I asked Rachel Kloepfer what she actually asks fund managers. Not the scripted portfolio construction question. The one that gets her an answer she would not otherwise get.

She started with what does not work. A manager says sourcing, then says events, then says a network from whatever large company they worked at. “That’s an obvious give,” she said. Then community building, which she flagged as a buzzword before I could. Then the question itself: “pull up your calendar and let’s talk about what you’ve done this week, last week, whatever.”

That request is good because it cannot be answered with a narrative. Everything above it can.

Rachel spent her first career in investigative and crime reporting before landing at Lenora Capital, the private investments arm of Brighton Jones. Most of what I took traces back to that.

Three things.

First, the edge is the records, not the questions. I expected the journalism answer to be about asking hard things of people who do not want to talk. She said that, and it is real. The sharper piece was procedural. She filed a lot of public records requests, and the skill that transferred was knowing how the request process blocks access to information that is already public by law. Mess up the form and you lose a month. Know it and you get in. She described knowing “who you needed to get on the phone and yell at” at day thirty. She now uses FOIA to pull endowment and pension investing strategies and watch how they shifted over time, since all of that has to be publicly filed. I have been around this seat for years and had not thought of that as an LP tool.

Second, my own analogy came back to me changed. I put the barbell to her as an NFL team against a tennis pro. The big firm has a brand that moves on its own. The emerging manager is one person walking into tournaments on their own merits (and own money, look up how a professional tennis player operates!!!). She took the framing, then put the advantage somewhere I did not expect.

The standard pitch runs the other way. The emerging manager claims the fast categories and the mega fund is the safe institutional choice. Rachel’s read: if you are underwriting SaaS or consumer, distribution is the moat, and in her personal opinion you want to be with a big fund. Where a smaller GP is genuinely strong is niche knowledge in a hard industry. She listed life sciences and bio, hardware, anything selling to the government, anything tied to blue collar logistics. In those, the knowledge is the moat.

Her reasoning is what makes it hold. Nobody can forecast this market, GPs included. “There is no all seeing eye that’s going to have the right answers.” So the riskier and faster the area, the more she wants it sitting with a firm that has enough manpower and brand power that “they can will their way through it.” That is not a quality judgment on the manager. It is a judgment about what survives being wrong.

Third, succession planning is a live reason she says no. Not a governance checkbox. She described firms she would have said yes to at first touch, among the best she can think of, that got a no on succession alone. The mechanism is specific. A GP who was senior at Google pulls founders out of that orbit because those founders respect the relationship. When the GP goes, the pull goes. The commitment does not. A ten year fund with extensions, re-upped to hold an allocation, means in her words “we’re looking at a 20 year relationship.” The line that stuck with me was about opening a deck years in and not recognizing a single name on it.

Where I am not settled is what that does to the media argument. Rachel thinks owned media decides the next decade. Streaming wars, entertainmentification of the whole economy, an A16Z documentary she would not be surprised to see. Eyes bring LPs, founders, talent, and deals at more disciplined prices. I co-host a podcast about LPs, so I did not need convincing. This conversation confirmed what I already believed rather than changing it, and I would rather say that than manufacture a turn.

But she also said the underinvestment in succession happens partly because so much focus goes to the media presence of the firm. Those two sit uneasily next to each other. The same build that wins attention now may be the one starving the bench that has to carry the brand in year fifteen. I did not push on it in the room. I do not know how she would resolve it.

Kelli Fontaine came at this from the other end on episode 105. Her team runs 78 dashboards and she is the most data-forward allocator we have had on. Rachel came in from crime reporting. They land in the same place. There is tactical benchmarking, Rachel said, and then there is the person to person work that sits “outside of what you can put on a spreadsheet,” and she called that second part a finger to the wind. Two unrelated careers, same conclusion about where the numbers stop.

The calendar question is what that conclusion looks like once it is operational. Not more data. One question a story cannot survive.

Listen here: https://swimmingwithallocators.com/podcast/what-lps-really-want-from-venture-funds/

With gratitude,
earn

I read a line off Henrik Reimavuo's own website and asked him to defend it. Firm Capital backs next-gen managers who operate at the esoteric edge. I wanted to know what that meant and how he underwrites it. He laughed. His colleagues had asked him the same thing when he wrote it, he said, and they kept the phrase because "we like the sound of it." Then he gave the real answer, which was better than the line.

The real answer is that he is not shopping for agreement. "Everybody talks about consensus," he said. He wants managers who can teach him something he cannot read from the outside. That is only a strategy if the market really does have parts nobody can read.

I brought Henrik the consensus framing on Europe, the one that comes up on this show constantly: the gap with the US is a capital shortage, so put more euros in. He does not buy it. He calls it a quality problem and a misallocation problem, and he puts the shortage at the wrong end of the pipe. By his numbers, angel capital in Europe runs roughly 20x below the US and pre-seed runs 10x to 13x below. The gap does not start at growth. It starts at the top of the funnel.

Then the mechanism. He puts governments at around 40 percent of the capital backing European venture funds. Government money carries government objectives: policy, employment, regional development. None of those is performance.

The detail that made it land was not a number. Henrik said managers who used to be sustainability investors are now investing in weapons. Same people, same funds. The narrative moved from decarbonization to European resilience and defense, and the money moved with it, because it was following the narrative and not the edge.

Three things I took.

First, the unmarked manager. Henrik sorts first-time GPs into archetypes, blunt about the cost of each. The spinout competes forever against the brand they left. The operator wears the tattoo of the company they built. The community catalyst has neither. He calls that being "unmarked." In soccer, unmarked is not a skill. It is a fact about where the defenders chose to stand. LP coverage is built to scout spinouts and operators, so the person who spent four years convening a group nobody scored as a market is standing free. The run only counts if somebody passes. That is the gap Henrik wants Firm inside: those managers usually get their day-one check from a US GP or from angels in their own network, and almost never from a European institution.

Second, the sequencing test. I pushed on community, because being a node in the ecosystem is not new and half this industry has a podcast now (this guy!!!). His filter is order of operations. Did the community come first and the fund arrive as a consequence, or was the community assembled to justify a raise? His examples were specific enough to check. One person built for the non-executive operator layer inside European startups, a layer with no room of its own while founders, juniors and investors all had theirs. Another convened post-Soviet founders across cities worldwide. Both are raising funds on top of networks they did not build in order to raise a fund.

Third, fragmentation as an asset. Henrik argues Europe's cultural fragmentation is "a feature, not a bug," that you can be the Swiss German investor to a Swiss German founder permanently and hold your seat because you understand each other. I believe that at entry. I have watched local relationships hold cap table position through rounds where nothing else would. I am less sure it survives scale, because the fragmentation that protects you at entry also caps how many companies you can serve. He would say that is why he writes into sub-$100M funds and keeps the portfolio concentrated. I want to see a vintage.

I asked Senia Rapisarda of HarbourVest a near-identical question about Canada in episode 103. She refused the capital answer too. She called ecosystem building agriculture rather than geology: a bigger drill does not find the nugget. What is unresolved for me is that Senia helped design VCAP, a government program, and credits it with catalyzing Canada, while Henrik points at state capital as the source of the misallocation. I cannot draw that line cleanly yet. My working guess is that the state is good at planting and bad at picking.

