The Pitch (60s / 3min / 5min)

In this module

Learning objectives
Why three pitches, not one — pitch length is a function of audience time, not content depth
The 60-second CEO pitch — when you have an elevator
The 3-minute CTO pitch — discovery-stage technical credibility
The 5-minute security/compliance pitch — procurement-gate readiness
The research foundation — Gong + Dunford on what actually works
The practice loop — record-watch-iterate
Common failure modes
Discussion prompts

Learning objectives

After this module you should be able to:

Deliver each of the three pitch versions from memory, with appropriate pacing
Recognize when to switch from one version to another mid-conversation
Diagnose which beat of the pitch is failing when a buyer doesn't engage
Practice with a peer using the record-watch-iterate loop

Why three pitches, not one

The mistake most early-stage GTM teams make is treating "the pitch" as a single artifact. It's not. It's a family of artifacts tuned to time + audience attention budget. Three versions cover ~90% of real-world situations:

Version	When	Buyer goal	Your goal
60-second CEO pitch	Elevator, conference floor, intro email reply, end of a discovery call when you have leftover time	"Is this worth a follow-up?"	Earn the next meeting
3-minute CTO pitch	Scheduled discovery call, technical first-look, kickoff intro	"Is this technically defensible?"	Earn the demo
5-minute security/compliance pitch	Procurement gate, CISO intro, vendor risk review	"Can this clear our review process?"	Earn the security questionnaire

The lengths are buyer-attention budgets, not content quotas. A 3-minute pitch is not a 60-second pitch with more detail — it's a different artifact with different beats. Mixing them is the most common failure mode.

Gong's research, based on 67,000+ analyzed sales demos, found that unsuccessful demos contain monologues up to 106 seconds uninterrupted while successful demos are short bursts with frequent buyer chime-ins. The implication: the "60-second pitch" isn't really 60 seconds of you talking — it's 4 beats of ~15 seconds each, with buyer reactions creating natural breaks. Plan the beats, not the script.

The 60-second CEO pitch

Four beats, ~15 seconds each.

Beat 1 — The pain (compressed)

"Most product orgs don't actually know which shipped work moved their strategic metrics. They have AI tools that ship faster, but they're shipping the same wrong things faster."

Beat 2 — The differentiation (sharpened)

"PM33 closes the loop. Every Brief declares its predicted impact before shipping; the platform measures realized impact after; the model recalibrates against your workspace's actual history."

Beat 3 — The competitive urgency

"First-mover advantage is real here. Workspace-specific Bayesian priors take 12-18 months to converge. The orgs that start now will out-learn their competitors structurally."

Beat 4 — The specificity hook

"Last sprint, our team's onboarding-area Briefs landed within their predicted 80% confidence band. Quarter-end attribution wrote itself."

Practice rule: memorize the beats, not the words. Word-for-word memorization sounds canned; beat memory sounds rehearsed-but-real. The buyer remembers ~one beat — make sure it's Beat 2 (closed-loop differentiation) or Beat 4 (specificity).

The 3-minute CTO pitch

Discovery-stage pitch with technical credibility. Six beats, ~30 seconds each.

Beat 1 — Acknowledge the AI coding tools they already have

"You're probably running Cursor or Copilot. Great. That layer's table stakes now."

Beat 2 — Frame the next layer

"The question we ask: when your AI-augmented team ships, do you know whether you shipped the right thing? Not 'did it pass tests' — did it move the strategic metric you cared about?"

Beat 3 — Introduce the Brief schema

"Every PM33 unit of work — we call it a Brief — declares a predicted outcome before shipping. Machine-verifiable acceptance criteria. Specialist + LLM tier. TDD phases. The platform's harness executes against the spec."

Beat 4 — Execution architecture (technical credibility)

"Multi-agent execution runs in per-agent git worktrees. Lifecycle events stream to an audit log. A coordinator agent dispatches gauntlet reviews on security/auth/contract surfaces. We treat AI agents like peers who happen to have known failure modes."

Beat 5 — Closed-loop attribution

"When a Brief ships, the outcomeHook measurement window opens. Predicted vs. realized metric movement is recorded. AR(1) priors update with each shipped Brief — your workspace gets a predictive model that's specifically yours, not industry default."

Beat 6 — Compliance + lock-in framing

"The audit trail is what makes this enterprise-ready. Every state transition is recorded. Your CISO sees who-did-what-and-when. The compounding advantage is the workspace prior, which is yours, not ours."

The 3-minute pitch is harder than the 60-second pitch. You're not just saying more — you're proving you can reason about the technical architecture without losing the strategic frame. The CTO listens for two things: "do they actually understand the stack?" and "is the strategic claim defensible?" Both have to land.

The 5-minute security/compliance pitch

This is a different beast. Read it as a security questionnaire summary delivered out loud, not as a sales pitch. The CISO's question list drives the structure.

