💥Fundraising Tips in a Consensus…

1. Key Themes

The Seed Market Is Barbelling: Winners Get Priced Like Growth Rounds, Everyone Else Falls Behind

Carta data shows an accelerating split between top-decile and median seed valuations, with VCs chasing a narrow set of AI-infra and frontier-tech deals regardless of ownership discipline.

"The 90th percentile of seed post-money valuations for US companies on Carta reached $99.8M in Q2 2026, up from $20.0M in Q1 2018, a roughly 5x increase over the period."
"The multiple between the 90th percentile and the median grew from 2.2x in Q1 2018 to 3.7x in Q2 2026, the widest spread in the dataset."
"Shad argues VCs are competing fiercely for the top 10% of early-stage deals tied to AI infrastructure, specialized hardware, and frontier tech, even where funds can't hit their normal ownership targets."

AI Is Commoditizing Intelligence — Value Is Migrating to Distribution and Workflow Ownership

a16z's framework argues that as model capability becomes cheap and interchangeable, defensibility shifts to whoever owns the customer relationship and workflow, not whoever has the best model.

"Silicon Data's LLM token price indices put the closed-model index at $3.07 versus $0.66 for open models; Acharya frames this gap as a routing signal, with real differentiation still sitting at the application layer."
"Network effects, distribution, brand, and scale still favor incumbents, Acharya argues, but the edge from linking software together is fading as agents learn to do it themselves."
"intelligence is becoming a cheap, commoditized input, so real value shifts to whichever company controls distribution, brand, and workflow adoption."

SaaS Is Being Replaced by Outcome-Ownership Models

Slow Ventures' Yoni Rechtman sees a shift from selling software tools to selling guaranteed outcomes — via forward-deployed services, AI-powered rollups, or agent/labor networks.

"The idea behind FDE (forward-deployed engineering) is selling the hammered nail: a company guarantees the finished business outcome itself, backed by hands-on service, and prices that guarantee directly."
"In this rollup model, a firm buys existing service businesses and uses AI to compress their costs; Slow's Metropolis used this approach to roll up parking and is now valued at $5B."
"Rechtman's dividing line across all five models is ownership: a company either does the work itself or it doesn't, and he argues the customer already knows which one they bought."

Fundraising in a Consensus Market Rewards Legibility, Not Just Metrics

With VCs increasingly reading signals off each other, founders' credibility is built through visible proof points (hires, obviousness of opportunity, discretion) well before the pitch.

"No revenue figure guarantees a Series A anymore... the real test is whether the opportunity looks obvious, and justifying its size to investors is already a bad sign."
"Claiming a term sheet from one fund often reaches competitors within a day in a small, connected market... so founders should assume anything they say will circulate."
"Few founders showcase recent hires in their pitch, per Kothari, even though strong hires help VCs underwrite the downside, since a struggling company can still be acquired for its talent."

2. Contrarian Perspectives

Fund Concentration ("Conviction Investing") Is Statistically Overrated

Against the popular GP narrative of doubling down on winners, the data shows even skilled managers can't reliably identify their own top performers, making concentration a bet that usually backfires.

"A Correlation Ventures census of about 21,000 US financings (2004-2014) puts the loss-or-flat rate at 65.3%, while just 0.53% of deals land at 50x or higher."
"After the deployment period for 20VC Fund I, Stebbings ranked his predicted top 5 portfolio companies; three years later, not one of them matched the fund's actual top performers."
"Fund managers who believe they can consistently pick winners tend to overconcentrate on favorites; a 2022 paper found this backfires more than it pays off, hurting weak funds more than it helps strong ones."

Justifying Market Size to Investors Is a Red Flag, Not a Strength

Conventional pitch advice emphasizes articulating TAM — but Kothari (via Rechtman) argues that needing to explain why an opportunity is big signals it isn't obviously big enough.

"The real test is whether the opportunity looks obvious, and justifying its size to investors is already a bad sign."

Higher Founder Pay Can Be Good for the Company, Not a Sign of Weak Commitment

Against the bootstrapping-martyrdom ethos, NFX argues underpaying founders creates a real risk of early departure before the outcome is realized.

"Underpaid founders tend to quit before the real outcome arrives."
"What a founder pays themselves also sets a soft ceiling on what they can justify paying senior hires."

