Nov 21, 2022 · 1h 24m · news
Semil Shah: Lessons Learned Scaling from a $1M to a $50M Fund | 20VC #951 · 20VC with Harry Stebbings
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
In this 20VC episode, Semil Shah of Haystack joins Harry Stebbings to share vital lessons on maintaining a disciplined fund size, optimizing ownership targets, navigating LP dynamics, and avoiding the systemic valuation traps of market bubbles.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. Harry holds 26% of the talking time here. How this is scored →
Harry as informed peer 5.3 Guest teaching 4.0 Guest disagreement 2.3 Harry pushing back 4.0
speaking balance: gold is Harry, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 17:01 Rejecting competitive obsession over lost deals
Semil directly rejects Harry's premise that investors must be hyper-competitive and upset whenever losing a deal, advocating instead for a collaborative, Zen-like mindset.
Hardest push from Harry ▶ 13:23 Challenging Semil's refusal to scale fund sizeHarry aggressively pushes Semil on why he continues to keep fund sizes at $50M rather than scaling to $100M to expand diversification and check size.
Biggest teaching moment ▶ 1:18:10 Explaining early fund economics through HashiCorpSemil educates Harry on the extreme power law of tiny micro-funds by breaking down how a $25k initial check in HashiCorp returned over $30M for a $1M fund.
Harry holds his own ▶ 54:33 The Chanel vs. Walmart VC market thesisHarry demonstrates deep market knowledge by synthesizing venture landscape trends into a clear thesis arguing that only boutique product firms (Chanel) and mega-funds (Walmart) will thrive.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | Harry as informed peer | Guest teaching | Guest disagreement | Harry pushing back | Why |
|---|---|---|---|---|---|---|
| Semil's Motivations: Running Away vs. Running Towards | 2 | 1 | 1 | 1 | Harry opens with an exploratory question about Semil's personal and professional motivations regarding running away versus running towards something. Semil reflects thoughtfully on early fears of failing out of the Bay Area ecosystem without any friction. | |
| Determining Haystack's Optimal Fund Size | 5 | 3 | 2 | 3 | Harry introduces specific market data on pre-seed round valuations, noting he rarely sees rounds under $3M on $15M caps. Semil breaks down how inflation doubled sub-$1M cap rounds and explains the discipline behind keeping Haystack funds under $100M. | |
| Identifying the Overlap of Talent and Entrepreneurial Grit | 6 | 4 | 2 | 4 | Harry shares counter-intuitive portfolio data showing high-profile heads of product from major tech firms often stall as founders. Semil elaborates on separating raw corporate talent from genuine entrepreneurial grit and hustle. | |
| The Power of Flexibility: Resisting AUM Scaling and Leading vs. Participating | 6 | 4 | 4 | 7 | Harry explicitly pushes Semil on why he refuses to scale fund sizes to $100M and questions whether co-investing flexibility leaves him vulnerable to aggressive lead investors. Semil resists the framing, advocating for a Zen approach to syndicate building. | |
| Ownership Floors, Deployment Speed, and Fund 4 Mistakes | 5 | 4 | 2 | 5 | Harry probes Semil's ownership thresholds and deployment pace, highlighting how competitors rushed to 12-month deployment cycles. Semil candidly admits selection errors in Fund 4 while defending his 24-month pace. | |
| Loss Ratios, Picking vs. Access, and Salvaging Fund 4 | 6 | 5 | 2 | 4 | Harry shares his tactical playbook for securing LP introductions and questions the utility of traditional pitch decks. Semil shares candid insights on LP psychology and emerging manager positioning. | |
| LP Terms, GP Stakes, and the Standard 2-and-20 Model | 5 | 4 | 2 | 4 | Harry presses on whether emerging managers should offer discounted carry or sell GP stakes to launch. Semil utilizes an airplane takeoff analogy to explain when temporary concessions make sense. | |
| Navigating LP Relationships: Lessons in Churn and Bad Behavior | 4 | 5 | 3 | 3 | Semil recounts stories of being abruptly dismissed by prominent institutional LPs and having to block unreasonable LPs from future funds. Harry sympathizes with the rarity and difficulty of managing bad LP behavior. | |
| Macro Venture Dynamics and the 'Chanel vs. Walmart' Framework | 7 | 4 | 3 | 5 | Harry introduces his structured 'Chanel vs Walmart' framework on venture capital bifurcation between boutique specialists and mega-funds. Semil offers a counter-perspective, noting that a single breakout deal can elevate mid-tier funds. | |
| LP Mistakes and Running a Modern Venture Endowment | 7 | 4 | 3 | 5 | Harry plays devil's advocate regarding LP allocation rigidities and outlines how he would run an endowment by anchoring hungry spin-out GPs. Semil details the ground reality of LP relationship maintenance. | |
| The Sins of VCs and Founders in the New Era | 5 | 5 | 2 | 3 | Semil outlines the structural sins committed by VCs and founders during the zero-interest-rate environment, including treating venture capital like speculative real estate. Harry agrees on the impact of poor founder communication. | |
| Quick-Fire Round: Big Wins, Major Misses, and Crucial Advice | 5 | 5 | 2 | 4 | Harry conducts a fast-paced quick-fire round covering portfolio write-downs, angel picks, massive early wins ($25k turning into $35M on HashiCorp), and painful misses like OpenSea. |
Statements from this episode (39)
Shah: California is the only place to operate Haystack effectively
“I feel like it's the only place I can live and do what I want to do in the way that I want to do it.”
