Alex Rampell on Venture at Scale and Founder Incentives

Rampell explains why venture funds must either be large generalists or small specialists to succeed. He argues that while small funds can generate higher multiples mathematically, large funds can return more absolute dollars by winning the best consensus deals. The key is finding, picking, and winning investments - with 'winning' being the hardest part as it requires selling entrepreneurs on taking your money.

• •Asset classes tend toward 'death of the middle' - either be a large generalist or small specialist
• •Small funds can achieve higher multiples (Mickey Malka's 55x on $85M fund) but large funds return more absolute dollars
• •Companies now stay private much longer - Series D didn't exist in 1992, Amazon IPO'd at $600M market cap
• •Winning the best deals requires either deep specialization or broad scale/resources to help founders
• •The best consensus deals (Uber, Facebook) often go to the best-known firms with resources

Rampell outlines his investment framework for evaluating founders: their ability to materialize labor (recruiting top talent for pay cuts), capital (fundraising ability), and customers (especially first enterprise customers). He emphasizes the importance of founders who study the history of their space and have 'Count of Monte Cristo' motivation - a chip on their shoulder driving them beyond just making money.

• •Invest in founders who can materialize labor, capital, and customers - especially critical as top talent is expensive
• •Best entrepreneurs study the complete history of their space (Patrick Collison meeting Dee Hock, Brian Chesky studying 1800s hotels)
• •'Count of Monte Cristo' motivation - founders need revenge/redemption drive beyond $50M-100M exits
• •High agency is critical - founders who won't be told what to do and take matters into their own hands
• •Getting first 5 customers can be harder than first 5 employees, especially in enterprise

Rampell introduces the concept that 'the best companies have hostages, not customers' - systems of record that are nearly impossible to switch away from. He explains the 'greenfield bingo' strategy of selling to new companies rather than trying to displace incumbents, and why high rates of new company formation enable this approach. The discussion covers how knowing too much about a space can be a liability.

• •Best companies have 'hostages not customers' - switching costs make them sticky (Workday, NetSuite)
• •Greenfield bingo strategy: sell to new companies who pick the best product vs trying to convert hostages
• •Mercury never stole an SVB customer until SVB failed - won by capturing new company formation
• •Knowing too much about a space can cause you to miss opportunities (Rampell passing on Stripe initially)
• •Rate of new company creation must be high enough to support greenfield strategy

Discussion of how venture firms balance ownership targets with winning competitive deals. Rampell explains the 'out of the money call option' framework for early-stage investing and why ownership matters more at Series A when buying options on potential. He addresses the challenge of successive quick rounds and when to pay up for winners versus maintaining ownership discipline.

• •Venture is buying 'out of the money call options' hoping they expire in the money
• •Two strategies: buy any percent of something absolutely working, or high ownership of something that could work
• •At seed, consensus is about the founder's agency and domain expertise, not metrics
• •If you win 100% of deals, you're probably not testing the efficient frontier on ownership
• •Series A ownership critical for large funds - need 15-20% to make fund math work given future dilution

Rampell strongly criticizes massive founder secondaries, arguing they create moral hazard by disconnecting founders from employees and investors. He explains how excess capital leads to 'foie gras-ing' of startups - doing 10 things instead of 2, hiring unnecessary layers, and losing the necessity that drives invention. The discussion covers founder-capital fit and why constraints often lead to better outcomes.

• •Hate massive secondaries - creates fundamental disconnect when founder is rich but employees/investors aren't
• •Moral hazard applies to both primary and secondary capital - removes necessity that drives invention
• •More input doesn't equal better output - smaller teams with less capital often outperform
• •Founder-capital fit matters: some founders stay motivated with wealth, others lose edge
• •Raising at too high a price creates problems for future rounds - first question is always 'what was last round price?'

Rampell outlines his three core investment theses: (1) Greenfield systems of record selling to new companies, (2) Software that does the job of labor with AI, and (3) Walled garden companies with unique data moats. He explains how AI is enabling software to replace labor in new categories while creating stickiness challenges for thin wrappers on foundation models.

• •Thesis 1: Greenfield bingo - systems of record (Rilla, Mercury) selling to new companies with sticky revenue
• •Thesis 2: Software replacing labor - selling $20K/year software instead of $80K/year employees (Eve for plaintiff attorneys)
• •Thesis 3: Walled gardens - unique data that even AGI can't replicate (Vlex with Spanish legal records, Open Evidence with health data)
• •Companies doing labor replacement must back into sticky systems of record to avoid commoditization
• •Three types of SaaS in AI era: impervious (Workday), decimated (Zendesk), and middle (Adobe)

Discussion of how AI will affect labor displacement across different sectors. Rampell argues that while some jobs (Zendesk users, tedious legal work) will be eliminated, companies may reallocate labor to higher-value activities like relationship building. He emphasizes that high-EQ roles in wealth management and customer relationships will become more valuable, not less.

• •Labor displacement will vary by sector - customer support and tedious work most affected
• •Smart companies will reallocate labor to revenue-generating activities (Zappos model of customer love)
• •High-EQ relationship roles (wealth management, account management) will become more valuable
• •Technology historically shifts labor rather than eliminating it entirely
• •Upskilling may mean moving from tedious work to relationship building, not 'everyone should code'

Rampell shares tactical advice on selling companies, emphasizing that M&A is a 'highly choreographed dance' requiring years of relationship building. He explains why you should spend time with potential acquirers' business unit leaders (not corp dev), ideally under the guise of partnership discussions. The best time to sell is when you're doing great, not when you need to.

• •Spend 5% of CEO time as 'background process' getting to know 3-4 potential acquirers years in advance
• •Build relationships with SVPs who would manage you post-acquisition, not corp dev (they just execute)
• •Approach under guise of partnership - genuine business value discussions, not 'please buy us'
• •Best time to sell is when rocket ship is growing 100x YoY, not when falling
• •Same advice applies to fundraising - maintain investor relationships as background process

Rampell reflects on passing on Plaid's Series B over a $5M valuation difference ($1.30B vs $1.35B), then correcting by doing the Series C at $2.4B. He shares his core investment advice: find high-agency people who know their space's history, can materialize labor/capital/customers, have Count of Monte Cristo motivation, and don't second-guess giving them money.

• •Biggest miss: debating $5M on Plaid Series B valuation, later invested at $2.4B Series C
• •Importance of correcting mistakes - willing to admit being wrong and do later round
• •Being 'burdened by what has been' (Yodlee's $600M exit) prevented seeing Plaid's potential
• •Core advice: 100% about people at every stage (seed through late stage)
• •If willing to take less ownership, don't do the deal - signals lack of conviction