mean the fund made its category bets and moved on to a new thesis, or that the partner who drove those deals has left.
The pattern is easy to see in public portfolios. Ribbit Capital’s own portfolio page names Coinbase, Robinhood, Revolut, and Credit Karma inside a single consumer-fintech thesis, four companies spanning crypto trading, equities trading, neobanking, and consumer credit, which is a cluster by any reasonable count. Founders Fund’s portfolio, by contrast, is built around deep tech and defense bets like SpaceX, Palantir, and Anduril; fintech sits at the edges rather than at the center of that book, so a single fintech company there reads as an opportunistic add, not a thesis.
Sector pages are a fast shortcut to firms with real category density, because someone has already done the clustering work for you. Top VC Investors for AI and Machine Learning Startups in 2025 and Top VC Investors for Healthcare and Biotech Startups both group firms by demonstrated category commitment, which gives you a peer set to compare against rather than judging density in a vacuum. A firm with three AI companies looks thin next to a peer with twelve; the same three companies might look strong in a smaller, newer category.
| Category signal | Company count | Vintage spread | Read |
|---|---|---|---|
| Cluster | 4+ | Multiple years, including recent | Real thesis, prioritize |
| One-off | 1 | Single year | Opportunistic, deprioritize |
| Stale cluster | 3-6 | All pre-2023 | Investigate before pitching |
Signal 4: Conflict Mapping: The Analysis That Saves You Months
Nothing wastes founder time like a promising first call that ends the moment a partner says “we’re already invested in a company doing something similar.” That conversation is avoidable, and the fix is a competitor-overlap grid you build before you ever send an email.
Start with a practical dividing line: a direct conflict is a company solving the same problem for the same customer with a similar model; an adjacency is a company in a related space that doesn’t compete for the same budget or the same buyer. Funds treat these very differently, and so should you.
Build the grid by listing your own competitor set down one axis and the target funds across the other, then marking every cell where a fund’s portfolio page shows one of your competitors. Food delivery is a clean public illustration of how this plays out in practice: Sequoia Capital backed both DoorDash and Instacart, two direct competitors for the same delivery budget, while Founders Fund backed Postmates instead, and a fintech-focused firm like Ribbit Capital has no position in the category at all.
| Your competitor | Sequoia Capital | Founders Fund | Ribbit Capital |
|---|---|---|---|
| DoorDash | Direct conflict (invested) | No overlap | No overlap |
| Instacart | Direct conflict (invested) | No overlap | No overlap |
| Postmates | No overlap | Direct conflict (invested) | No overlap |
Sequoia’s willingness to back two direct competitors in the same category is itself a useful reminder: a conflict policy is a per-partner and per-fund judgment call, not a hard rule every firm applies the same way, so don’t assume a fund will auto-disqualify you just because a competitor is already in the book. If you’re working in a regulated category, the Healthcare and Biotech sector page is a good place to run this same exercise, because FDA pathway, payer relationship, and clinical indication overlap tend to produce the clearest direct-conflict cases.
When you do find a likely conflict, raise it in your first email instead of hoping it goes unnoticed. Something as direct as “I saw you’re invested in [company], want to flag we compete on X but differentiate on Y in case that’s disqualifying” signals maturity and self-awareness, and it often surfaces information you didn’t have, including the fact that the partner who led that deal already left the fund. Not every portfolio company is a blocker either: sometimes it’s a warm intro path, particularly if the overlap is adjacent rather than direct and the founder of that company would happily make an introduction.
Signal 5: Partner Slots
Funds don’t invest. Partners do, and each partner has a hard ceiling on how many new companies they can take on, regardless of how much capital the fund has left.
Attribute every portfolio company you can to an individual partner. Firm websites sometimes list this directly; when they don’t, check the company’s own announcement post (which frequently names the specific partner who led), the partner’s LinkedIn profile, or the partner’s public deal history. Over time you’ll build a rough map of who is actively deploying and who has gone quiet.
