Growth Sales Cro Orchestrator

Growth-Sales-CRO Orchestrator

The single entry skill for moving a number — conversions, activation, revenue, retention, closed deals. It locates the task on the funnel stage × lever map and delegates to one of 24 specialist spokes. The cross-cutting model every spoke shares — the funnel itself, the baseline → hypothesis → test → decide measurement loop, and the rule that you optimize on evidence rather than opinion — lives in growth-sales-cro-core; read it before proposing any change you intend to call a "win", before designing a test, or before touching pricing.

Routing map (intent → spoke)

Measure first (shared substrate — almost always step zero)

• Instrument events, GA4/GTM/Mixpanel/Segment, tracking plan, attribution → analytics-tracking (if you can't see the metric, you can't move it)
• Design the experiment that proves a change works → ab-test-setup (hypothesis, variant, significance, runtime)

Convert (self-serve funnel — the heart of CRO)

• A marketing page (home, landing, pricing, feature, blog) isn't converting → page-cro
• A non-signup form (lead, contact, demo, application, checkout) → form-cro
• A popup, modal, overlay, exit-intent, sticky bar → popup-cro
• A signup / registration / trial-start flow → signup-flow-cro
• Post-signup activation, first-run, time-to-value, "they sign up but don't stick" → onboarding-cro
• In-product upgrade screen, upsell, feature gate, free→paid → paywall-upgrade-cro

Monetize (what & how you charge)

• Pricing tiers, packaging, value metric, freemium, price increase → pricing-strategy (public pricing; for the in-app upgrade moment use paywall-upgrade-cro)

Acquire (paid demand)

• Google/Meta/LinkedIn/X campaigns, ROAS, CPA, targeting, retargeting, bidding → paid-ads

Retain & expand (keep and grow the base)

• Churn, cancel flow, save offers, dunning, win-back, retention → churn-prevention
• Referral / affiliate / ambassador / viral-loop programs → referral-program

Sell (human-assisted path, runs parallel to self-serve)

• Lead lifecycle, scoring, routing, MQL→SQL, marketing→sales handoff, CRM hygiene → revops
• Sales narratives, talking points, value framing from product docs → sales
• Sales collateral: decks, one-pagers, objection handling, demo scripts, battle-card content → sales-enablement
• Investor / fundraising pitch experience (scrollytelling deck artifact) → pitchdeck-skill
• Find and qualify target accounts/contacts for outbound → lead-research-assistant

Research (cross-cutting evidence — feeds every lane)

• Voice-of-customer: interviews, transcripts, reviews, JTBD, personas, "why they churn/convert" → customer-research
• Company brief: exec quotes, product strategy, org context for a target/partner → company-research
• Competitor comparison / "vs" / alternative pages (SEO + sales enablement) → competitor-alternatives
• Structured teardown of a rival's positioning, messaging, channel behavior → competitor-teardown

App-store growth (mobile listing surface)

• ASO: keywords, metadata, ranking, store listing optimization → app-store-optimization
• Store screenshots / preview gallery to platform spec → app-store-screenshots
• Codebase-driven, ASO-optimized generated screenshots → aso-appstore-screenshots

Standard Operating Flow

1. Locate the task on the map: which funnel stage (acquire → convert → activate → monetize → retain → expand, or the parallel sell path) and which lever.
2. Establish the metric first. If the request implies a "win" but no baseline or instrumentation exists, route through analytics-tracking before changing anything, and through ab-test-setup if the change is worth proving. The model is in growth-sales-cro-core.
3. Delegate to the spoke(s). Multi-step asks fan out in funnel order — e.g. "more paying users" → page-cro (top) → signup-flow-cro → onboarding-cro → paywall-upgrade-cro, each gated by a measured hypothesis.
4. Ground in evidence. Conversion and messaging changes should cite customer-research (voice of customer) or competitor findings, not taste.
5. Return: chosen spoke(s), the single metric each change targets, the test plan (or why none), and the next action.

Guardrails

See growth-sales-cro-core. In short: one change, one metric, one hypothesis — name the number before you touch the funnel; don't call an untested change a "win"; respect statistical significance and minimum runtime before declaring a winner; ground copy and offers in voice-of-customer evidence, not opinion; never invent metrics, lift percentages, or competitor claims. Pricing and paywall changes are revenue-affecting — state the assumption and the blast radius before shipping.

External companions (not in this cluster)

• alex-hormozi-pitch and customer-sales-automation ship as a separate bundle (an empty/agent-only package without a top-level SKILL.md), so they are archived outside this cluster. Reach for the bundle when you want Hormozi-style offer/pitch framing or the prebuilt sales-automation/support agents; this cluster's sales, sales-enablement, and pitchdeck-skill cover the same intents natively.

Loading spokes on demand

To keep CLI startup context lean, this cluster's spokes are not separately registered as skills — only this orchestrator and its *-core are enumerated. When you route to a spoke named above, load it on demand by reading its file:

~/.agents/skill-clusters/skills/<spoke-name>/SKILL.md (or skills/<spoke-name>/SKILL.md inside the skill-clusters repo).