AI Hypergrowth Gtm

AI Hypergrowth GTM

Use this skill to turn the AI Sales-Led Hypergrowth research corpus into concrete GTM decisions. Keep the answer operational: diagnose the current motion, identify the highest-leverage pattern, stress-test platform/distribution risk when relevant, and propose the next sales/pilot/proof step.

Source basis: Jina Reader exports of https://sevaustinov.me/hypergrowth-research/, saved in references/ on May 19, 2026.

Core Workflow

1. Identify the current stage: pre-product discovery, first design partners, founder-led sales, pilot conversion, first sales hire, or expansion.
2. Name the active wedge: buyer, painful workflow, measurable output, current human/labor cost, and why this can reach meaningful ARR before broadening.
3. If the idea touches agent infrastructure, coding agents, assistant operating systems, workflow entry points, generic MCP/tooling, or platform-owned data, run the platform-risk screen in references/distribution-endgame.md before recommending a wedge.
4. Apply the growth laws and sales laws below. Do not give generic SaaS advice if one of these laws gives a sharper move.
5. Load detailed references only as needed:
   • Full map: references/source-index.md
   • Fast case selector: references/case-metaprinciples.md
   • Platform/distribution risk: references/distribution-endgame.md
   • Company cases: references/companies/<company>.md
   • Growth law details: references/growth-laws/<law>.md
   • Sales law details: references/sales-laws/<law>.md
   • Matrices: references/data/pattern-matrix.md, references/data/comparison-tables.md
6. Output a concrete recommendation: target buyer, offer, pilot terms, proof metric, demo input, pricing anchor, platform-risk verdict when relevant, and next 5-10 founder actions.

Reference Loading Rules

• Start with references/case-metaprinciples.md when choosing which company cases matter. It maps each case to the transferable principle, when to load it, and what not to overgeneralize.
• Load the full company file after selecting 1-3 matching cases; do not answer a serious GTM question from the summary table alone.
• Load references/distribution-endgame.md whenever the product could be described as generic agent infrastructure, generic coding-agent workflow, agent OS, agent runtime, tool marketplace, connector layer, eval/observability layer, or "AI for work" surface.
• For agent-infrastructure ideas, default to the inversion question: what will the distribution owner not do, not be trusted to do, or not be economically incentivized to do?

Growth Laws

1. Start with the wedge that prints value. Pick one workflow with obvious economic value, not a horizontal platform. Scale the wedge hard before broadening. Use references/growth-laws/wedge-clarity.md.
2. Win the buyer the market follows. In prestige-sensitive categories, do not start mid-market and hope to move up. Use a top-tier buyer, design partner, or anchor logo to create a trust cascade. Use references/growth-laws/prestige-first.md.
3. Domain-expert GTM beats generic sales. Prefer sellers, solutions leads, and CS people who have done the buyer's job. A lawyer selling to lawyers or clinician selling to clinicians beats a generic AE memorizing a product. Use references/growth-laws/domain-expert-gtm.md.
4. Build proof that cannot be argued with. Engineer paid pilots/design partners with pre-agreed metrics, baselines, timelines, and conversion terms before scaling GTM. Use references/growth-laws/proof-before-scale.md.
5. Price against labor cost, not software alternatives. Anchor to the human work being replaced or compressed: hours saved, headcount avoided, agency/process cost removed, or outcome delivered. Use references/growth-laws/labor-budget-pricing.md.
6. Build expansion into product logic, not sales motion. Land in one workflow/team, then expand because usage, outcomes, seats, data, departments, or resolved cases naturally grow. Use references/growth-laws/expansion-flywheel.md.

Supplemental growth patterns exist for trust architecture, high-touch implementation, non-black-box design, ICP discovery filters, founder timing, and product arc. Load the matching files under references/growth-laws/ when the problem touches those topics.

Platform-Risk Law

Distribution eats generic agent infrastructure. If Google, Microsoft, Apple, OpenAI, Anthropic, or a cloud/platform owner can ship the feature into the workflow entry point, assume the base layer will be commoditized. The survivable wedge is usually a neutral, workflow-specific, compliance-sensitive, budget-owned control point that the platform owner cannot or will not own cleanly. Use references/distribution-endgame.md.

