launch-guide
Guide the complete product launch process across Product Hunt, social media, blogs, and other platforms. Use when preparing a launch, filling out Product Hunt submission forms, planning screenshots an
SKILL.mdskills/launch-guide/SKILL.md
GitHubLaunch Guide
Coordinate a launch that the product can support and that produces learning
after the announcement. This skill owns launch strategy, readiness, operations,
assets, analytics, and follow-through. Use product-launch-announcement-writer
for the detailed copy deck.
Operating Rules
- Inspect the product, repository, website, analytics, audience, and existing copy before asking questions.
- Verify platform rules and form fields on the execution date. Do not rely on cached character limits, categories, account requirements, or timing advice.
- Never recommend purchased votes, coordinated upvoting, review manipulation, fake accounts, undisclosed incentives, or spam.
- Do not fabricate users, logos, testimonials, rankings, scarcity, or performance claims.
- Prefer product readiness and audience fit over a supposedly ideal weekday or hour.
- Build an owned follow-up path; a platform launch is an event, not the whole go-to-market strategy.
Phase 1: Launch/No-Launch Gate
Assess:
| Area | Ready when |
|---|---|
| Product | Core path works with realistic data and known limitations are disclosed |
| Reliability | Monitoring, backups/rollback, incident owner, and status communication exist |
| Onboarding | A new user can reach first value without founder rescue, or high-touch help is planned |
| Positioning | Audience, problem, outcome, alternative, and differentiator are specific |
| Proof | Demo, real examples, or permissioned customer evidence supports the claims |
| Conversion | Pricing/CTA, signup, email delivery, attribution, and support flow work |
| Compliance | Privacy, terms, consent, claims, licenses, and platform rules are reviewed |
| Capacity | Someone can respond to users and incidents throughout the launch window |
Return launch, soft launch, delay, or cancel, with blockers and owners.
Do not delay merely for cosmetic perfection when the core learning loop is safe
and functional.
Phase 2: Define the Launch Thesis
Write:
- Audience: exact people the launch should reach;
- Message: one problem, one outcome, one differentiator;
- Primary action: try, install, join, buy, book, star, or discuss;
- Primary learning goal: the assumption this launch should test;
- Success measures: downstream behavior, not only reach;
- Guardrails: uptime, support queue, refunds, complaints, unsubscribes, and abuse.
Choose channels based on audience presence and native intent. Do not post everywhere by default.
Phase 3: Verify Platform Requirements
For each selected platform, record source URL, checked date, account eligibility, fields, limits, media specifications, promotional rules, and moderation risks.
Product Hunt
Use current official Product Hunt guidance. As of the latest documented rules, makers should use eligible personal accounts, may schedule/create a draft, can select up to three categories, and should invite people to visit and join the discussion rather than directly asking for upvotes. Current guidance describes a short description field around 260 characters, but verify the live form because Product Hunt changes its workflow.
The product must be usable or meaningfully available; a page that only collects emails may not be homepage-eligible. Review the form guide immediately before submission.
Other Communities
Read current rules for Hacker News, Reddit, Slack/Discord groups, app stores, and newsletters. A native explanatory post usually performs better and is safer than a bare link. Disclose affiliation and do not coordinate artificial engagement.
Phase 4: Build the Asset System
Create a shot list that tells one coherent story:
- hero: product and core outcome;
- problem/current workflow;
- key workflow steps;
- result or artifact produced;
- differentiator or integration;
- proof, limitations, or who it is for;
- clear next action.
Use realistic, permissioned data. Remove secrets, notifications, personal information, test accounts, and misleading dashboards. Verify current dimensions; Product Hunt has historically recommended 1270×760 gallery assets and multiple images, but the current interface is authoritative.
Prepare:
- a short demo that reaches value quickly;
- accessible captions/transcript and alt text;
- fallback static images if video/GIF fails;
- consistent product name, URL, pricing, and claims across channels;
- UTM/tagging convention and analytics event checklist.
Use first-comment templates for structure, then make the comment specific and candid.
Phase 5: Produce the Runbook
Build a time-sequenced runbook with owner, dependency, status, and fallback for:
- final smoke test and backup;
- submission/publishing order;
- analytics verification;
- email/social/community posts;
- comment and support coverage;
- incident severity and rollback;
- abuse/security escalation;
- refund or billing issue handling;
- end-of-day reconciliation.
Do not optimize solely for “12:01 AM Pacific.” Product Hunt’s day has historically followed Pacific time, but choose a launch time that preserves response coverage and verify current platform behavior.
Phase 6: Operate the Launch
During the launch:
- respond substantively rather than posting ranking updates;
- capture objections, bugs, use cases, and source attribution;
- update known limitations transparently;
- pause promotion if reliability or customer harm crosses a guardrail;
- avoid arguing with criticism; investigate and correct factual issues;
- keep a decision log for changes made under pressure.
Phase 7: Convert Attention Into Learning
Within one to three days, reconcile:
- impressions/visits → qualified visits → signup/install → activation → payment;
- retained behavior by source and segment;
- support volume, defects, refunds, and complaints;
- message/asset performance;
- qualitative themes and surprising use cases;
- what to repeat, revise, or stop.
Publish a useful follow-up where appropriate, then schedule the next product or distribution experiment. Do not present correlation as causal uplift without a valid comparison.
Output Contract
Return:
- readiness gate and blocker list;
- launch thesis and channel rationale;
- verified platform requirement table;
- asset and screenshot shot list;
- launch runbook with fallbacks;
- measurement plan and dashboard specification;
- post-launch retrospective template;
- copy handoff requirements for
product-launch-announcement-writer.