growth-sales-cro-core

name	growth-sales-cro-core
description	Shared reference for the growth-sales-cro cluster: the conversion funnel, the baseline → hypothesis → test → decide measurement loop, the metric-per-stage matrix, and the rule that you optimize on evidence not opinion. USE WHEN proposing a conversion/pricing/retention change, designing an A/B test, or planning a growth experiment — the interlocking discipline every spoke shares.
cluster	growth-sales-cro
version	1.0.0

Growth-Sales-CRO Core

Shared model for the growth-sales-cro cluster. Every CRO, growth, monetization, retention, and sales spoke depends on the same two things: a funnel you can see and a loop that proves a change worked. Keep them consistent here so no spoke optimizes on vibes or double-counts a win.

1. The one decision this domain turns on

Optimize on evidence, not opinion. Before any change earns the word "win", it must be tied to a single metric, framed as a hypothesis, and validated against a measured baseline — ideally by a controlled test.

Everything below exists to make that decision honest. The failure mode of growth/CRO work is shipping opinion-driven changes, declaring victory from noise, and stacking unmeasured tweaks until nobody can attribute the result. The cluster's job is to prevent that.

2. The funnel (the map every spoke lives on)

Pirate-metrics order, plus the parallel human-assisted path:

ACQUIRE ─→ CONVERT ─→ ACTIVATE ─→ MONETIZE ─→ RETAIN ─→ EXPAND
   │          │           │           │          │         │
 paid-ads   page-cro   onboarding   pricing    churn-    referral-
            form-cro    -cro        -strategy  prevention program
            popup-cro              paywall-
            signup-flow-cro        upgrade-cro

          ── parallel SELL path (human-assisted) ──
   revops → sales → sales-enablement → pitchdeck-skill → lead-research-assistant

          ── cross-cutting EVIDENCE (feeds every stage) ──
   analytics-tracking · ab-test-setup · customer-research · company-research
   competitor-alternatives · competitor-teardown

Rule: locate every task at exactly one stage and one lever. "Grow revenue" is not a task; "lift trial→paid on the paywall" is. Vague asks get decomposed into stage-specific, single-metric tasks before any spoke runs.

3. The measurement loop (how a change becomes a "win")

Self-serve or sales, the loop is the same:

BASELINE ─→ HYPOTHESIS ─→ TEST ─→ DECIDE ─→ (ship or kill) ─→ re-baseline

• Baseline — instrument the metric before changing anything. No instrumentation → start at analytics-tracking; you cannot improve what you can't see. → analytics-tracking
• Hypothesis — "Changing X will move metric M by ~N because reason grounded in evidence." One change, one primary metric. Evidence comes from customer-research or competitor findings, not taste.
• Test — a controlled comparison where stakes justify it. Respect statistical significance and a minimum runtime / sample size; don't peek-and-stop. → ab-test-setup
• Decide — ship, kill, or iterate based on the result and its guardrail metrics (did the win cannibalize another stage?). Then re-baseline; the funnel moved.

4. Metric-per-stage matrix

Stage	Primary metric	Lead spokes	Evidence input
Acquire	CPA / ROAS, qualified traffic	paid-ads	competitor-*, analytics-tracking
Convert	conversion rate (visit→action)	page-cro, form-cro, popup-cro, signup-flow-cro	customer-research, ab-test-setup
Activate	activation rate, time-to-value	onboarding-cro	customer-research, analytics-tracking
Monetize	trial→paid, ARPU, win rate	paywall-upgrade-cro, pricing-strategy, sales, sales-enablement	pricing-strategy, customer-research
Retain	churn rate, recovered revenue	churn-prevention	customer-research, analytics-tracking
Expand	referral / viral coefficient, NRR	referral-program	analytics-tracking
Sell (parallel)	pipeline, MQL→SQL, close rate	revops, sales, sales-enablement, pitchdeck-skill, lead-research-assistant	company-research, competitor-*
Mobile listing	store conversion, keyword rank	app-store-optimization, app-store-screenshots, aso-appstore-screenshots	customer-research, competitor-*

Cross-cutting evidence spokes (analytics-tracking, ab-test-setup, customer-research, company-research, competitor-alternatives, competitor-teardown) are not a funnel stage — they are the substrate the stages cite.

5. Conventions

• Voice of customer over copywriter instinct. Headlines, offers, and objection handling should quote real customer language (customer-research), not invented benefits.
• One change per test. Stacking changes destroys attribution. If you must ship a bundle, say so and accept you've traded learning for speed.
• Guardrail metrics. Every primary metric has a guardrail (e.g. lifting signups but tanking activation is a loss). Name it before shipping.
• Public pricing vs in-app upgrade are different surfaces. pricing-strategy sets what you charge on the marketing site; paywall-upgrade-cro is the in-product moment after value is felt.
• Self-serve and sales are two paths to the same revenue. revops is the seam — lead scoring/routing and the marketing→sales handoff — not a funnel stage of its own.

6. Shared guardrails

• One change, one metric, one hypothesis — name the number before touching the funnel.
• No win without a baseline. Don't call an untested change a win; instrument first.
• Respect significance + minimum runtime. No peeking-and-stopping; no declaring victory from noise.
• Ground copy and offers in evidence (voice of customer, competitor data) — never fabricate metrics, lift percentages, or competitor claims.
• State the blast radius of revenue changes. Pricing and paywall edits affect real money and often other stages — call out the assumption and what could regress.
• Decompose vague asks. "Grow faster" → a specific stage, lever, metric, and test before any spoke runs.
• This cluster owns convert → monetize → retain → sell; top-of-funnel creation (brand voice, long-form, launch campaigns, investor narrative) lives in the business-content cluster — hand off rather than duplicate.