3. Companies Identified

Carta — Cap table and valuation data platform. Mentioned as the data source for seed valuation trends. Quote: "Hamza Shad, Insights at Carta, shared new Carta data tracking the gap between the top and the middle of the US seed market."
a16z — Major VC firm. Mentioned via GP Anish Acharya's LP deck on where AI value accrues. Quote: "published the deck he presented to LPs covering the broad AI market, the application layer, and consumer products."
The Odin Times / Odin — Research/commentary outlet. Mentioned for its analysis of fund concentration math. Quote: "Dan Gray at The Odin Times examined portfolio construction in venture capital... using Pitchbook benchmarks and academic research."
20VC Fund I (Harry Stebbings) — VC fund. Used as a case study showing even top investors can't predict their own winners. Quote: "Stebbings ranked his predicted top 5 portfolio companies; three years later, not one of them matched the fund's actual top performers."
Correlation Ventures — Research source cited for base-rate venture return data. Quote: "A census of about 21,000 US financings (2004-2014) puts the loss-or-flat rate at 65.3%."
Slow Ventures — VC firm. Source of the five-business-models-replacing-SaaS framework. Mentioned throughout via partner Yoni Rechtman.
Metropolis — AI-powered parking rollup, portfolio company example. Cited as proof of the "AI rollup" business model. Quote: "Slow's Metropolis used this approach to roll up parking and is now valued at $5B."
Recurrence (Phoebe) — Caregiver-staffing agent network. Cited as an example of the "agent network" model. Quote: "like Recurrence's Phoebe filling caregiver shifts, then resells the worker network it built to new demand, an end state he calls a 'for-profit union.'"
NFX — VC firm. Source of the founder compensation framework via Omri Drory.
Granola — AI meeting-notes/copilot tool. Newsletter sponsor, cited as a product recommendation. Quote: "Granola transcribes directly from your computer or phone audio... no meeting bots."

4. People Identified

Hamza Shad — Insights lead at Carta. Why mentioned: authored the seed valuation data study. Quote: "shared new Carta data tracking the gap between the top and the middle of the US seed market."
Anish Acharya — GP at a16z. Why mentioned: authored the AI value-accrual framework for LPs. Quote: "Acharya splits AI progress into capability... and economic impact."
Dan Gray — Writer at The Odin Times. Why mentioned: analyzed fund concentration and portfolio construction research. Quote: "managers can't reliably predict winners... yet overconfidence pushes many to concentrate bets anyway."
Harry Stebbings — Founder of 20VC. Why mentioned: case study on inability to predict own fund's top performers. Quote: "not one of them matched the fund's actual top performers."
Luke Sophinos — Interviewer. Why mentioned: conducted the interview with Yoni Rechtman on post-SaaS business models.
Yoni Rechtman — Partner at Slow Ventures. Why mentioned: originated the five business models replacing SaaS. Quote: "a company either does the work itself or it doesn't, and he argues the customer already knows which one they bought."
Nikunj Kothari — Author/investor. Why mentioned: wrote the fundraising signals piece for a consensus market. Quote: "legibility gets built well before the pitch meeting itself."
Omri Drory — Partner at NFX. Why mentioned: authored the founder compensation framework. Quote: "founders should protect the sense that their success and the company's are the same."

5. Operating Insights

Showcase hires as a fundraising signal. Founders rarely highlight recent hires in pitches, but strong hires act as downside protection for investors — "a struggling company can still be acquired for its talent" — making this an underused, low-cost credibility lever.
Assume zero confidentiality in a small market. "Claiming a term sheet from one fund often reaches competitors within a day," so founders should treat all fundraising claims and conversations as public information that will circulate.
Revisit founder comp only at defined triggers, not continuously. NFX's four triggers — material life change, dilution to employee-level equity, reaching escape velocity, or a justified secondary — give operators a disciplined framework rather than ad hoc or emotionally-driven pay decisions.

6. Overlooked Insights

The closed-vs-open model pricing gap (4.6x) is being reframed as a routing/commoditization signal rather than a moat. This is a subtle but important reframe: "Silicon Data's LLM token price indices put the closed-model index at $3.07 versus $0.66 for open models" — suggesting sophisticated operators should architect for model-agnostic routing rather than lock-in to a single frontier lab.
A $100M seed valuation is being used as a leading indicator of down-round risk, not just a fundraising win — "a $100M seed valuation sets a demanding bar for Series A roughly two years later, worth tracking as a leading indicator of down-round pressure" — a risk signal that's easy to miss amid celebration of large seed rounds.