The Twenty Minute VCMitchell Green: Why 50% of VCs Should Not Exist
CHAPTERS
- 0:00 – 2:35
SaaS sell-off: why Mitchell is buying incumbents on the dip
Mitchell argues the SaaS drawdown is more about expectations resetting than incumbents becoming obsolete. He explains why Lead Edge is actively buying public software names and why distribution, data, and balance sheets still matter.
- •Lead Edge buying public software: Procore, Workday, Appian, Toast; notes on Clearwater going private
- •Incumbents won’t vanish overnight; disruption creates winners and losers but not mass extinction
- •Core moat ingredients: distribution, data, balance sheet strength
- 2:35 – 5:02
Workday, seat-model fears, and the real reason numbers got cut
Harry challenges whether slow growth and weak agent rollouts signal structural decline. Mitchell responds that law of large numbers and overly optimistic sell-side estimates drove the sell-off more than AI replacing the business.
- •Workday as a $10B revenue / high FCF business; growth naturally slows at scale
- •AI revenue inside incumbents can grow fast even if consolidated growth decelerates
- •Sell-side estimates were too high; down cycles often reflect estimate revisions catching up
- •Stocks tend to work again once numbers are reset and beats resume
- 5:02 – 6:09
How to invest into falling markets: averaging in, floors, and earnings discipline
Mitchell outlines a pragmatic approach for buying amid uncertainty: don’t try to call the bottom, focus on cash-flow support, and scale in over time. He contrasts companies with earnings (a valuation floor) versus those without.
- •If there’s no earnings/EBITDA, there may be ‘no floor’
- •Practical tactic: buy in tranches over weeks; add on down days
- •Long-run data favors consistent buying over timing bottoms
- •‘Catching falling knives’ is less risky when fundamentals and cash flow anchor value
- 6:09 – 9:20
Founder-led advantage and the leverage trap during tech transitions
They debate whether non-founder CEOs are disadvantaged in AI-era transformation. Mitchell argues mindset (growth vs margin) matters most, and highlights leverage as the true vulnerability when disruption hits.
- •Founder-led or growth-minded leadership tends to navigate transitions better
- •Companies run purely for margins often correlate with higher leverage
- •Leverage limits innovation capacity during disruption
- •1999 retail analogy: some incumbents adapted; heavily levered ones died
- 9:20 – 16:35
ByteDance as an AI leader—and why the biggest AI winners may not exist yet
Mitchell claims ByteDance is the world’s most advanced AI company and is underrated in the West. He also argues today’s AI wave resembles 1999: the truly massive businesses may emerge in the next 2–5 years, not necessarily from current obvious categories.
- •ByteDance’s AI deployment and investment are underappreciated outside China
- •AI will reshape industries, but the largest new categories may be unforeseen today
- •Many current AI startups will win big, many will fail—classic disruption outcomes
- •Analogy: in 1999 no one predicted social media’s eventual scale
- 16:35 – 18:16
AI won’t kill software: legacy durability, retention, and where disruption really hits
Harry presses on whether AI is an unprecedented inflection that could obsolete incumbents. Mitchell argues legacy software persists for decades and the real disruption may hit leveraged, non-innovating firms and non-software sectors like healthcare and manufacturing.
- •Mainframes still run banks; legacy giants (Oracle, Microsoft, SAP) aren’t going away
- •Focus on companies with very high gross dollar retention
- •AI-driven disruption likely bigger in healthcare, manufacturing, drug discovery, robotics
- •Unlevered firms can invest through the transition; levered firms struggle
- 18:16 – 21:36
Jobs and productivity: change is slower than people think
Mitchell downplays near-term mass unemployment fears and predicts retraining and gradual adoption, especially in regulated industries. He frames AI as a productivity boom rather than an immediate job killer and notes government intervention would occur in extreme scenarios.
- •Historical analogy: switchboard operators and other displaced jobs; new work emerges
- •Regulated firms often can’t even use LLM tools yet—adoption friction slows impact
- •Companies likely retrain staff; productivity increases can offset headcount growth
- •If unemployment spikes dramatically, policy response would be swift
- 21:36 – 24:56
Meme-ified markets, social media volatility, and the SBC dilution blind spot
They discuss how viral narratives and retail flows amplify swings, creating opportunities for long-term buyers but complicating price discovery. Mitchell also flags stock-based compensation as an under-discussed driver of ‘not actually cheap’ valuations.
