06-reference

mike vernal death of three act playbook

2026-06-01·reference·source: X / Mike Vernal (long-form article)·by Mike Vernal
startup-strategyenterprise-saasai-disruptionambition-over-wedgego-to-marketcost-of-software

"The Death of the Three-Act Playbook" — @mvernal

Why this is in the vault

Mike Vernal (Investor + Engineer @ Conviction; ex-Facebook VP, ex-Sequoia) on why the classic enterprise-software GTM playbook is obsolete in an era of collapsing software cost. Shared by the founder 2026-06-01 ~21:54 ET — the third of three pieces he shared today, all on the same spine: the cost of software/intelligence trending to zero voids the old playbooks. The trilogy is the reason this is worth keeping as a set, not just individually. 182k impressions, 567 bookmarks.

The core argument

The classic enterprise-SaaS playbook had three acts, each gated by calendar time:

Vernal's claim: the playbook is dead — the world moves too fast. The three-act sequence relied on calendar time because founders were single-threaded on Act I. The wave of companies going ~$0→$100M ARR in a couple years (Cursor, Cognition, Clay, Harvey, Sierra, Baseten, Fireworks, Lovable) is the evidence. As software-engineering cost plummets, the time to finish Acts I+II approaches zero, so the rational move is to build it all quickly, mostly from the start.

His own investing shift: he used to hunt for the "protective wedge" (the safe harbor to $10-50M ARR); now the wedge feels like "small ball" and he wants founders to "jump straight into the deep end." Anecdote: Anysphere/Cursor at seed planned to straight-up replace VS Code (not build an extension first) — he thought it was crazy; he was wrong, and now replacing VS Code "almost feels under-ambitious." Close: as the cost of writing software drops to zero, he values "unreasonable, unrelenting ambition" above all; relying on a wedge in a period of rapid change is too timid — "if you're going to go for it, you should probably just go for the whole thing."

Mapping against Ray Data Co

  1. The wedge-vs-ambition tension, applied to RDCO (surfaced to founder as a provocation, not a verdict — it's his strategic call). On the surface, RDCO is wedge-shaped: Squarely, Sanity Check, and MAC each built incrementally as small bets. Vernal's frame would call that the timid three-act instinct. The reconciliation: RDCO's actual unreasonable-ambition bet isn't any single product — it's the COO agent itself. Per the L5 north star ([[project_l5_north_star_strategic_direction]]), bets are downstream of agent capability, and the primary focus is unhobbling the agent (toolset + visibility), NOT operating small bets first. Read that way, RDCO is already playing Vernal's "go for the whole thing" game — it's just pitched at the agent-platform level rather than the per-bet level. The open gut-check for the founder: are the small bets the obsolete wedge instinct, or the right portfolio sitting under an agent-capability platform play?
  2. The today-trilogy through-line. All three pieces the founder shared 2026-06-01 share one spine — the collapsing cost of intelligence/software voids the old playbooks — at three altitudes:
    • Builders → [[2026-06-01-garry-tan-stop-building-foxconn-factories-for-agents]] (markdown skills, not 540k lines of caging code; tokenmax).
    • Enterprises → [[2026-06-01-jaya-gupta-token-budget-wars]] (allocate tokens to outcomes; marginal token utility).
    • Startups → this piece (skip the wedge; ambition over safety). The convergence is itself the signal: three credible operators independently describing the same regime shift. Candidate raw material for an original-reframe Sanity Check angle (NOT a restatement — would need RDCO's own synthesis to clear the no-derivative bar).

Related