Launch content and distribution plan

Announcement post (outline)

SignalQL is an open query language for behavioral product analytics. Teams describe questions in a small, predictable grammar; the Postgres reference path compiles to parameterized SQL, while other adapters document their own safety contract, so AI tools and humans share the same analytics semantics without locking analytics to a single vendor.

Draft angles

1. Use SignalQL with Cursor to analyze your app — link Cursor guide, grammar pack, MCP stdio command.
2. Let Claude analyze product data — Claude MCP setup + first successful query path.
3. Replace dashboards with AI + SignalQL — emphasize structured SignalQL over prose→SQL; pair with playground.

Launch checklist

Area	Check
Docs site builds (npm run docs:build)	VitePress bundle green
Playground (apps/playground)	Compiles examples offline
Repo (README, LICENSE, CONTRIBUTING)	Present and accurate
MCP (@signalql/mcp)	stdio tools respond
Grammar + prompting docs	Linked from nav
Neutral integration examples	Follow docs under integrations/

Distribution channels

• GitHub repository public release with tagged v0.1.x packages (when publishing to npm).
• Documentation at signalql.org (build artifact from docs/).
• Blog/social posts referencing docs and playground URLs—avoid duplicating long prose; link canonical docs.

Versioning

• Language v0.1 tracks docs/spec/v0.1.md and schemas/signalql-ast-v0.1.schema.json.
• npm packages may use semver independently; note supported language version in each package README.

Distribution targets

• Developers & data folks: GitHub, Hacker News–style launch post, and X/Twitter thread linking to the spec and playground.
• AI tooling users: Cross-post short “Cursor + SignalQL” and “Claude MCP” tips to communities that already discuss MCP and coding agents.
• Product analytics: Keep messaging neutral (open language, not a single-vendor product).

CTA copy bank

Use these as reusable button/link lines and one-line support copy across docs, README, and launch posts.

Developers

• Run your first SignalQL query — Go from question to deterministic SQL output in under a minute.
• Browse SignalQL examples — Start with copy-paste analytics questions and expected shapes.
• Compile SignalQL with CLI — Keep intent readable while generating deterministic SQL.

AI tooling users

• Use SignalQL with Cursor — Ask for SignalQL, validate with MCP tools, and keep semantics explicit.
• Use SignalQL with Claude — Generate queries from your event taxonomy, then compile safely.
• Use SignalQL with ChatGPT — Prompt with grammar + schema context, then validate locally.

Analytics engineers

• Adopt SignalQL in your stack — Map once to the portable model and reuse queries across workflows.
• Ship bounded analytics queries — Use explicit limits and allowed aggregates for safer ad hoc analysis.
• Contribute a dialect implementation — Track v0.1 AST bounds and document dialect differences.

CTA funnel checks

Track CTA outcomes with lightweight, activation-focused checks.

Entry points to monitor

• Docs home hero actions (/): run-first click, no-setup click, spec click.
• README Start here links: playground path, AI path, CLI/SDK path.
• Integration guide close-out blocks: "Run this query now" completion.

Baseline and comparison

• Record current weekly counts for each entry point before CTA rollout.
• Compare 2-week and 4-week post-rollout windows against baseline.
• Note source channel for each jump (docs home, README, integrations, social launch links).

Activation outcomes

• First-query completion: user reaches a compiled query output with params.
• Guide completion: user follows one integration path end-to-end.
• Repo action signals: issue comments, external PRs, and setup-related questions dropping over time.

Success metrics to watch

• GitHub stars and clone traffic on the public repo.
• Doc site unique visitors and time on the example library and AI guides.
• Reported issues and external PRs against the spec or compiler.
• Qualitative: third-party blog posts or videos referencing SignalQL as an open standard.