- •‘Casinoization’ of markets: narratives can move billions quickly
- •Volatility can create value opportunities for fundamentals-based investors
- •If you don’t have earnings, there’s no floor; earnings + buybacks matter
- •Stock-based comp and dilution materially affect true shareholder returns
- 24:56 – 31:11
China discount, listing realities, and why Mitchell thinks China can win AI
Mitchell explains how he thinks about ByteDance valuation via earnings multiples and peer comps, while acknowledging it won’t list in the US. He then argues China has structural advantages in AI—especially power buildout, resources, and STEM depth—while warning about US local pushback on data centers.
- •Valuation framing: compare to Alibaba/Tencent earnings multiples vs Western peers
- •No US listing; likely Hong Kong—yet sentiment on China can swing materially
- •China advantages: power build speed, resources, PhDs, tech prioritization
- •US/EU may face local backlash over data centers, electricity prices, and regulation
- 31:11 – 36:14
Selling discipline: re-underwriting, ‘in the money in 18 months,’ and price vs quality
Mitchell shares his core approach to position sizing and exits: continually re-underwrite and treat selling as the job. He introduces a key heuristic—ensure an investment can be ‘in the money’ on reasonable multiples within ~18 months—and separates ‘good company’ from ‘good investment.’
- •‘Buying is glamorous, selling is the job’—constant re-underwriting
- •Heuristic: can you be in the money 18 months out at a reasonable multiple?
- •Exit thinking: probability-weighted doubling vs taking chips off at extreme prices
- •Good company ≠ good investment; seek the intersection of quality and price
- 36:14 – 41:29
Too many VCs: fund size math, tourists, and why DPI beats marks
Mitchell argues the venture market is overcrowded and undisciplined on price, with too much capital chasing too few truly scalable outcomes. He emphasizes liquidity management: ‘marks are opinions, DPI is math’ and predicts LPs will increasingly demand cash returns.
- •Claim: 50%+ of VCs add negative value; too many ‘tourists’ and too much money
- •Mega-fund math forces underwriting $150B+ outcomes; few companies reach that scale
- •Liquidity windows open/close—sell portions even of winners to return capital
- •LP focus shifting hard toward DPI rather than paper marks
- 41:29 – 44:55
How investors destroy value: bad burn advice, false expertise, and real founder support
Mitchell outlines common investor behaviors that harm companies—pushing reckless burn, meddling without operator experience, and giving generic advice. He argues the best investor value is recruiting and connecting founders with experienced operators, then getting out of the way.
- •Negative value patterns: ‘burn at all costs,’ overstepping into operations, mis-hiring
- •Most investors haven’t run companies; humility and operator networks matter
- •Best help: recruiting leaders who have scaled from $20M to $200M
- •Strong brands (e.g., top funds) help portfolio hiring and credibility
- 44:55 – 46:43
The metric that matters most: Gross Dollar Retention (GDR) and the ‘living dead’ problem
Mitchell calls gross dollar retention the most important metric for software durability and efficient growth. He argues low GDR creates ‘living dead’ companies that must overspend on sales just to refill churn, which becomes fatal at scale.
- •GDR definition: revenue retention from same customers excluding upsells
- •Benchmarks: ~90% good, 95% great, 98% exceptional; sub-80% is dangerous
- •Low GDR forces high S&M just to maintain revenue; scaling magnifies the issue
- •Explains why many venture-backed companies stagnate despite headline growth playbooks
- 46:43 – 1:00:09
Private equity’s leveraged SaaS portfolios and the worsening liquidity environment
They discuss how debt constrains PE-owned SaaS assets’ ability to invest through disruption, and why leveraged models are vulnerable in tech shifts. They also connect meme-driven public markets to founders delaying IPOs, deepening the liquidity crunch—until LPs force distributions.
- •Debt-heavy SaaS portfolios have less flexibility to invest in AI and innovation
- •Operational margin expansion can be fine if not achieved by gutting engineering
- •Volatile publics incentivize companies like Stripe/Canva to stay private
- •Liquidity pressure likely resolves when LPs refuse to re-up funds without exits
- 1:00:09 – 1:04:14
Quick-fire: lessons, misses, fund picks, and preparing for the next downturn
In closing, Mitchell shares what he’s changed his mind on, memorable founder meetings, and his biggest career misses. He says the next decade likely includes a major downturn—and that having capital available then will define the best investors and vintages.
- •AI will be even bigger than expected and impact more sectors than currently modeled
- •Biggest miss: not paying up in 2016–2018; Shopify IPO sell too early
- •Fund preferences: Benchmark (Series A); respects Iconiq; notes Founder Fund performance
- •Expects a major downturn within 10 years—best time to deploy capital if prepared