Shah: Most venture investors cannot build companies like entrepreneurs do
“A lot of us as investors, for the most part, we can't do what entrepreneurs do. We choose not to do what entrepreneurs do.”
Shah: Haystack missed fundraising targets for its first four funds
“Well, remember the first four, I was never able to reach the target of what We wanted to raise. So there were always underwhelming in that sense”
Shah: Future Haystack funds will remain capped under $100 million
“Our next funds, by the way, will modestly grow, but I still think, you know, for what we do, it's gotta be under a hundred million dollars.”
Shah: Pre-seed startup round sizes doubled over the last decade
“Over the last 10 years, just simply because of inflation, those are all doubled, you know?”
Shah: Haystack achieved nearly 7% initial ownership in recent funds
“We've been able to get close to seven percent initial entry ownership in the last two funds.”
Shah: Haystack's 10-year deal valuations clustered between $8M and $15M
“If you look at the activity of all the deals over, over 10 years, I would say there's like a tight clustering between eight and fifteen million dollars as the entry point and the rounds being, you know, three million dollars or less.”
Stebbings: Hot seed rounds raised by executive founders consistently underperformed
“I looked at the seeds, and I looked at those that haven't really worked, and all of those that haven't really worked have often been head of product at Twitter, head of product at Twitch, head of product at amazing company, the hottest of hot seed rounds with …”
Shah: Big Tech product managers hold political, not entrepreneurial, roles
“Once you make it through to like certain organizations and you are in this PM role, you do have more of like a political job than you do a breaking glass job.”
Stebbings: Competitive Series A rounds correlate to hot Series B rounds
“The series A trend that I noticed was the hot A's absolutely do correlate to hot B's, weirdly. The hot seeds often don't. Hot A's very commonly moved to hot B's.”
Shah: Haystack led only about 10 deals over three years
“If you look over the last three years we've maybe led, you know, 10 deals. Ten-ish.”
Shah: Forcing Haystack to lead every deal lowers portfolio quality
“If we tried to lead everything, we would lower the I think the quality of the portfolio that we're in.”
Shah: Scaling seed funds to $100M+ breaks collaborative syndicate dynamics
“If I had a hundred, fifty million dollar fund, that whole collaboration layer of how we work would change, which would then set off a chain reaction of how the deal flow would change and how we would have to operate with the companies would change.”
Shah: Haystack sets initial ownership floor at 5%
“I would say we stop around five percent. So if you look, yeah. So if you look at like what we tell the entrepreneur is like, look, we're trying to own five and 10% to start. We want to earn the right to maintain that over a few rounds, right? And we'll never o…”
Shah: Large VCs are too slow for seed deals despite Twitter claims
“And also like these funds are really big and slow. So they're not going to be able to like move and prosecute the things in the same way that they say on Twitter.”
