# Gtm System Architecture > >- - **Type:** Skill - **Install:** `agentstack add skill-leadmagic-gtm-skills-gtm-system-architecture` - **Verified:** Yes — security-reviewed for prompt injection and unsafe behavior - **Seller:** [LeadMagic](https://agentstack.voostack.com/s/leadmagic) - **Installs:** 0 - **Category:** [Communication](https://agentstack.voostack.com/c/communication) - **Latest version:** 0.1.0 - **License:** MIT - **Upstream author:** [LeadMagic](https://github.com/LeadMagic) - **Source:** https://github.com/LeadMagic/gtm-skills/tree/main/skills/analytics/gtm-system-architecture - **Website:** https://github.com/LeadMagic/gtm-skills ## Install ```sh agentstack add skill-leadmagic-gtm-skills-gtm-system-architecture ``` Requires the [AgentStack CLI](https://agentstack.voostack.com/docs/cli). Works with Claude Code, Cursor, and any MCP-compatible agent. ## About # GTM System Architecture ## Overview A GTM system is the operating system underneath your revenue. It's the answer to "how does money flow through this company?" — not "what tactics are we running this quarter?" Winning by Design's GTM Index framework scores 1-10 across six interconnected models. Most companies break because they optimize one model while neglecting the others. This skill designs all six as one unified system. ## When to Use - "Design our GTM system" - "Audit our revenue architecture" - "Build a GTM operating model" - "Our GTM is fragmented — fix it" - "Align sales, marketing, and CS" - "Design how revenue flows" ## Authoritative Foundations - **Winning by Design GTM Index** — six models (Revenue, Data, Math, Operating, Growth, GTM). 1-10 scoring per model. Average across 1000+ companies benchmarked. - **Winning by Design Bowtie** — revenue is a continuous loop: acquire → retain → expand. Not a funnel that ends at "closed won." - **SiriusDecisions Demand Waterfall** — demand flows through stages with defined handoffs and conversion metrics. ## Step-by-Step Process ### Phase 1: The Six Models (WbD GTM Index) Score each 1-10, then identify the weakest link: **1. Revenue Model — "Do we know how we make money?"** - Pricing clarity, packaging, discount authority, expansion model. - Weak: "We charge whatever the customer agrees to." - Strong: "Standard tiers, defined discount rules, clear expansion path." **2. Data Model — "Can we answer basic questions about our business?"** - CRM hygiene, pipeline data, conversion rates, attribution. - Weak: "We don't know our MQL→SQL conversion rate." - Strong: "Every deal has stage, source, close date, and MEDDIC dimensions." **3. Math Model — "Do unit economics work?"** - LTV:CAC, payback period, CAC by channel, churn rates, NRR. - Weak: "We don't track CAC by channel. We think churn is around 3%." - Strong: "CAC payback is 8 months, LTV:CAC is 4:1, NRR is 115%." **4. Operating Model — "Is there a documented process someone could follow?"** - Sales process stages, playbooks, enablement, handoff documentation. - Weak: "Reps sell however they want. No two deals look the same." - Strong: "Stage definitions with Goals and Exit Criteria. Playbook documented. New reps ramp in 90 days." **5. Growth Model — "Do we know which channels work and at what cost?"** - Channel mix, CAC by channel, channel ROI, experimentation velocity. - Weak: "We try things and see what sticks." - Strong: "Paid search CAC is $X, content CAC is $Y, outbound CAC is $Z. We allocate budget to highest-ROI channels." **6. GTM Model — "Is the whole system connected?"** - Sales-marketing-CS alignment, shared metrics, unified customer journey. - Weak: "Marketing generates MQLs sales ignores. CS doesn't know what was sold." - Strong: "Unified Bowtie. Shared metrics across functions. Single customer view from first touch through renewal." ### Phase 2: Bowtie Architecture Revenue doesn't end at "closed won." It continues through retention and expansion. Design the Bowtie: ``` Awareness → Education → Selection → Onboarding → Adoption → Expansion → Renewal [Marketing] [Marketing] [Sales] [CS] [CS] [CS/Sales] [CS] ``` Each stage has: owner (which team), goal (what success looks like), handoff criteria (when to pass to next stage), and metrics (how to measure). ### Phase 3: System Integration The six models don't operate independently: - **Revenue Model ↔ Math Model:** Pricing choices determine unit economics. Lower price → lower ARPU → longer CAC payback. Know the tradeoff. - **Data Model ↔ Operating Model:** If CRM data is unreliable, the process can't be inspected. Fix data before adding process. - **Growth Model ↔ GTM Model:** Marketing channels produce demand. Sales converts it. CS retains it. If any link breaks, the system fails. - **Weakest link principle:** The lowest-scoring model caps the entire system. A 9/10 Revenue Model with a 3/10 Data Model is a 3/10 system. ### Phase 4: Implementation Roadmap Don't fix all six at once. Sequence by impact: 1. **Fix the Math Model first.