Where I did not move: I walked in believing "more capital" is almost never the real diagnosis for an ecosystem, and nothing here changed that. It sharpened it. The question is not how much capital. It is who the capital answers to.

He closed on something I did not expect. I asked what he wished more LPs would say out loud. Answer your emails, he said, and say no clearly, face to face, with the actual reason, which he says is usually a human issue rather than a portfolio one. He described himself as a notorious people pleaser who has worked through some of it. In a business that hides behind portfolio construction language, that is harder than it sounds.

Listen to the full conversation here:

With gratitude,
earn

I said something on the record I have mostly kept to private conversations. When I moved to San Francisco after business school and started interviewing for associate and principal VC roles, I was grateful that nobody could tell how old I was. My genes bought me cover. I needed it, because I heard over and over something I called the rule of 27. If you had not done something exceptional by 27, why would a top tier firm take you.

I do not think I was wrong that the rule existed. I was wrong about what it measures.

Santosh opened by asking whether there is a sweet spot age for a high functioning VC. I swapped the word almost immediately. Not age. Lifestyle. This work feels a lot like life. You meet people and nothing comes of it. You back people and it goes to hell or heaven. What the job asks for is room to think and the willingness to keep showing up, and both of those track with age without being caused by it.

Then he brought numbers that cut against the folklore. Across roughly three million U.S. founders the mean age is 42. Among the highest growth one in a thousand it is 45. Founders at 50 and over are twice as likely to deliver a top tier outcome. The industry that invented the rule of 27 writes its checks to people who are, on average, fifteen years past it.

An aging curve is a population object and we keep using it to price a person. Basketball solved this a while ago. The curve says a wing declines after 30. LeBron at 40 does not refute the curve. He is the residual, and the residual is the whole business. I said on air that the goal is to be an exception that invests in exceptions. You are trying to be the weird dot on the chart, the one that should not be there. If that is the job, a hiring band is a strange instrument to run it through.

Three things I took.

Risk tolerance may be inherited rather than aged into. Santosh gave the standard version. You will never be more risk seeking than right out of school, because responsibility becomes liability. I pushed on where that belief comes from. We were raised by boomers, personally and professionally, and that generation was handed a world that mostly went up and to the right. Some of what they taught us was true then and is not true now. I am not sure the risk curve is a fact about human beings. It may be a fact about one fifty year run.

Staying too long is a franchise cost before it is a personal one. Santosh is the youngest of four partners at Dynamo and the firm is coming up on eleven years. He looks at other professions carrying partners at 65 and 75 and 85 and sees damage. "It doesn't matter if your name's on the door," he said. "That's still net negative for you and your legacy if you're a founder of a firm." He raised Jordan going to the Wizards and was honest that ego sits underneath the whole question. LPs ask him how many funds he has left in him. It is a rude question and it is the right one, asked badly.

Experience might be the depreciating asset. I did not walk in with this one, and it went against me. NVCA data shows the share of partners at their current firm ten years or more has risen over the last four or five years. I offered the flattering read. Founders want expertise, pattern recognition compounds, the market is finally paying for people who have seen a few cycles. Santosh agreed that every firm benefits from someone who has seen multiple cycles, because history rhymes. Then he kept going. "Experience alone, I don't think, will be valuable because in the age of intelligence, you can tap into experience in a non-humanistic way that's still sufficient."

I have been sitting with it since. If experience is retrievable, the ten year tenure number stops being evidence of compounding and becomes a question about what those ten years still buy.

Where I am not settled. I still believe pattern recognition compounds. What I can no longer defend is the assumption that it compounds faster than it commoditizes. That is the same clock Santosh put on every other venture advantage on episode four, and I did not notice he had turned it on us.

I wrote in Everyone Shoots Threes Now that a weakness is a hiring problem and not a prompting problem. The other half of that showed up here. Knowing precisely what you spike at beats being adequate at three things. When you are young you try to spike at all of them and dilute the one you are actually good at.

There is no perfect age. There is a weird dot, and the work is getting yourself onto it.

Listen here.

with gratitude,
earn

Some of my favorite teams ever never won anything. Hell, my Cowboys have not won a Super Bowl since I was eleven.

Read more

I asked Michael Wooten a question I have been chewing on for months, and he answered a better one.

My question was about telling versus showing. I keep meeting emerging managers who are excellent at telling, and whose response to a hard fundraise is to tell more. I keep meeting others who are quieter, who have something real, and who cannot get it across a table. I wanted to know how a manager shows differentiation when it is not obvious, not early, and not loud.

Michael did not take the communication framing. “Sometimes people want you to be differentiated, but they want you to be on the edge of the box,” he said. “If you’re too differentiated, then that’s probably going to be tough for the investment committee.”

I had been treating differentiation as a storytelling problem. He handed it back to me as an underwriting problem.

In basketball (yes another basketball anology) the corner three is the best shot on the floor and the tightest one. Twenty two feet instead of twenty three point nine, worth the same three points. That foot and a half is the entire edge. It is also the spot where the sideline crowds your heels. The efficiency and the out of bounds line sit on the same piece of floor. You do not get one without standing next to the other.

That is the edge of the box. It is where the return lives and it is where a committee stops being able to price you.

Three things I took from the hour.

Differentiation has to be legible, not just real. Michael’s point was not that managers should sand down their edges. It was that an LP has to be able to write you up. Storytelling is necessary and it is not sufficient. He was direct that execution has to back it, and that at the earliest stage you are selling the sizzle because the steak does not exist yet. By growth you have the steak and you had better lead with it. The failure I see most often is not the manager who is too plain. It is the manager who is genuinely differentiated in a way that has no comparable, and who reads a committee’s inability to underwrite them as a lack of conviction.

Sourcing is the product, and the product is for founders. Michael walked through why SignalFire stood out to him. They built proprietary data early, used it internally for sourcing, then turned the same thing around and gave it to their founders. That second move is the one that mattered. It stopped being a sourcing edge and became a winning edge. He said they took him through the track record and showed where they beat tier one firms already on the cap table, and ended up with more ownership than they had rights to. Underneath it is a claim I now think is just true. Of capital, he said, “that’s not going to be enough because that’s a commodity.” Founders want three things from an investor, in his telling. Introductions to customers, capital, and help recruiting. Two of the three are not money.

The thesis push may be eating the thing it was meant to prove. This was his hot take and it is the part I am still turning over. He thinks sector theses work at growth and do not make sense early. “I think your thesis should be to find great founders who are trying to tackle hard problems.” Then he closed the loop. LPs want managers in a box. Managers reverse engineer a thesis to get into the box. “you’ve crowded out that market and now you’re no longer differentiated because everybody else is doing that.” The differentiation requirement manufactures sameness. That is a mechanism, not a complaint.

Where I am not settled. I let that one sit in the room without pushing on it. I have watched sector focus do real work at pre seed, but the work it does is access, not picking. A narrow lane gets you into founder networks a generalist never sees. Michael would probably say fine, that is a sourcing edge, call it what it is and stop calling it a thesis. He may be right. I have not landed.