Sales Laws

1. Founders sell every deal until the motion is proven. Do not hire a VP Sales to discover the motion. Founders validate ICP, pricing, objections, demo, success metrics, and deal structure first. Use references/sales-laws/founder-sells-first.md.
2. Demo against their own work, not sample data. Use the prospect's documents, workflows, public artifacts, tickets, filings, calls, or ecosystem data. Generic AI demos create distance. Use references/sales-laws/demo-against-own-work.md.
3. Willingness to pay is a signal, not enthusiasm. Ask directly what they would pay. Specific high-dollar urgency from a budget owner matters; "interesting, maybe later" is a weak signal. Use references/sales-laws/wtp-as-qualification.md.
4. Use paid pilots with pre-agreed exit conditions. A strong pilot has fixed scope, fixed time, success metrics, baseline, full-contract pricing, and a conversion decision before it starts. Use references/sales-laws/paid-pilot-structure.md.
5. The hardest objector may be the best champion. The person pushing back often has the most operational risk, authority, and accountability. Convert them with proof and they can move the deal. Use references/sales-laws/objector-as-champion.md.
6. Sell to practitioners first, then let them pull procurement. Solve personal pain, embed where work happens, and let practitioner dependency create internal pull toward budget owners. Use references/sales-laws/practitioner-pull.md.

Supplemental sales patterns exist for economic-buyer pre-close, peer-reference hierarchy, and multi-year contracts as switching costs. Load the matching files under references/sales-laws/ when needed.

Offer Patterns

Use these as starting shapes, then adapt them to the user's product:

• Design partner: You get <measurable outcome> for <10-20% of expected TCV> in <4-8 weeks>; if <pre-agreed metric> is not met, you get your money back or no full rollout.
• Paid pilot: One workflow, one owner, one baseline, one success metric, one timebox, full-contract price agreed before kickoff.
• Labor pricing: Current cost is <hours/headcount/vendor/process cost>; AI cost is <10-15x cheaper or outcome-aligned>; buyer pays only when value is delivered when possible.
• Demo: Bring the prospect's own work into the product: their docs, tickets, calls, filings, data, codebase, community, or public footprint.
• Expansion: Start with one team or workflow; expand when usage/output naturally crosses departments, seats, cases, data volume, or customer-facing surface area.

Useful Company Cases

• Sierra: design partners, 10-20% TCV upfront, outcome pricing, Fortune 500 CX, founder credibility. See references/companies/sierra.md.
• Harvey: B2C2B legal adoption, prestige-first Big Law, personalized PACER demos, legal-engineer GTM. See references/companies/harvey.md.
• Decagon: 100+ founder interviews, hard WTP filter, 4-week paid pilots, support automation. See references/companies/decagon.md.
• Gong: founder-led ICP pivot, alpha trial-close, pilots as conversion engine, category creation. See references/companies/gong.md.
• Glean: founder network design partners, paid POCs, adoption-data expansion. See references/companies/glean.md.
• Abridge: clinical trust architecture, physician-founder sales, health-system proof, Epic distribution. See references/companies/abridge.md.
• Legora: paid lawyer interviews, live-demo FOMO, reliability pause before scaling. See references/companies/legora.md.
• Deel, Wiz, Ramp, Moveworks, Writer, Hebbia, Cognition, Incident.io, Intercom/Fin, Listen Labs: load their company files when the user's situation matches payroll/compliance, security, finance, IT support, enterprise writing, research, software engineering, incident management, support, or research automation.

Output Contract

For strategy asks, answer in this shape:

1. Diagnosis: the current motion and likely bottleneck.
2. Best-matching pattern: cite the relevant law and 1-3 company cases.
3. Concrete offer: buyer, workflow, pilot terms, success metric, price anchor, and demo input.
4. Next actions: founder-owned steps for the next 1-2 weeks.
5. Risks: the main anti-pattern to avoid.

For agent-infrastructure or agent-workflow ideas, include a platform-risk verdict inside Risks: likely commoditized, survivable with reframing, or structurally defensible.

Avoid broad inspiration. The value of this skill is applying specific observed mechanics to the user's current GTM decision.