Original Title: How to Build a VC Firm
Original Author: @hosseeb
Translation: Peggy, BlockBeats
Editor's Note: Amidst the cyclical noise of "crypto is dead" being repeatedly proclaimed, the author Haseeb Qureshi (Managing Partner of Dragonfly) draws on his own experience to review the process of building a crypto VC from scratch to scaling up, discussing specific issues like fundraising, positioning, winning deals, post-investment support, and team building.
This article dissects the operational logic of VC from a practical level: under a structure of power-law distribution of returns, how to understand "non-consensus judgments," how to view hit rates and concentrated investment strategies, why "winning deals" is more crucial than "picking the right projects," and why this is a business that requires long-term patience.
For those wanting to understand how VCs operate, this is a direct and concrete experience share.
Below is the original text:
I have a bad habit: every time I accomplish something, I can't help but write down how I did it.
We just completed the fundraising for Dragonfly Fund IV, a $650 million crypto VC fund (and at that time, nearly half the media once again felt "crypto is dead"). Currently, we manage about $4 billion in assets, with approximately 45 people in New York, San Francisco, and Singapore, and have become one of the largest VC platforms in this industry where "most people didn't make it."
So, when a few people asked me to write about how Dragonfly got to where it is today, I thought: why not.
To be honest, if someone had given me a blueprint on "how to build a VC from scratch" when I first started Dragonfly, that would have been incredibly valuable. But the reality is—almost no one will tell you these things.
Honestly, this article will probably only be useful to about 0.01% of readers, so taking up this much space to write it might not have much significance. But whatever. If you are considering building a VC, or if you happen to be me from ten years ago—this article is for you.
When I first entered crypto VC, most people thought the industry was already "dead." It was 2018, the ICO bubble had just burst, and the entire industry was in free fall. Most of the people who entered the industry around the same time with me had already left.
But I always believed that crypto is something that is destined to last long-term—it belongs to that category of ideas that once you truly understand, you can never "pretend not to understand." So, when people ask me why I remain so optimistic about crypto, my answer is actually quite simple: if I didn't believe in it, I would have left a long time ago. Now it is too late for me; this optimism has spread to the back of my head.
Thus, when I met Bo and we decided to build Dragonfly together, we didn't expect the market to be particularly enthusiastic. But every VC has to start from scratch.
Lesson #0: For your first fund, you have to bet your life on it
The lifeline of a VC is just one thing: money.
To have a fund, you must first be able to raise money. If you don't have access to funding (or a partner who can help you raise funds), then you are not ready to build a fund.
For the first fund, you must start by raising money from friends. Your boss, your boss's boss, anyone you know who is wealthy and reputable—even if it's just an acquaintance.
If your reputation isn't linked with this fund, it means you are not taking enough risk. I've seen too many first-time fund managers fantasize about keeping their reputation intact "even if the fund fails."
This is a fantasy.
If you're not all-in, you fundamentally have no chance of success. If you fail, yes, you will embarrass yourself and lose some important people's money. But if you want any chance of success, you must leverage every resource available to you to make the first fund happen. If you're not willing to do that, then you shouldn't try to build a VC.
Once you've secured startup capital from those "who have plenty of reason to back you," you have to move onto larger pools of money: family offices (super-rich families), fund-of-funds (funds that invest in other funds), "institutional capital" (university endowments, foundations, sovereign wealth funds).
Generally, this goes from easy to hard, from low to high.
Now you start pitching your fund to these investors with "more money than they know what to do with." But here’s the problem: as a first-time fund manager, what gives you the right to manage their money?
The answer is simple: you must have a clear, expressible advantage.
Lesson #1: Find a niche angle where you excel more than anyone else, no matter how small
When we founded Dragonfly, crypto VC was still a very small field. But even so, at that time, there were already a few dominant players: Polychain, Pantera, a16z. In our eyes, they were untouchable giants.
So, at the beginning, we couldn't lead any projects. No one wanted our money. We had to find an angle that would allow us to "sneak into rounds." Like startups, new funds must focus.
The initial idea was: Bo is in Asia, I am in the US, we would do "East-West connections." Crypto is global, and we could serve as a bridge between Asia and the US, helping founders on both sides enter each other's markets.