One more thing worth saying plainly. When I asked what he actually watches in diligence, none of it was the deck. He watches how a manager treats waitstaff. He hands them a question thirty minutes early, teach me something, anything, and then watches whether they can carry a conversation while it sits in the back of their head. He built that from hiring people at a former job and thinks manager selection is not that different. And his pet peeve was not a bad quarter. It was silence. The manager who stops responding after the money is in.

I wrote in The Songwriting Credit that sourcing is the part of this business people credit last and depend on most. Michael got there from the other side of the table and landed in the same place. That is usually a sign the thing is true.

Listen to the full conversation with Michael Wooten here.

With gratitude,
earn

Two years ago I spent a lot of money on Replit trying to build a Cerebro, an agentic CRM that would let me blow past Dunbar’s number. My family asked what I was doing. The best answer I had was that it was cool. Vibe coding, just vibes.

I started that story on episode four of Carry On, and Santosh cut in with the specifics before I got to them. Five hundred plus. Over a weekend. I finished his sentence. That you’re not even using. Everybody building has one of those.

I brought the pattern; Santosh supplied the part I did not have. My list was firms that focused on data, firms that built platform services, firms that specialized in one industry. Each started as a reason an LP should pick you and ended as a line in every deck. What I asked was what happens to whoever shows up late.

He answered with a clock. In the prior cycles, novelty commoditized in roughly one to two fund cycles. He claims that AI accelerates that, because commoditizing defensibility is what AI does to the industries it enters. Then he ran it forward. We are sitting here in 2026; three years ago was 2023, and by his read the cycle has arguably run once already. He kept the hedge and so will I. On adoption, he is not hedging. “You cannot do it.”

I reached for basketball. When the Sloan people let us idiots know that a thirty-five percent three-point shot beats a fifty percent two-point shot, every team started shooting threes. The insight was correct, and it stopped being an advantage the moment the market was corrected. What separated the Warriors was not the math. It was Klay Thompson and Steph Curry, and Draymond Green, who survived the space they opened.

He put it as a maybe rather than a verdict. “If you don’t have anything unique and advantageous without AI in the equation, AI isn’t going to actually bring that to you.”

My framing has been that tools should magnify what you are already great at, by ten or a hundred times. I said it again on air. Santosh took it and added the half I had never stated. “Don’t use AI to help you improve a weakness that might be better done by actually adding to your team.”

That is a hiring rule wearing an AI costume. A gap in your firm is a person-shaped problem. Covering it with a model leaves you mediocre at that thing permanently, with better documentation. I said absolutely, and we moved on. I have chewed on it since. In the same stretch, he described asking his principal what she thought her superpower was, and framed his read as an argument rather than a finding: most funds probably have duplicative superpowers on the team and have done okay anyway.

The thing I keep returning to is the least glamorous. He thinks the skills file will become one of the more important artifacts inside a firm, along with every reference it reaches into. “A lot of how you get AI performant is not different than when you have new hires coming in. You need to document how work is done.”

At Dynamo, an agent drafts the portfolio update and headline numbers, then several people comb it for accuracy and completeness, which frees the quarter for the deep thesis piece. What I liked most is what AI did to his old process docs, some untouched for two years. Some he improved, because a model holds more context than a person. Others he decided did not need to exist, since they only existed to account for a human in the loop. That is operations hygiene in a new hat, and the least copyable thing in the episode.

Where I am still working it out is taste. Santosh asked whether you could teach it to a model. I said I never like to say never with technology, and then I said today, no. I believe both halves and know they do not sit together.

The bigger tension is with my own first episode. Motion is not movement. I argued in July that venture mistakes activity for progress, and spent a good part of this one advocating for activity. Tinkering that changes what you ship is movement. Tinkering that only changes what you use is motion. I am not sure the line holds. I have the Replit receipt.

The moment I keep coming back to landed before the harnesses and the agents. Santosh asked whether it was strange that we had circled back to what our parents were teaching us at eight, nine, ten years old. Just be a good dude. I said being a good person is an advantage, and I meant it as a claim about competitive dynamics, not a sentiment.

Then we spent the back half on harnesses and agents and skills files, and I closed by saying the individual contributor is dead forever. I wrote a version of that in Everyone (and Everything) Is a Manager Now and still think it is right. But we settled the human part first and spent the rest on tools. Either that is the joke, or it is the point.

Listen here: https://carryonpodcast.com/podcast/ai-wont-get-you-access-the-human-skills-that-still-matter-most/

With gratitude,

earn

In venture, when a thesis starts to play out, you don’t write a new memo. You make a follow-on investment. Consider this a follow-on field note.

Two months ago I wrote that the refinery was arriving: that data had spent twenty years as unrefined crude because nobody had built the pipelines, the tankers, or a market price for the barrel. This week the barrel got a price. Google just won the bankruptcy auction for Spirit Airlines’ enterprise data, agreeing to pay $10 million and outbidding Mercor at $7.5 million. Roughly 100 million emails, 500 million Teams messages, the code, the workflows. Decades of operational exhaust from a company once worth $6 billion, sold off right alongside the gate slots.

Corporate development keeps finding new things to buy. First it was buy the product for synergies. Next, it was acqui-hires: buy the team, shut down the product. Then firms got acquired for their Rolodex, if you remember what a Rolodex is. Now companies are being bought for their data. The through line is that what an acquirer actually wants is rarely what's on the balance sheet.

If the auction feels like a one-off, the numbers say otherwise. Zach Bratun-Glennon at Gradient estimates the labs are spending $10 to 15 billion a year on expert-generated training data, $1.5 million a day at a single vendor, in an industry that did not exist 24 months ago. Gradient’s own view is that the headline run rates understate the category by three to five times. Zach’s sharpest point has stuck with me: the gap between models is being purchased, not trained. Compute you can rent. Techniques get published. Data bought exclusively compounds inside one org and never diffuses.

The supply side is organizing just as fast. I caught up with an investor friend today whose firm backed a startup that winds down companies. The original business was closing companies cleanly. The new product is acquiring the digital footprint of every company they close, scrubbing the PII (Personally Identifiable Information), and selling it to the foundation labs. Demand is outrunning their hiring, and word is the labs want to go straight to the source. Because a dead company is not dead data. Spirit did not collapse because its data was bad. That archive is a complete record of how an airline actually ran.

How long does this window stay open? I honestly don’t know. But while it is, there is real opportunity for three groups: the AI data-training companies brokering and refining the supply, the wind-down and liquidation startups sitting closest to it, and private equity, which is holding distressed assets with a line item nobody ever priced. And because these are data deals, not equity deals, you don’t need a fund or a license to be in the flow. Brokers, scouts, and service shops eat here too.

There is no playbook yet. Nobody agrees on how to value a company’s data, how to diligence it, how to clean it, how to officially onboard it, or even who owns it. Data rights don’t show up in a term sheet. That is exactly what the early innings of an asset class look like: no comps, no standards, every deal cut from scratch. The refinery arrived in June. The traders showed up in August.