Signal 6: Board Load
This is the constraint that actually matters, more than fund size. A partner sitting on eight to ten active boards has little bandwidth left, no matter how large their fund’s most recent close was. A partner with two or three board seats, or one whose recent deal pace has visibly slowed, either has room, or has quietly stepped back from active investing. Both are worth knowing before you pitch.
The partner whose deal flow stopped eighteen months ago is a particularly useful find: it can mean they moved to a portfolio-support role, they’re raising a new fund, or they’ve left the firm entirely and nobody updated the website. Any of those changes who you should actually be talking to.
Once you’ve attributed a deal to a partner, you have a pitch angle, but attribution is where this section stops. Turning that attribution into an actual pitch (understanding what that partner has said publicly, what they look for beyond the thesis, how they’ve behaved with founders before) is exactly what the 9-step framework for researching VC partners is built to do. Portfolio analysis tells you who to research. It doesn’t replace the research.
Signal 7: Fund Cycle and Reserve Pressure
The single most overlooked signal is timing: a fund’s position in its own deployment cycle determines whether it can write you a check at all, independent of how good your fit is on every other signal.
Start with deployment pace. A burst of several new deals announced in a tight window usually means a fund is actively deploying and hungry for dealflow. A long gap since the last announced deal can mean the fund is between vehicles, heavily focused on existing portfolio support, or deliberately slowing down ahead of a raise for its next fund.
Fund vintage and announced fund size matter for the same reason. A firm that just closed a new fund is typically eager to deploy early and fast; a firm three or four years into a fund’s life is more likely conserving capital for follow-on rounds in its existing portfolio, a group of companies sometimes referred to as “reserve-heavy.” If a fund is reserve-heavy, it’s fair to ask directly, early in the conversation, how much of the current fund is earmarked for follow-ons versus new investments.
The pattern shows up clearly in public fund announcements. Andreessen Horowitz announced a series of multi-billion-dollar funds in rapid succession through 2021 and 2022, a public signal of a firm positioned to deploy aggressively across many stages at once during that window. A firm that hasn’t announced a new fund in several years, or whose last few announced deals are concentrated in follow-ons for existing portfolio companies rather than new logos, is telegraphing the opposite: capital preservation, not new-deal appetite. Check the firm’s own fund announcements (press releases, its own site, or coverage from outlets that track funding news) rather than estimating. If you want more names to run this against, the fintech sector page groups several firms whose deployment pace differs sharply without picking a side for you.
Turn Your Analysis Into a Ranked Target List (Do This Next)
Seven signals per firm, scored across dozens of candidates, becomes unmanageable fast without a system. The fix is a simple scorecard: one row per firm, one column per signal, weighted by what matters most for your raise.
| Signal | What you check | Suggested weight |
|---|---|---|
| 1. Stage Reality | Last entries match your round | High |
| 2. Check-Size Drift | Stated stage vs. actual behavior | Medium |
| 3. Thesis Density | Cluster count in your category | High |
| 4. Conflict Mapping | Direct conflicts with your competitors | Disqualifying if direct |
| 5. Partner Slots | Attributed deal volume per partner | Medium |
| 6. Board Load | Active board seats per partner | High |
| 7. Fund Cycle | Recent deployment pace and vintage | Medium |
Cap your initial pass at around 30 firms. Beyond that, the scorecard stops being useful and starts being busywork. Seed the list from whichever sector page matches your category: top VC investors for AI and machine learning startups, top VC investors for fintech startups, or top VC investors for healthcare and biotech startups, score each firm against the table above, then split the results into tiers: a top tier you pitch first, a middle tier you hold in reserve, and a bottom tier you deprioritize or drop.
This is exactly where portfolio analysis ends and list-building begins. The scores you just produced are the direct input to how to build a VC investor target list, which takes a scored firm list and turns it into a sequenced outreach plan. Do this next: pick your sector page, score ten firms this week, and you’ll have a working tier list before your next investor call.
Common Portfolio-Analysis Mistakes That Kill Otherwise Good Raises
Four mistakes account for most of the wasted outreach we see, and each one maps to a step in one of the two companion guides that would have prevented it.