This positioning wasn't enough to let us lead. No founder would want an "East-West fund" as the main investor. But it was strategic enough to get us a small seat— and that was enough for us to start pushing our way in.
Lesson #2: Do the dirty work
It turns out that this East-West arbitrage was hardly a competition at all. I puzzled at first: why wasn't anyone doing this obvious opportunity?
Then I understood the answer: because it's just really damn hard.
This means we had to operate a fund across both Asia and the US simultaneously, with an extremely high daily workload; more coordination, more late-night Zoom calls, more language barriers, and almost no normal life.
If it could be done without this level of effort, who would choose this path? But we had no choice. So we persevered. We worked harder than others and dealt with more jet lag.
Many people imagine VC as an elegant profession: summer vacations, quarterly team-building ski trips. We did none of that. No money, no time, no breathing room. The closest we got to "winter sports" was the repeated crypto winters.
Lesson #3: Optimize to the extreme like a startup
Once you have your entry angle and start getting into rounds, the next step is to establish a feedback loop. Investing is essentially a feedback loop, the tighter the better.
Investors demand that startups be highly data-driven and quantified, but they often do the opposite themselves.
You should record everything: your discussions, missed projects, use AI to record and analyze your fundraising and investment committee meetings; review the biggest deals in the industry, figure out why they succeeded, and summarize them into theories; study those great investors you admired before, identifying the commonalities in their successes. Now, with AI, all of this is much easier than before.
But most investors do not care about this. They primarily rely on "feel." Success often hinges more on their luck and networks than anything systematic.
Luck might be temporarily beneficial, but it's not a strategy, nor will it compound in the way that ruthless optimization will.
Lesson #4: Talent is everything
Overall management levels in VC are quite poor, and I mean organizational management. One-on-one communication, mentorship systems, KPIs, clear responsibilities, transparency, all-hands meetings... many VCs do these basics atrociously.
It took me a while to understand the reason: VCs do not "filter for management ability" like companies do.
Companies with poor management will eventually fail; but VC is a power-law industry—if just a few people can produce power-law returns, the fund can survive even if overall management is a mess.
But in the long run, good management itself is an advantage. It can retain the strongest talent and help them grow into the next generation of core personnel. VCs notoriously perform poorly on "generational succession" and internal promotions; many partners even fear hiring younger people who might be smarter than themselves.
At Dragonfly, we attracted and retained individuals who would otherwise go to larger, better platforms. We provided them stability, a voice, and independence, proving through actions that we value them—this is a crucial reason we can outperform our peers.
Lesson #5: Be stupidly ambitious
What continually astonishes me is that most new VCs, when asked "What kind of institution do you want to become?" can't articulate a clear vision. "We want to invest in good companies and be the best partner for founders."
Ugh. That's like an entrepreneur saying, "My goal is to maximize shareholder value."
You need a real ambition, and you need to articulate it.
When we first started, our ambition was straightforward: beat Polychain.
Just that one thing. At that time, Polychain was the benchmark for crypto VCs. Later, when we truly began to surpass them, I realized we needed to upgrade our goal: to be one of the Top 3 crypto funds. This goal drove us for a long time. Now, in my mind, we are already Top 3, so the goal becomes Top 2, then Top 1. As for where we currently stand, I'll leave that for the readers to judge.
Lesson #6: First pretend you've made it, then actually make it
With the first fund, you don’t have a brand. So you must quickly fabricate a sense of branding with the little social endorsement you have.
Get into hot projects, even if the amounts are small. Collect logos, exchanging them for more logos. In Fund I, we wrote very small checks in many popular companies: dYdX, Anchorage, Starkware. These amounts were inconsequential, but these names provided us the foothold to continue going forward.
We called ourselves a "research-driven fund." The so-called research just meant I wrote some blog posts saying, "What if this could be crazy?" We dubbed it Dragonfly Research, and at the time, that actually counted as research.
We claimed to have the strongest connections in Asia. This was theoretically true, but initially, we had no idea what others wanted from Asia. We were telling stories while figuring things out on the fly, and later gradually systematized it. At first, we just did our best to push the narrative out—and it actually worked.
Lesson #7: Trends are not your friend
Resist the temptation to chase trends. The crypto space is full of foolish fads: NFTs, TCRs, P2E, chatbot tokens, VC-backed meme coins...