With gratitude,
earn

Kelli Fontaine worked in the athletic media relations department at the University of Colorado. She described what you do after a bad football game. You go find the stat that is true. “We could say we had the longest punt in history.” The number is real. It is also a record of a loss.

That is the whole conversation, sitting there in the first three minutes.

Kelli is a partner at Cendana Capital and may be the most data-forward allocator we have had on. Her team runs 78 dashboards. What I did not expect was how carefully she keeps her own numbers from hardening into rules.

The conversation sharpened something I had been circling. In early-stage venture, a finding is not a law. It is a timestamp. A good analysis has a shelf life the analyst does not control.

Three things I took.

First, she dates her own findings instead of defending them. When Kelli joined Cendana she built the data systems, then ran the analysis on their own book. Pre-seed against seed. Same graduation rate. Same mortality, which she thinks was probably acqui-hires. Better multiple on the year of entry. More risk, better outcome.

In our prep call she had put this to me as having grown skeptical of that conclusion, and I read that framing back to her on air. She did not repeat the word, and she did not defend the finding as durable either. She placed it: “it’s one of the things that worked in a market at that point in time.” Then she gave the mechanism that closed it. San Francisco filled with capital focused on early stage. Institutionalized angels became a category. Pre-seed, in her words, “probably arbitrage away in the Bay Area specifically.”

That is more useful than recanting. She keeps the analysis and puts a date on it. “That was a learning for me,” she said, and the learning was not that she had been wrong. It was about why a thing worked, and what that means now.

Second, a harder market argues for more shots, not fewer. I asked how the change in graduation rates shows up in diligence on a Fund I or Fund II, and said the part that seemed obvious: it is tougher now, a manager has to get so many things right to land even one or two fund returners.

She agreed, then went somewhere I did not expect. The instinct she keeps seeing in response is concentration, and she gets pitched a lot of 15 company portfolios. Her read runs the other way: “if graduation rates are lower, you know that would mean that you should probably have a few more shots on goal today, because things can look good at the early stages.”

Run the mechanics. Concentration only pays if your ability to identify the winner improved. A lower graduation rate does not improve identification. It lowers the base rate. Same shooting percentage, fewer makes, so you need more attempts. Her ask of a concentrated manager is not that they drop the structure. It is that they articulate why it is right, and why a prior hit rate repeats.

Third, acqui-hires are a return line. Looking back at Cendana’s historical returns, roughly 10% came from acqui-hires and roughly 10% from modest exits. Same contribution. I have spent years hearing acqui-hires described as what you settle for when the real outcome did not happen. In her math they pay about what a decent exit pays. She paired it with a colder point: it is very rare that a company comes back to life after year seven. Venture built enormous muscle around sourcing, picking, supporting, and winning. Portfolio management, she said, “has been missed along the way.”

Where I am still working it out is founder secondaries. I brought the alignment framing into the room, that founder selling should track GP selling. She came at it differently, and hers is better. Venture runs on hospitals, endowments, foundations, and pensions, so a founder taking cash ahead of those LPs is a fact about sequence, not only optics. Then read it as information. If a founder is de-risking, she said, “almost taking that as a signal is it’s time for you to de-risk some too.”

I prefer her reading to mine, and I am still not sure it resolves cleanly. She drew the line herself: is it enough “to buy a house and like give them some breathing room, or is it enough to make them extremely wealthy?” That line gets drawn by circumstance as much as by conviction. I have watched founders with no cushion make choices that looked like doubt and were just rent. I did not push on it in the room. I would want more resolution before treating sale size as a read on belief.

The phrase I keep sitting with is what she is not underwriting: “the king making marks.” I wrote about a version of this in The Motion Problem, that when capital concentrates hard enough, kingmaking starts doing the work judgment used to do. Kelli is running the LP-side version of the same worry, with dashboards. She would rather know the customer quality and the ACVs than know who led.

That is the data discipline I trust. Not the dashboard that answers. The dashboard that keeps asking whether last year’s answer still holds.

Listen here: https://swimmingwithallocators.com/podcast/finding-alpha-before-consensus-data-judgment-and-early-stage-venture/

With gratitude,

earn

A few weeks ago in San Francisco, over Korean fried chicken and a glass of non-alcoholic beer, a friend told me what her company’s customers say about the startups nipping at their heels. Not to the startups. To her.

“Don’t worry,” they tell her. “You build. Let us play around with these new startups for a year and tell you what’s working, what’s not. Once you’ve built the competing product, we’ll come back to you guys.”

The customer isn’t threatening to leave. It’s offering to run free competitive research and hand-deliver the results, because switching was never really on the table. They’ll pilot the scrappy new thing, learn what it does well, wait for my friend’s company to build the equivalent, and quietly come home. The startup thinks it’s winning a proof-of-concept. It’s actually running an unpaid discovery sprint for someone else’s product roadmap.

I’ve known this friend a long time and actually invested in her last startup. She built a company from nothing, self-serve, the kind where growth comes from a credit card and not a handshake. A fast-growing tech company bought her startup last year, and now she leads product there. Which means she has watched this entire game from every seat at the table: the founder who has to win the business, and the growth incumbent who barely has to try. That’s a rare vantage point, and everything she told me over lunch has shaped how I view the opportunity set for founders and venture capitalists alike.

Here’s the sentence that started it. I asked her what founders need right now, more than at any point since she started her career. She didn’t hesitate. “Building is easy,” she said. “It’s the only thing that got easy. Distribution is the only thing that still matters.” Scoping and shipping a real product used to take her about two weeks, she told me. Now it takes about two days. I won’t pretend that number is peer-reviewed; it’s her felt sense of her own working life, but I believe her, because I’ve felt the same compression in mine. Take her word for it for a second, because the implication is the whole essay. When production speeds up seven times over, the bottleneck doesn’t disappear. It moves. Right now, it’s moved entirely into the channel.

AI collapsed the cost of building software. It did not collapse the cost of being believed. Distribution, which is really just trust plus a channel plus a signature on a renewal, is the last scarce thing left, and it is repricing accordingly.

My friend’s CEO is proof of what that kind of trust actually buys you. She says he can put a name and a face to nearly every account the company has. He flies out to see them, not once at signing and never again, but every year, on purpose, the way you’d visit family and old friends. When she panicked about being late to market with a new feature, watching an emergent competitor already shipping it, he told her not to worry. He had his customers. Her job was to build the right product at the right time. His was to keep the company’s word and sell whatever she built, whenever it was ready, because the account was never really up for grabs to begin with.

And it’s working, she told me. Crazy how well.

That’s the real prize of distribution, and it isn’t just winning more deals. It’s buying time. A company with a genuine distribution moat gets to move slower on purpose (like I said about Luka’s deceleration), in a market that punishes everyone else for moving slowly by accident. That’s also, not coincidentally, exactly what customers keep asking for: fewer vendors, not more, one place for their data to live. When her company ships something new, it doesn’t have to win a bake-off against three well-funded startups. It gets folded into an existing contract as what she called an incremental ten percent spend. A new product isn’t a new sale. It’s an upsell dressed up as a line item, on a relationship that was never actually in question. Customers will even hand over a rival’s pricing unprompted, the same way you’d mention what a mutual friend makes. All the secret stuff, she called it. When the relationship is strong enough, competitive intelligence just arrives.