Shah: Haystack operates on a 24-to-30-month fund deployment cycle
“What I kind of concluded over that time and what I told our LPs is that for the seed market I felt like the thing we could promise is 24 to 30 month deployments. And then doing, doing new, new funds. And I would say that since we last spoke, I would say it's c…”
Shah: Haystack Fund IV made selection errors by prioritizing ownership targets
“I think in fund four, I always worry about like the carnage in that fund. And I knew there was going to be turbulence because it was a first time was really going for ownership. And I knew in going for ownership to our earlier part of the conversation that the…”
Shah: LPs expect one spectacular fund and one failure across five funds
“I remember an LP telling me a long, long time ago, like just when I was starting that typically they'll stay with good managers for five funds and they assume that three of them will be so, so maybe two to three X funds. One of them will be a complete turd and…”
Shah: Haystack Fund 4 holds 5%-10% stakes in potential IPO candidates
“So the good thing about all of this is that we have like five to 10% positions in, in three or four, like really amazing companies that all could be public actually at some point.”
Shah: Emerging managers must build an AngelList track record before pitching LPs
“I think one is like on AngelList now, there's really no excuse not to have some kind of track record before you go talk to an institutional LP.”
Shah: Institutional LPs are predisposed to pass on debut funds
“If you go meet an institutional LP, they're predisposed to passing on your first fund, primarily out of the fact of like, they probably don't think you're going to raise your first fund or deploy it well, if you do.”
Stebbings: Meeting an LP for the first time during a raise guarantees failure
“I meet two new LPs every single week, even when I'm not raising, because I think if you ever meet them when you're raising for the first time, you've got no chance.”
Shah: Raising institutional capital requires organic industry chatter and name-checking
“I think that like, if you're not getting name checked by institutional LPs and other people aren't talking about what you're doing, it's almost impossible to raise institutional capital, whether it's a three million dollar check or thirty million dollar check …”
Shah: Limit preferential LP terms strictly to a fund's first iteration
“And so like, basically if you, if you're trying to offer preferential terms to get the jet off the runway and get some thrust, That's fine, but just don't have it be successive in, in future funds. Like, gate it to that one fund.”
Shah: Union Square Ventures is the best pound-for-pound VC fund
“I just, I follow a lot of honestly what USV does of like try to learn. I think they're the best pound for pound fund in the world both operationally and in every way.”
Shah: Haystack maintains a standard 2-and-20 fee structure across all funds
“And so I think Fred has been on the record of just saying like, they just, you know, do a standard two and 20 on every fund. So I haven't really Seeing the need to deviate from that.”
Shah: I blocked two LPs from re-investing in future Haystack funds
“I've blocked two LPs from investing in, in future funds.”
Shah: Crosslink acquired over 20% of Chime through a modest early investment
“Look at Crosslink. I mean, no one really talks about Crosslink. I don't even know if you've had an investor from Crosslink on, on 20 Minute VC, but they ended up owning like over 20% of Chime you know, through, like, a pretty modest deal because no one was loo…”
Shah: LPs' biggest mistake was permitting fund size escalation and blind pools
“I think on the LP side, the sin was going along with fund size escalation and allowing and enabling those firms to allocate within those funds. So I believe like In having separate vehicles for separate separate activities, because that's the best governor. An…”
Shah: Only 50 to 75 VC funds globally can effectively guide IPOs
“I think that there are a set of funds, maybe like 50 to 75 VC funds that in partnership with other funds and entrepreneurs know how to guide a company towards a big outcome or a public offering.”
Stebbings: Board member value follows a barbell distribution, skipping mid-stage VCs
“The other truth is I think there's a utility value to board members that resembles a barbell, which is like Fenton, Gurley at the A you know, Alfred Lynn at the A, Roloff Botha at the A, amazing Pat Grady at the A, great, and then your pre-IPO people who are a…”
Shah: Multiple startups hold $50M+ cash without product-market fit
“There's probably multiple, multiple companies with like over fifty million dollars of cash with no product market fit.”
Stebbings: Series B investors are offering founders $1M payouts to return capital
“You know, it is, but I'm seeing now I've seen four this month where the series B's are saying, Hey, preemptively we fucked up, but give us our cash back and we'll give you a million each.”
Shah: Traditional VCs avoid asking for capital back over reputational concerns
“I think a lot of traditional VCs are loath to do that for reputational reasons.”
Stebbings: Skipping investor updates halves follow-on commitments in next rounds
“I think if you don't send the updates, then you're halved in terms of re-commitments for next round.”
Shah: The venture capital downturn will be prolonged
“It feels like we're just in the early innings of this like administration of pain or like the game hasn't even really started. So it feels like it will be quite prolonged.”