Our most successful investments often came from avoiding the madness— and doubling down when others dropped out of a track. We didn’t touch Terra, Axie, or Yuga; after Terra collapsed, we invested in the seed round for Ethena; we invested in Polymarket before the 2024 election frenzy.
Every cycle has a narrative that is hard to resist. You’ll feel pressure from your team, LPs, and Twitter. But most hot topics will ultimately prove to be a waste of money.
The real difficulty lies at the psychological level. When you’ve passed on a project everyone is rushing to get, and it doubles in value the next week, you’ll feel like an idiot. But chasing trends often results in a portfolio of "projects that were popular 18 months ago"—the worst kind of allocation.
Your job is to invest in what will be important 3–5 years down the line, while hot markets almost never embody this forward-thinking.
Lesson #8: Control your distribution capabilities
It was once said that a16z is a "media company with a VC business," which was a joke, but is now just fact.
VC is fundamentally a storytelling business. You must build an audience, making the entire team a signaling source. Encourage team members to build personal brands and reward them for speaking up. The brand of a VC, unless you're Sequoia, is almost entirely attached to specific individuals. This is a "people" business.
Some funds even prohibit employees from tweeting, which I can’t understand at all. If you want founders to be adept at social media, why can’t you be?
Lesson #9: Cultivate power
This is a crucial step for a fund transitioning from novice to heavyweight player.
As Dragonfly began to gain influence, many doors started to open automatically. Exchanges, banks, market makers, even projects we had never invested in sought to build relationships with us. Initially, I felt this was a distraction: why chat with old institutions instead of looking at new projects?
Then I realized: the essence of VC is branding money. You win a deal because founders believe your money is better than others’. In reality, all money is green.
Marc Andreessen once said: the job of a VC is to lend their brand and power to those who don’t have them yet. Therefore, you not only need a brand, but also influence. Founders want to know if you can get them into rooms and if your words carry weight.
As the fund grows, you must evolve from a mere investment institution into a platform. The best founders want more than just capital; they want to know if you can truly help them move things forward. We built a platform team at Dragonfly that supports everything from token design to exchange listings to executive hiring. It’s not sexy and doesn’t directly create returns, but it compounds. Once the flywheel starts turning, it becomes hard for competitors to replicate.
Lesson #10: Almost all the money comes from a tiny number of deals
There’s a simple matrix that can describe the essence of VC investing.
Many hot projects are actually transactions where "consensus is correct." In other words, most people believe this company will win, and it ultimately does win. Such transactions are usually not bad, but it’s hard to make a lot of money because the prices are often bid up high due to intense market competition.
Almost all genuinely profitable returns come from those transactions that are "non-consensus but correct." This is because these kinds of transactions often have a structural underpricing, and the probability of achieving returns over 100 times mainly comes from here.
Venture capital returns follow a power-law distribution, and the math is ruthlessly cold. In a typical fund, the returns from the top three projects often exceed the total returns from all other projects combined. This means most of the deals you make are relatively unimportant when looked at individually. What truly matters is whether you hit that one or two projects that define the entire fund cycle.
This leads to a counterintuitive conclusion: your hit rate is almost unimportant. What's really important is how many "big punches" you've thrown. Thus, when looking at each project, you should ask yourself one question: does it have the potential to be a "fund-returning project"?
If the answer is no, then why would you make this investment?
And that equally ruthless corollary: consensus deals almost never lead to such outcomes. If everyone thinks a project is great, the price has already reflected that, and your upside is capped. The truly intergenerational investments are often those that other smart people would think you’re foolish to invest in.
Lesson #11: If you can't win the deal, everything above doesn’t matter
The value chain of VC can be broken down into four stages: Sourcing (finding projects) => Selection (judging) => Winning (winning deals) => Supporting (post-investment support)
Finding projects is the first step for new VCs. You must establish an engine that can reliably source new projects.
Judging is what most people think is the most important skill ("selecting projects"), but in reality, it only constitutes a small part of the whole game.
Winning deals is the most important step. Even if you have the world's best project sources and the sharpest judgment, if founders choose others, then everything is meaningless. At the highest level of venture capital, the truly scarce resource is "access opportunities." The best founders are often over-subscribed; they can choose their investors at will. Therefore, you must provide them with a reason to choose you. This comes back to brand, platform capability, and the relationships and reputation you have built over the long term—all the lessons learned above will ultimately converge here.