She sees the sharpest version of this with the founders she advises. Everybody gets a proof of concept fast these days, because everyone wants to play with the new thing. Converting that pilot into a real, multi-year contract is the hard part, and by her account it’s getting harder. Back when she was building her own company, she remembers enterprise pilots converting something like seventy percent of the time. That’s memory, not data, but it lines up with everything else she described. Now, almost nobody wants to sign anything longer than a year, because the fear in the room has changed shape. It isn’t “will this product work.” It’s “what happens when the platform we already pay just builds this feature themselves.” She told me about a startup she watched get most of the way through a term sheet for a security product, only to feel the whole conversation cool the moment an LLM announced something similar. The deal didn’t die because the product got worse. It died because the buyer got scared of being embarrassed for picking the smaller, riskier option. Her read: either you’re first into the account, or you already have real distribution power. There isn’t much room left in between.

Then she told me something that clearly cost her a little to say out loud, from the winning side of an acquisition.

Her old company never built any of this. It was self-serve from day one, put in a credit card, and that was the whole relationship. They had one person doing anything resembling sales. Never went top-down into an enterprise, never took a champion out for a beer, never learned a single budget holder’s name. For a while that wasn’t a weakness. It was the whole appeal. Bottoms-up growth let a generation of builders skip the relationship layer entirely and still win. If her old company still existed today, she said, it would be in real trouble. Not because the product got worse. Because it never had a relationship with anyone who actually controls a budget, and in the world as it exists right now, almost every serious purchase decision gets made top-down. The self-serve era let you route around the relationship. This one built a toll booth right back into the middle of the road.

A founder autopsy, delivered by the founder herself, from the other side of the table. That’s not a comfortable thing to say out loud, and she said it anyway.

I couldn’t stop thinking about The Wire.

Read more

Bands almost never break up over the music.

Ask anyone who has been in one. It does not end because the songs got worse or because somebody stopped being able to play. It ends over the credit. Who wrote the bridge. Whose name goes on the record. Who gets paid when a car commercial licenses the chorus eleven years later. Nobody has that fight in year one. In year one there is a van and a shared amp and nothing to split, so the question never comes up, and everyone tells themselves the silence is alignment. It is not alignment. It is an unpriced asset. The fight arrives on schedule the moment there is something worth arguing about, and by then the arrangement between the people is the fragile part, not the music.

Near the end of episode three I put three archetypes on the table, because I think every firm is one of them whether or not it has admitted it.

Some firms are the Rolling Stones. This is the band. We are doing this into our seventies, or until we quit. We ain’t adding nobody new. Some are small market franchises. Utah, Oklahoma City. We are not getting top free agents, so we had better be excellent at the draft. And some are the Lakers. We do not care about draft picks. Once somebody proves they are amazing, our brand is better than theirs, and they will come. We are watching that one happen right now, with people folding firms they started themselves to go join a bigger name.

The taxonomy sorts firms by how they hold together and how they replace people, which is more useful than sorting them by AUM or stage. And the band is the one most people are actually building. It is also the only one of the three with no plan for its own failure. The franchise has a draft. The brand has a recruiting machine. The band has a handshake.

Most of the episode is about the least glamorous part of this job, which is sourcing. Santosh’s framing is that the shift already happened and a lot of firms have not noticed: you stop tracking companies and you start tracking people, sometimes years before they come to the realization that they want to go found something. At pre-seed and seed there is no database and no portfolio page to scrape. There is only whether they know your mettle before they need you. That is a different job than the one most of us were trained to do, and it does not produce a pipeline you can screenshot for an LP update.

At my stage the problem inverts. When I started in 2015, sourcing at Series A still meant something, because there were fewer firms and much less information moving between them. Now everything ends up in a database, so I am not sure proprietary deal flow at traction stage is a real thing anymore. Anybody can find it. Can you win it. Can you show a founder why you are the right partner, and then actually help them inflect.

Santosh’s answer for how you earn that is plainer than it sounds. Make a promise, know in advance which ones you can keep, and then keep them. Founders run the list afterward, and so does everyone watching. Which is really a conversation about brand. We learned it at Kauffman and I have never found a better version: your brand is what people say when you are not around.

The middle stretch of the episode is solo GP versus partnership, and the way that question usually gets asked is about economics and speed. The way it should get asked is about self-knowledge. Do you know how you behave when you are wrong in front of someone whose opinion you need. A partnership is not a structure for making decisions. It is a structure for surviving disagreements, and Santosh is right that the ones without a real conflict resolution process do not blow up. They poison slowly. Resentment compounds while everyone stays professional about it. In a firm the unpriced asset is carry splits, decision rights, and whose name is on the deal when it works. That is the songwriting credit again.

Do not run a partnership like a dentist’s office. One admin at the front desk, three dentists, one building, nobody talking to each other. That is not a firm, it is a lease.

Sam Heshmati from Citizens Private Bank sits down partway through, and his stretch is the best articulation I have heard of why problem solving beats product pushing. He says their job is not to be the reason a company succeeds, it is to slightly increase the probability of success, and to do that for a lot of people. Products and services get commoditized. People and relationships do not. He is describing banking, and he is describing our job.

There is more in the forty-four minutes than I can carry into one essay. We get into why the easy version of your strategy is usually just a different hard thing, and what an associate is even for when diligence takes ninety minutes.

So here is the ask, same as the last two times, because it keeps working. Listen to the episode. Then tell us where we got it right and, more usefully, where we did not. Reply to this post, comment wherever you listen, or send me a note directly. If there is a question about this craft you think two GPs owe you a straight answer on, that is how episodes get made.

Write the songs. Then write down who wrote them.

Grateful to AngelList, WilmerHale, KPMG, and Citizens Private Bank for backing this from the start, and to the Heard Media team for turning two people talking into something worth your time.

with gratitude,
Earnest

Listen on your favorite platform.

• Full episode page

• Apple

• Spotify

Columbia sent me a book before I ever set foot on campus. Homer’s Iliad, in the mail, over the summer. I remember thinking, with real optimism, that this was how it was going to work from here on out. They were just going to send us our books.

They did not send us our books. That was the only one.

But you were supposed to read it before you arrived, because the first day of Literature Humanities (Lit Hum IYN) was not a syllabus review. It was a discussion.

I walked in ready. I had done the reading. I was in college now, I was taking this seriously, and I had the distinct sense that the whole thing was going to be a breeze because I was putting in more effort. The professor, a grad student, asked a question. I do not remember the question. I remember my hand going up high and straight, the way it does when you want to be picked. He pointed at me. I gave my answer.

He said, “No.”

Then he moved to the next person.

I did not speak in that class again for two or three weeks.

What that “no” installed in me was a rule I did not choose and could not see. I have to be a thousand percent sure I am right before I open my mouth.