Map directly to engineer-track Module 5 (Governance & Trust) — slide 05 is the visual. Hit these five gates in order:

Multi-tenant isolation — every row in every table is RLS-scoped. Tenant context is set per request; nothing crosses tenants.
Audit log completeness — every state transition fires a structured event. Audit log is queryable by your security team.
Role-based access — 15-role hierarchy, with bypass tracking when overrides happen. Bypasses are observable, not blocked, by design (covered in security-compliance track Module 4).
Agent oversight — every PR goes through human review at the merge gate. AI agents work in isolated per-agent worktrees; no shared state.
Data residency + retention — your data, your retention policy. Self-hosted option available for enterprise tier.

5-minute pitch closing line: "Our security & compliance track has a 20-minute procurement-decision read. I can send it after the call so your team can pre-review before we meet." Always close with a specific next step that respects the CISO's time.

The research foundation

Two sources do most of the heavy lifting for this module:

Gong's call-analysis research (primary)

Gong analyzes call recordings at scale. Their published findings:

67,000+ demos analyzed — unsuccessful demos contain 106-second uninterrupted monologues; successful demos have frequent buyer interruptions
326,000+ calls analyzed for talk-to-listen ratios — winning discovery calls are 46:54 (rep:prospect). Winning demo calls are 65:35 (rep talks more once buyer wants to see it work)
Implication for the 3 pitches: the 60-second pitch should land 1-2 buyer reactions. The 3-minute pitch should land 3-5. The 5-minute security pitch can run more linearly but should pause for "any questions on that section?" between gates.

April Dunford's Sales Pitch (2023)

Dunford's framework for B2B SaaS pitch structure. Her core insight applicable here: don't show product features — demonstrate why your differentiated value matters. The 4-beat / 6-beat structures above implement Dunford's "Insight → Alternatives → Differentiated Value → Proof" pattern, compressed for time.

Dunford is the canonical book for sales-pitch architecture in B2B SaaS post-2020. If your reps have time to read one book, this is the one.

Honest caveat

Frameworks like MEDDIC, Challenger, and Force Management are widely used but the primary research behind them is mostly proprietary. The Challenger Sale (Dixon & Adamson, 2011) is the exception — it's backed by a 6,000-rep study via CEB (now Gartner). The 3 PM33 pitches incorporate Challenger principles (insight-led, teaching-oriented, comfort-with-tension) without slavishly following the framework.

The practice loop

This is the part most reps skip — and it's the part that produces certified-pitch-ready AEs.

Record yourself doing each pitch out loud. Phone voice memo is fine. No script in front of you.
Watch back at 1.25x speed. Fast playback reveals filler words, lost-attention moments, and rushed beats.
Score yourself on three axes (1-5):
- Did I land Beat 2 (the differentiation) clearly?
- Did I pause for buyer reactions (or imagine them)?
- Was my closing specific (next step) or vague?
Iterate weekly until each pitch is at 4+/5 on all three axes.

Most reps over-rotate on "we use AI" (table stakes, doesn't differentiate) and under-emphasize the closed loop (the actual differentiator). The fix is structural: practice forces you to confront the imbalance.

Common failure modes

Pitching all 3 versions in 60 seconds. You lose the buyer at second 30. Don't try to fit the full case into the elevator pitch — earn the next meeting, then earn the next, then close.
Word-for-word memorization. Sounds canned; buyers tune out. Beat memory + improvised words = rehearsed-but-real.
Skipping Beat 4 (specificity hook). Generic claims ("we help orgs ship better") don't stick. A specific example ("last sprint, our onboarding Briefs landed within predicted CI") is the part the buyer remembers.
Burying the differentiation. Beat 2 (closed-loop) must land first. Buyers who don't catch the closed-loop framing assume PM33 is "another AI coding tool" and you've lost.
No closing CTA. Every pitch ends with a specific next step. "Would you have 30 minutes for a discovery call next week?" beats "let me know what you think."

PM33 ships Builder-track Module 6 with the full pitch scripts in three lengths. Internal version of this document is the canonical reference. The sales team has its own copy in this track.

The engineer-track Module 6 is technical-foundation pitch language. This Module 1 is sales-delivery context. Read both before your first pitch — engineer-track gives you the what, this gives you the how and the when.

Discussion prompts

For team practice sessions:

The Beat-2 test: deliver your 60-second pitch to a peer. After, ask them to repeat back the differentiation in their own words. If they can't, your Beat 2 isn't landing.
The version-switching test: a peer plays a CEO who interrupts your 3-minute pitch to say "give me the short version." Practice switching mid-pitch to the 60-second version without flailing.
The objection-during-pitch test: a peer raises objection #4 ("we tried AI dev tools before; it was a fad") during Beat 1 of your 3-minute pitch. Practice handling the objection without losing the pitch's structural integrity.
The talk-ratio test: record a discovery call simulation. Compute your talk-to-listen ratio. Gong's data says winners hit 46:54 in discovery. Where are you?