Treating a marquee logo as proof of stage fit. A famous exit from five years ago says nothing about whether the fund’s current team, current fund size, or current stated stage still matches your round. Signal 1 (Stage Reality) exists specifically to catch this.
Analyzing the firm and ignoring the partner. A fund can look perfect on paper while the one partner who covers your category has no bandwidth left, or has quietly left the firm. This is why how to research VC partners before a pitch exists as a mandatory second step, not an optional one.
Over-indexing on exits from a fund the current team didn’t invest in. Partner turnover at venture firms is common, and a marquee outcome is frequently credited to a partner who left years before you’re reading the page. Attribute the deal to the actual person still at the firm before you weight it.
Letting the scorecard replace a real conversation. The seven signals filter a long list down to a workable one. They do not replace a first call, a reference check, or the judgment a founder builds by actually talking to a partner. Use how to build a VC investor target list to sequence outreach, but treat the score as a starting point, not a verdict.
The 90-Minute Portfolio Analysis Workflow
Here is the whole process chained into a single sitting, start to finish.
Minutes 0-20: pull the firm list. Start from whichever sector page matches your category, top VC investors for AI and machine learning startups, top VC investors for fintech startups, or top VC investors for healthcare and biotech startups, and build a working list of 20 to 30 firms.
Minutes 20-60: score the seven signals per firm. Move through Stage Reality, Check-Size Drift, Thesis Density, Conflict Mapping, Partner Slots, Board Load, and Fund Cycle for each firm, using the scorecard from the section above. Tools like Crunchbase, PitchBook, and CB Insights speed up pulling deal history, though the firm’s own portfolio page remains the source of record for what they actually claim to have invested in.
Minutes 60-90: conflict pass, partner attribution, tiering. Run your competitor list against every firm’s book, attribute standout deals to individual partners using LinkedIn and public announcements, and sort the scored list into pitch-first, hold, and drop tiers.
The reusable template. Save the scorecard as a template and refresh it every two to three weeks during an active raise, since fund announcements, new deals, and partner moves shift the picture faster than most founders expect. Sources like NVCA for broader fund-formation trends, and the SEC’s EDGAR full-text search for fund-related filings, are worth bookmarking alongside your portfolio-tracking tools for exactly this reason.
Frequently Asked Questions
What is VC fund portfolio analysis, and how is it different from just browsing a firm’s website? Browsing a website surfaces logos. Portfolio analysis systematically extracts entry stage, category density, conflicts, partner attribution, and deployment pace from that same page, then scores each firm against your specific raise instead of forming a general impression.
How many portfolio companies should I review before deciding a fund is a fit? Review enough to see a pattern, typically the firm’s last 8 to 12 announced deals for stage and pace, and their full category list for thesis density. A single company in either direction isn’t enough signal to decide anything.
How do I tell a real conflict of interest from a merely adjacent portfolio company? A direct conflict solves the same problem for the same buyer with a similar model. An adjacency is related but doesn’t compete for the same budget. When in doubt, ask the fund directly rather than guessing.
Can I figure out whether a fund still has money to deploy without access to LP data? Yes, approximately. Recent deployment pace, fund vintage, and the firm’s own announcement of its most recent close are all public and give a strong directional read, even without private LP reporting.
Should I analyze the fund or the individual partner first? The fund first. Portfolio analysis filters a long list down to firms worth pursuing, then partner research (via the 9-step framework) tells you who at that firm to actually pitch.
How do I attribute portfolio companies to a specific partner when the firm doesn’t publish it? Check the portfolio company’s own announcement post, which frequently names the leading partner, along with that partner’s public deal history and professional profile.
How often should I refresh my portfolio analysis during an active raise? Every two to three weeks is reasonable during an active raise, since new deals, fund closes, and partner departures can shift a firm’s fit faster than founders expect.
Does a fund’s exit track record matter if the current investing team is new? Less than founders assume. An exit credited to a partner who has since left tells you little about how the current team will behave. Weight track record by who is still actually at the firm.
A fund’s portfolio page was never designed to help you decide anything. It’s a list, not an argument. Run these seven signals against it, and it turns into exactly the kind of constraint map that tells you where to spend your limited outreach time, and where not to.