Read more

The last time Chris Schelling was on the show, the episode carried a title I still think about: Private Detective to Private Investments. He is back, and the seat has changed. Chris recently made the move to Aksia as a Managing Director, which hands him the kind of cross-asset vantage point over private markets that most of us only get to guess at. Alexa and I took full advantage.

What struck me this time is how much of the AI conversation happened on the allocator’s side of the table. Everyone asks how AI changes the companies we underwrite. Chris is just as interested in how it changes the underwriting. He gets into using AI for diligence and memos, training analysts with it as a kind of behavioral coach, and what it would actually mean to seat an AI agent at an investment committee. That last one sounds like a bit until you hear him walk through it.

On the bubble question, Chris threads it the way only someone who has lived a few cycles can: “You can have a bubble without true technological disruption occurring.” Both things get to be true at once. The technology is real. The prices can still be wrong. And he applies the same discipline to the fashionable short thesis of the moment: “It’s not legacy SaaS goes boom. Like that’s way too simplistic.”

Then there is the section I suspect people will replay. Chris walks through SPV sandwiches, layers of access stacked on layers of fees, and why the structure rhymes uncomfortably with 2008. He even sketches the start of a checklist for evaluating SPVs and the access claims that come with them. I will not spoil it here. If you have wired money into an SPV in the last couple of years, that stretch alone is the episode. Nick Cassin, who heads Sidley’s secondaries practice, joins the episode as well, and the back half moves through how the mega RIAs are building private markets platforms and what product diversity across the wealth channel actually requires.

The line I keep coming back to is the one Chris lands near the end: “In a world of AI tools, differentiated relationships, networks, and deep domain expertise matter more than ever.” If you read The Delusion Requirement last week, you can guess why that one stayed with me. The tools are getting evenly distributed. The human parts are not.

If the show has been useful to you, the best thing you can do is leave a rating and review wherever you listen. It matters more than most people realize.

with gratitude,
earnest

I’ve had one improv class (maybe two) in my life, usually at a Kauffman Fellows module. There is a rule they teach you in the first week of improv, before you are allowed to try to be funny. Two people walk onto a bare stage with nothing. No script, no set, no idea where any of it is going. The first person makes an offer, and it is usually a bad one. The scene does not survive on the quality of the offer. It survives on whether the second person accepts it and adds something. Yes, and. I’m sure you have heard of it. The moment somebody blocks, corrects, or waits for a better scene to arrive, the whole thing dies where it stands.

I have been thinking about that rule since Santosh Sankar and I sat down to record the first three episodes of Carry On. Santosh opened with the math. Over the last couple of years, something like 75 percent of everything LPs put into venture went to the top thirty funds. Everyone else, and that is not just emerging managers, that is everyone else, split the quarter that was left. Then he asked the question the math begs for: why would anyone start a venture fund right now?

My answer, on the record, was that you have to be delusional. Not kind of delusional. Definitely delusional. And I meant it as a compliment, because delusion is the exact trait we screen for in founders. Someone who can see every problem in front of them, assess all of that information honestly, and still expect to be the exception. We ask founders to carry that contradiction every single day. Starting a venture firm is just finally taking your own medicine.

Santosh has a story about this. A few years into building Dynamo, his wife looked over at him and asked, casually, so you think this is going to work? Then she answered it herself: what a dumb question. Of course he believed it was going to work. Why would he be doing it otherwise? Your self belief is a different thing from the actuality of what happens. Holding both of those at once is the job.

But delusion alone is not a strategy. It is noise with confidence. The thing that has to sit underneath it is a why, and this is where the episode got uncomfortable in the useful way. An allocator friend of mine recently passed along what another allocator told her: she needs one more venture firm the way she needs another bullet in the head. That is the room every new manager is walking into. Your why is not marketing copy for that room. It is your right to exist in it.

Differentiation is the same conversation wearing a suit. It is the word we all use and never define, and as this industry matured it collapsed into a sterile monoculture: identical websites, identical value add, identical posts, everyone quietly afraid that the tallest flower gets cut. Real differentiation is not a slide. It is the compounding of all the small ways you approach the craft differently. A 24 hour response rule. How you actually run diligence. What you still do for a founder in month nine that you promised in the first meeting. Most of us are some shade of blue. The work is knowing exactly which shade you are and saying it out loud.

We also get into the part I keep having to relearn: none of this works if you wait to be perfect. Somebody prominent in this business, I heard it secondhand so I will not name them, said the optimal level of cringe is not zero. The best firms right now, as Santosh put it, operate like media companies with investment capabilities attached, which means the reps are public, and some of the reps are bad, and the bad reps are where the feedback lives. That is improv again. You do not get to workshop the scene in private first.

There is more in the episode than I can carry into one essay. Sam Heshmati of Citizens Private Bank joins in the middle and traces the white glove lineage from First Republic, which is itself a case study in a why surviving a change of jersey.

So here is the ask, same as last time, because it worked. Listen to the episode. Then tell us where we got it right and, more usefully, where we did not. Reply to this post, comment wherever you listen, or send me a note directly. If there is a question about this craft you think two GPs owe you a straight answer on, that is how episodes get made.

The cards are stacked. The scene starts anyway. Yes, and.

Grateful to AngelList, WilmerHale, KPMG, and Citizens Private Bank for backing this from the start, and to the Heard Media team for turning two people talking into something worth your time.

with gratitude,
Earnest

Listen on your favorite platform.

• YouTube

• Apple

• Spotify

• Full episode page

A few weeks ago I got on a call with a founder friend of mine who spends most of his waking hours actually building things. He is a universe of technical depth ahead of me, and the whole reason I asked for his time was that my agent setup, which I had cobbled together over the last several quarters, had started to feel clunky in a way I could not diagnose. Too many gears, not enough grip.

About fifteen minutes in, he stopped me and said this:

Friend: So question. Because it’s going to help me identify, like, how I talk to you about this. How technical are you on, like, elements and agents?

Me: I’m just like a self-taught. I’m not that technical. I actually don’t.

Friend: Do you know what a sub-agent is? Do you know what an agent harness is?

Me: No.

Friend: Do you know what a system prompt is?

Me: A system prompt?

Friend: Yeah.

Me: No. What is that? I’m awful. I’m not technical at all.

Friend: Yeah, I know. Yeah, I’m just trying to gauge, like, what terms I can use and not. What about, like, a context window? Tokens?

Me: Yes.

Friend: Okay, damn. So, okay, yeah. So, like, if it’s not in a headline somewhere, you don’t know what it means.

I laughed. He was right. If it was not in an article or a post I was already reading, I did not know it. I am not out here reading technical documents or research papers like Chris Dixon, though I do read the Substack called Technically, and the only reason I made it to that call at all was that I was willing to say so out loud to someone with the vocabulary and the receipts to explain, with a lot of grace, that I had been running my entire operation the wrong way.

I want to talk about what he showed me. But more than that, I want to talk about what it implies for the future of labor. Because I think the individual contributor as a category of worker is about to become endangered, and the people who do not see it coming are going to be the same ones who have spent the last cycle telling themselves they were just a little behind on the technology and would catch up once the disruption settled. They will not catch up. The tools are not the thing. The builder mentality coupled with the (in)organic chart is the thing.