** If unit economics don't work, nothing else matters. Know your CAC, LTV, churn, payback. 2. **Fix the Data Model.** Clean CRM, define stages, establish reporting. 3. **Design the Operating Model.** Document processes, build playbooks. 4. **Optimize the Growth Model.** Test channels, measure CAC per channel. 5. **Refine the Revenue Model.** Pricing, packaging, discount rules. 6. **Align the GTM Model.** Cross-functional alignment, shared metrics. ## Output Format GTM system design: six-model assessment (current scores + target scores), Bowtie architecture with ownership per stage, weakest-link analysis, implementation roadmap with 90-day priorities, and measurement dashboard. ## Quality Check - [ ] All six models scored 1-10 with evidence - [ ] Weakest link identified (lowest scoring model) - [ ] Bowtie architecture with owner, goal, handoff criteria per stage - [ ] Implementation roadmap (not trying to fix everything at once) - [ ] Shared metrics across functions (not siloed KPIs) - [ ] Re-assessment cadence (quarterly GTM Index check) ## Common Pitfalls 1. **Optimizing one model in isolation.** Improving the Operating Model without fixing the Data Model means adding process to bad data. Fix data first. 2. **No Bowtie.** Treating revenue as a pipeline that ends at "closed won" ignores the 70%+ of lifetime value that comes after the sale. 3. **Self-assessment without evidence.** "I think our Math Model is a 7" without actually calculating LTV:CAC. Score with data, not feelings. 4. **Trying to fix everything at once.** Six-model overhauls take 12-18 months. Sequence one or two per quarter. 5. **GTM model owned by one function.** GTM alignment requires sales, marketing, and CS leadership in the same room with shared metrics. ## Execution Artifacts - `references/framework-notes.md` — named frameworks, citation anchors, and operating assumptions - `templates/output-template.md` — copy-paste deliverable structure for the user - `scripts/check-output.py` — local checklist validator for required sections This skill includes lightweight artifacts the agent can load on demand: Use the artifacts when the user asks for an implementation-ready deliverable, a repeatable workflow, or a quality check rather than generic advice. ## Related Skills - **gtm-metrics**: Model measurement and dashboards - **pipeline-management**: Operating model execution - **marketing-strategy**: Growth model design - **sales-team-building**: Operating model staffing ## Source & license This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code. - **Author:** [LeadMagic](https://github.com/LeadMagic) - **Source:** [LeadMagic/gtm-skills](https://github.com/LeadMagic/gtm-skills) - **License:** MIT - **Homepage:** https://github.com/LeadMagic/gtm-skills Install and usage instructions live in the source repository linked above. ## Pricing - **Free** — Free ## Security capabilities Automated source analysis of v0.1.0 — what this tool can access: - **Network access:** no - **Filesystem access:** no - **Shell / process execution:** no - **Environment & secrets:** no - **Dynamic code execution:** no *"Yes" means the capability is present in the source — more access means more to trust, not that it is unsafe.* ## Versions - **0.1.0** — security scan: passed — Imported from the upstream source. ## Links - Listing page: https://agentstack.voostack.com/l/skill-leadmagic-gtm-skills-gtm-system-architecture - Seller: https://agentstack.voostack.com/s/leadmagic - Browse the marketplace: https://agentstack.voostack.com/browse --- Listed on AgentStack — the marketplace for AI agent skills and MCP servers. Every listing is security-reviewed. Creators keep 70%.