Post-investment support is the final step and will also reinforce the earlier steps of "finding projects" and "winning deals." Support determines your NPS (Net Promoter Score) and whether this cycle can continue. If you truly stand by the founders, they will become your best salespeople: introducing you to the next excellent founders and supporting you in small groups. This industry is small and closed, and reputations spread rapidly. An angry founder could ruin dozens of future deals, while a genuinely satisfied founder could open doors for you for the next decade.
Lesson #12: Venture capital is a "get rich slowly" business
You will see many people in this industry rapidly rise to success like shooting stars.
You must endure through them. Some make their money too fast, others become lazy, gradually believing they "deserve this success." The crypto space is particularly brutal in this screening. Each cycle births a group of overnight millionaires; yet, in each cycle, most of them will disappear. Traders who made 50x will retreat into obscurity; founders financing at absurd valuations will quietly shut down their companies. Ultimately, tourists will leave the scene.
You are not a tourist. In VC, progress takes many years to measure. There are no real "overnight successes" here. Most of the value in your fund often remains unrealized for years. This means that you embody that famous New York Times article—
That’s okay.
Your job is to keep the ship steady. Floats, debris, rising tides, falling tides—all of these will happen. You must always stand there, with your team, with your founders, with the entire ecosystem. The compensation you receive is to act as long-term capital.
So, it really has to be long-term.
Lesson #13: Fundraising when things are going well
Founders hate fundraising as much as VCs do, and it’s not easy at all.
Fundraising as a VC operates under a completely different cultural system from that of founders. I came from a middle-class background. When I was a professional poker player, I thought I had seen "rich people." Later I realized—this is not even on the same scale.
Fundraising itself is an art and is highly dependent on who you are facing.
When raising funds from family offices, relationships are key. These are wealth families with intergenerational legacies, each operating under unique logics, and establishing trust takes time. They rely heavily on social endorsement.
Institutional capital and fund-of-funds are another breed: process-oriented, heavy on due diligence, valuing forms over intimate dinners. They want to see performance, processes, and a sustainable edge.
To be a truly excellent fundraiser, you must learn to speak both of these languages fluently.
But in general, successful fundraising has one prerequisite: you need to be operational, have returns already, or if you don’t have returns yet, you have to tell a damn good story about where those returns will come from.
Finally, and most importantly: timing is everything.
LPs almost always buy high and sell low. So, you should do the opposite. The reasoning sounds simple, but the actual implementation is excruciatingly painful.
Your best fundraising window often comes when the market is hottest, and LPs are most excited—yet that is precisely when you should be careful with your investment allocations. And when the market has hit the bottom and everyone is feeling low, that’s precisely when LPs least want you to invest—but that is actually the wrong approach.
The best top VC knows how to fundraise when conditions are optimal and deploy capital when asset prices are at their best. And these two things almost never happen simultaneously.
Above are some of the lessons I learned while building Dragonfly. I am sure I missed some and undoubtedly, there are many lessons I have yet to learn.
Building a VC is a constantly changing endeavor. Each cycle introduces a new cast of characters, always bringing mistakes you could entirely avoid yet are still lurking around the corner.
But the underlying principles remain unchanged: stake your reputation; find your advantages; do the dirty work that others refuse to do; hire people better than you and truly treat them well; and—maintain patience.
Venture capital ultimately rewards those who persevere long enough to see the other side of the cycle.
This is certainly not "the final answer on how to build a VC." But it is the kind of article I truly wish someone had written for me back then. I hope it helps you. If you are doing something cool in the crypto space, feel free to reach out to me.
Disclaimer: This article does not constitute any investment advice. Building a VC fund is hard, and you will likely fail. But who knows—maybe you should give it a try.
Good luck.
免责声明：本文章仅代表作者个人观点，不代表本平台的立场和观点。本文章仅供信息分享，不构成对任何人的任何投资建议。用户与作者之间的任何争议，与本平台无关。如网页中刊载的文章或图片涉及侵权，请提供相关的权利证明和身份证明发送邮件到support@aicoin.com，本平台相关工作人员将会进行核查。