Where I was stuck

I run a small operation with a lot of surface area. I am building a Fund. Two podcasts. A newsletter. Convenings. Consulting. It looks like more than it is because I built agent scaffolding underneath the whole thing to keep action items from falling through the cracks. Every meeting drops a transcript, every transcript gets scraped for to-dos, every to-do gets assigned or drafted or flagged. It runs while I sleep. And when it works, it works well enough that I forget it is running at all.

But it had started to break in the small ways that always add up. Duplicate drafts. Missed follow-ups. Two agents chasing the same task in two different tabs, neither aware of the other. Sand in the gears.

My instinct was to slice the tasks thinner, so I did. One agent for to-dos, one for the partner meeting, one for reengagement, one for intros, each of them polite and focused and operating with a sliver of the picture. Which sounded like discipline, and was actually the opposite. I was treating every agent like an employee I could dispatch, and I was staying in the loop between all of them, checking their work, nudging their progress, rerouting their outputs, playing the role of switchboard operator between a bunch of well-meaning digital ICs who could not see each other.

I was the manager. I had built myself a bullpen of ICs. And I had made myself the bottleneck.

What he actually drew

After I walked him through my misjointed workflows, he stopped me and drew what I should be doing instead. He drew it for me on a whiteboard app while we talked, and I have the screenshots. I keep looking at them because they are the simplest version of a thing I had been failing to see.

Read more

There is a moment early in this conversation when Senia Rapisarda describes how she spent years educating Canadian pension plans on what venture actually is. Her line stuck with me. Venture is agriculture, not geology. A bigger drill does not help you find a nugget that was never buried there. You plant kernels at the university level, at the incubator level, and then you provide a continuum of capital and wait. Twelve years later, Canada has managers on fund two, three, and four, and companies that stay home instead of heading south.

Senia joined Alexa and me from Toronto, where she leads HarbourVest’s Canadian strategy. Before that she was at BDC, where she helped design the Venture Capital Action Plan, a program built with one party and adopted by the next. In a moment this polarized, a policy that survives a change in government might be the strongest signal of all that innovation capital has become infrastructure.

That was the argument that landed hardest for me. Innovation capital is sovereignty. We tend to think of sovereignty as borders and defense budgets. Senia’s frame is that a country that cannot fund its own champions ends up de-risking companies at home and exporting the returns abroad. By her count, roughly 90 percent of late-stage Canadian tech rounds have been led by international managers. The talent is Canadian. The upside mostly is not. Breaking that cycle is the work.

The part of the conversation I keep replaying is her guidance for emerging managers. HarbourVest runs an open door policy, but the door is a starting line, not a finish line. Come back every six months with a fifteen minute update. Bring a co-investment before you ask for a fund commitment. She jokes that a fund commitment lasts longer than the average North American marriage, so date with a deal first. And know that everything gets logged. HarbourVest tracks what she calls your say/do ratio. Tell them what you will do, then do it, and you are building a track record before they ever wire a dollar. That framing feels right to me. Trust is not claimed in a pitch meeting. It accrues in the gap between what you said last time and what you did since.

She also gave the most honest answer I have heard to a question I have been asking a lot lately. She has four filters for durable companies: mission critical, big moat, capital efficient, and non-greedy founders. I asked her to pick the one that predicts survival through this cycle. Her answer was that it depends on your cash position. If you have cash, the moat matters most. If you do not, founder greed at the last raise is what kills you, because a too-high valuation destroys value in the round you cannot avoid.

Nick Cassin of Sidley closes the episode with a clear-eyed explainer on continuation vehicles, including why disclosure and equal information are the whole ballgame when a GP sits on both sides of the table.

Worth a listen.

With gratitude,
Earn

There is a version of a basketball player who never stops moving and never actually gets anywhere. Always cutting, always talking, always going east and west, and somehow the ball ends the possession in the exact same spot it started. Coaches have a phrase for this, and it is not a compliment: motion is not the same thing as movement. Most people confuse the two anyway.

I have been circling that same confusion in venture for a while now. Plenty of activity. Term sheets, panels, LinkedIn posts, another fund closing, another logo added to a deck. Not always much progress underneath it. I did not have a clean way to talk about that difference until Santosh Sankar and I sat down to record, and it turned out to be the first real conversation of the show we are calling Carry On.

Santosh runs Dynamo Ventures. I have known him long enough to trust that he will tell me when an idea is not as sharp as I think it is, which is most of what you want from a co-host and most of what is missing from venture capital content generally. We did not start this because the podcast market needed another one. We started it because the honest conversations about what it actually takes to build a firm mostly happen off the record, in DMs and side rooms, after the panel ends. We wanted to put more of that on the record.

Episode one is called Motion ≠ Movement. We get into why urgency and momentum get treated as the same thing when they are not, and what it looks like when capital and attention concentrate into a narrow set of AI deals and mega funds. We also spend real time on something less comfortable: the erosion of intellectual honesty in this industry, and what it costs a firm when kingmaking dynamics start doing the work that judgment used to do. Santosh walks through the eras of venture, the pendulum between concentration and unbundling, and why craft still matters even when capital alone can buy you a seat at the table.

None of that is a finished argument. Some of it we are still working out in real time, which you can probably hear. That is the point. We would rather practice this in public and be wrong on the record occasionally than polish something into a highlight reel that does not tell you anything you could not have guessed.

So here is a genuine ask: listen to the episode. Tell us where we got it right and, more usefully, where we did not. Reply to this post, comment wherever you are listening, or just send me a note directly. If there is a topic you think two GPs owe each other an honest conversation about, tell me that too. We are building the list of episodes off of exactly that kind of feedback.

Grateful to AngelList, WilmerHale, and Citizens Bank for backing this from the start, and to the Heard Media team for turning two people talking into something worth your time.

with gratitude,
Earnest

Listen on your favorite platform.

Youtube:
Apple
Spotify

Last week on Swimming with Allocators, we welcomed Teddy Repko of the Yuhaaviatam of San Manuel Nation, who shares how his global upbringing, risk-aware personality, and early trading career shaped his investing approach, and how a shift at Columbia University’s endowment from hedge funds to relationship-driven venture gave him a front-row seat to the asset class.

During the conversation he also breaks down how LPs really underwrite managers, balancing base rates, power-law math, and portfolio construction with deep people and relationship assessment, and flags both opportunities and risks in today’s market, from mega funds, AI-driven valuations, and rapid funding cycles to concentrated DPI, late-stage growth, and crypto.

A great episode that I hope you all enjoy.

In the summer of 2011, the Miami Heat had a welcome party for their three free agent signees, Dwyane Wade, Chris Bosh and of course LeBron James. James had recently made “the decision” to leave his original team, the Cleveland Cavaliers, for a chance to win the championship in South Beach. The clip that is laughably used to mock this moment in NBA history is when LeBron responds to the question, ‘how many titles can this team win?’ He responded, “not 1, not 2, not 3, not 4…”, eventually tailing off around 8.

They won 2 titles. Which is a hard achievement. But when your expectation is so high, 2 titles is “not 5, not 6, …” and feels like a failure. Even at the time it felt like the Miami Heat players and fans were literally celebrating the offseason like it was the championship parade.

And for a while now something similar has been going on in venture capital. Some misplaced sense of achievement. This has been sitting with me after a stretch of conversations with founders, pre-seed and seed investors, and a steady diet of consolidation headlines.

The fear is not about losing. The fear is about ‘winning’ the wrong way.

There is a moment in early-stage venture that looks like a breakthrough and is actually a trap. A marquee firm comes downstream, sees something they like, and writes a check that skips two stages of price discovery. The round closes at a number that requires everything to go right. The founder celebrates. The seed investors celebrate. And somewhere underneath all of that, quietly, the outcome distribution narrows.

That’s the kingmaking problem nobody talks about loudly. The brand-name firms have enough momentum capital and enough platform pull to set a valuation that the market has to treat as real. Which means the founder is now playing a different game than they signed up for. And the early investors, who did the hard work of finding them before anyone else was paying attention, are now along for a ride they didn’t choose.

A well-capitalized company raises at a number that requires generational scale to justify, stumbles on product-market fit, and then has to pivot the entire thesis just to keep the lights on. Meanwhile, a scrappier competitor, operating with less runway and more discipline, is quietly winning on the ground. The first company raised 10 to 20 times more and still lost.

The math on this is not complicated, but the psychology of it is hard. More capital feels like more options. It buys time, it buys talent, it buys distribution. But it also buys obligation. The moment you take money at a valuation you have not earned, you stop being a company navigating toward product-market fit and start being a company defending a number. Those are very different jobs. The first one leaves room to learn. The second one does not.

What I keep coming back to is how much of this is a timing problem. The capital is not the mistake. The sequencing is. Raise too much, too early, before you know what is actually working, and you lose the middle. The good-not-great outcomes, the ones that still build real companies and return real capital, mostly disappear. You are either returning the fund or you are a cautionary tale.

The investors I keep learning from are not asking “how do we get to the Series A?” They are asking “what does the Series A make possible, and what does it foreclose?” That orientation tends to produce founders who find fit before they find the big check.

The consolidation headlines keep coming. The mega-rounds keep landing at pre-revenue multiples. The pressure to participate is real on every side of the table. But the companies that feel durable right now are not the ones that raised the most confidence. They are the ones that earned it. They won the right way before anyone was watching.

That’s the only way the foundation holds.

with gratitude,
E

For two decades, we’ve heard that data is the new oil. It was a clean line, the kind that makes its way into every keynote, every board deck, and every LP deck, and like most punchy lines, it skipped a step. Oil is not valuable because it exists. Oil is valuable because somebody built refineries, pipelines, and tankers, and established a global price for the barrel. Until that infrastructure showed up, oil was a thick, corrosive sludge that ruined boots and poisoned the wells around it.

Data, for most of the last two decades, has been unrefined crude oil. Companies sat on enormous reserves of it and treated it like exhaust from the real business. Logs, tickets, transcripts, internal Slack threads, edit histories, the back-and-forth between a customer success rep and the engineer she escalated to at 11pm. Nobody was buying those interactions. These activities and data points were just the price of doing business. The refinery did not exist.

Read more

Kate Simpson started her career as a history major. No finance background, no roadmap. The UNC endowment took a chance on her anyway, and she spent the first stretch of her career doing exactly what a non-finance person does in that seat: listening, asking questions, and building mental models from scratch. She took intro accounting classes at night to supplement what she was learning during the day. She went deeper inside private assets because that was where her team spent most of its energy. And she kept going, from Parish Capital to TrueBridge, where she spent twelve years watching the venture industry scale and evolve in real time, to now leading the venture strategy at GEM, a multi-asset OCIO firm that has been around since 2007.

What struck me about Kate’s path is that it never looked like the fastest way to the top. It looked like someone trying to genuinely understand the craft before assuming they had mastered it. That patience, which is different from caution, runs through how she thinks about everything from building a new venture program to evaluating an emerging manager for the first time.

There is a concept that came up early in our conversation that I want to dwell on before we get to the barbell. Kate described what she looks for in any venture manager as three things: how they source, how they pick, and how they win. The sequence matters. You cannot pick or win consistently unless you are sourcing in the right places. That means the first thing she wants to walk away from an initial meeting understanding is the manager’s network, specifically what ponds they are fishing in and whether those networks carry a real edge.

That is a different question than most GPs think they are being asked. Most people walk into an LP meeting ready to talk about portfolio performance or investment thesis. Kate is asking something that comes earlier: where are you finding things that other people are not finding yet? The answer to that question is more predictive than almost anything else.

The barbell idea is where GEM’s strategy gets interesting. Kate is clear that access to scaled platforms still matters, but argues that the alpha in today’s market is increasingly at the edges. On one end, a handful of top-tier, established platforms retain real competitive advantage. On the other end, the small, craft-driven funds that most institutional programs cannot or do not prioritize are generating the kind of right-tail skew that makes the math work. The middle, the firms that are too big to win at the seed stage but too small to compete with multi-billion-dollar platforms at Series A, is where she is spending proportionally less time.

The math point is not abstract. GEM builds what Kate calls a “what you need to believe” model for every manager who advances in their pipeline. It overlays fund size against target ownership, number of positions, reserve strategy, and a practical range of outcomes to pressure-test whether it is reasonable to expect a 3x, a 5x, or better. One strong outcome returning the fund, not necessarily a unicorn, is the baseline they are testing for. The point is not to find reasons to say no. The point is to know, specifically, what has to be true.

The sub-$200M threshold GEM uses for their dedicated seed and micro fund vehicle is a version of that same discipline applied at the portfolio level. There is something they are protecting when they draw that line, and it is not just vintage diversification. It is the recognition that the fund math on a small, concentrated, high-ownership fund looks different from the math on anything larger, and that difference is worth constructing around deliberately.

Kate also talked about something I have been thinking about more lately, which is what this AI cycle actually looks like from an LP seat. Her read is that we are in early innings of a genuine paradigm shift, comparable to the move to mobile or the move to cloud but potentially larger in terms of the size of outcomes being created. Companies staying private longer has helped the secondary market grow into a real asset class rather than a niche workaround. That normalization changes what liquidity looks like for early investors and founders in ways that would have been unimaginable a decade ago.

Nick Cassin from Sidley also joined us to break down the secondary market side of this. The numbers alone tell the story: from roughly $20 billion in annual transaction volume when he started in 2010 to potentially north of $250 billion today. Continuation vehicles, once a niche exit mechanism, have become a mainstream strategy. The pool of buyers has expanded dramatically. That is not just a market structure observation. It changes how GPs should think about what they are building and how they are building it.

The through line I keep coming back to from this conversation is that the best managers Kate has worked with over her career share one characteristic above most others: they know what they are good at and they stay focused on it. Not because they lack ambition, but because discipline around the edges of your own competence is what lets you build a durable brand and a repeatable process. That is true for fund managers and, honestly, for anyone trying to build something that compounds over time.

If the show has been useful to you, the best thing you can do is leave a rating and review wherever you listen. It matters more than most people realize.

with gratitude,
earnest