← Back to Skills Marketplace
bullkis1

Growth Hacker (Early Stage)

by Bullkis1 · GitHub ↗ · v1.0.0 · MIT-0
cross-platform ✓ Security Clean
452
Downloads
0
Stars
3
Active Installs
1
Versions
Install in OpenClaw
/install growth-hacker-early-stage
Description
Rapid user acquisition, viral loops, conversion optimization, and growth experiments. Use when working on: getting first users, improving signup/activation r...
README (SKILL.md)

Growth Hacker

Find the fastest path from zero to traction. Experiment ruthlessly, double down on what works.

Mindset

  • Distribution beats product in early stages
  • Measure everything, assume nothing
  • One growth lever at a time — don't dilute focus
  • Cheap experiments before expensive ones
  • Users talk to friends → that's your best growth channel

The Growth Framework

Step 1: Diagnose where you're stuck

Growth problems usually live in one of these stages:

  1. Acquisition — people don't find you
  2. Activation — they find you but don't sign up / complete onboarding
  3. Retention — they sign up but don't come back
  4. Referral — they use it but don't tell others
  5. Revenue — users but no money

Fix in order. Don't run acquisition campaigns if activation is broken.

Step 2: Pick ONE metric to move

Define the North Star Metric (NSM): the single number that best captures value delivered.

Examples:

  • SaaS: Weekly Active Users who complete core action
  • Marketplace: Successful transactions per week
  • Community: Daily posts from returning users

Step 3: Run cheap experiments first

Channel Cost Speed Best for
Reddit (organic) Free Days Technical / niche products
Twitter/X threads Free Hours B2B, dev tools, thought leadership
Cold outreach (email/LinkedIn) Free Days B2B, high-value
Product Hunt launch Free 1 day Dev tools, SaaS
Hacker News Show HN Free 1 day Dev tools, open source
Content SEO Free, slow Months Long-term
Paid ads $$ Immediate When organic is working, not before

See references/channel-playbooks.md for tactical guides per channel.

Step 4: Build the referral loop

The best growth is built-in:

  • Viral coefficient > 1 = exponential growth
  • Viral coefficient 0.5 = still worth building — cuts CAC in half

Simple referral mechanics:

  1. User invites friend → both get value
  2. "Powered by X" / "Made with X" on user output
  3. Share result to social button in product
  4. Waitlist with referral unlock

Conversion Quick Wins

Landing page (typical low-hanging fruit):

  • Single clear CTA above the fold
  • Social proof (logos, numbers, testimonials) near CTA
  • Remove nav links on landing page
  • Headline = outcome, not feature
  • Add FAQ to kill objections

Onboarding:

  • Reduce steps to first value moment
  • Pre-fill example data so it doesn't feel empty
  • Celebrate first completion ("You did it!")
  • Send email at 24h if they haven't returned

A/B Testing

Only test when you have enough traffic (>100 conversions/variant/week):

Minimum sample size per variant: 
  n = (16 × σ²) / δ²
  Rule of thumb: 100+ conversions before reading results

Tools: Vercel Edge Config + flags, Posthog feature flags, GrowthBook (OSS).

Metrics to track from day one

Acquisition: Visits, signups, CAC per channel
Activation: % completing core action within 24h
Retention: D1, D7, D30 retention
Referral: Viral coefficient (invites sent × invite conversion rate)
Revenue: MRR, ARPU, churn rate

Critical Rules

  • Never run paid ads until you know your activation rate is > 40%
  • Always track source/medium for every signup
  • Never optimize for signups — optimize for activated users
  • Always talk to churned users (not just happy ones)

References

  • references/channel-playbooks.md — Reddit, HN, Product Hunt, cold email, Twitter tactics
Usage Guidance
This skill is internally coherent and contains practical growth tactics, but review before acting: many tactics (coordinated upvotes, cold emailing, scraping contacts) can violate platform policies or anti-spam laws—use only compliant methods and obtain consent. Because the skill names third‑party services, be mindful that integrating those services will require you to supply and protect API keys or account credentials separately (the skill does not request them). If you plan to automate actions, audit the automation for policy/legal risk and limit any agent autonomy when performing outreach or platform interactions.
Capability Analysis
Type: OpenClaw Skill Name: growth-hacker-early-stage Version: 1.0.0 The skill bundle consists entirely of strategic and tactical documentation for growth hacking, providing advice on user acquisition and conversion optimization. It contains no executable code, scripts, or instructions that would lead to data exfiltration, unauthorized access, or malicious behavior by the AI agent (SKILL.md and references/channel-playbooks.md).
Capability Assessment
Purpose & Capability
Name/description align with the content. The SKILL.md and channel playbooks provide tactics for acquisition, activation, retention, referrals, and A/B testing; there are no unexpected binaries, env vars, or config paths required.
Instruction Scope
Instructions stay within growth/marketing scope (channel tactics, referral mechanics, measurement). Some recommendations (organizing upvotes, mass cold outreach, use of email-finding tools) raise ethical/platform‑policy and legal considerations (spam laws, site rules), but they are consistent with a growth playbook rather than indicating hidden or unrelated behavior.
Install Mechanism
No install spec and no code files — instruction-only skill that does not write to disk or fetch external code. This is the lowest-risk install profile.
Credentials
The skill declares no required environment variables or credentials. The guide names third-party tools (Posthog, GrowthBook, Apollo.io, Hunter.io, Resend, Beehiiv, etc.) which, if you choose to use them, will require separate credentials handled outside this skill. That is proportionate but users should be aware they must provide and protect those credentials when integrating tools.
Persistence & Privilege
always is false, no config paths or system modifications, and the skill does not request permanent presence or elevated privileges. Autonomous invocation is allowed by default on the platform but is not unusual or excessive here.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install growth-hacker-early-stage
  3. After installation, invoke the skill by name or use /growth-hacker-early-stage
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.0
0-to-traction framework: acquisition diagnosis, A/B testing, channel playbooks (Reddit, HN, Product Hunt, cold email, Twitter). Adapted from agency-agents.
Metadata
Slug growth-hacker-early-stage
Version 1.0.0
License MIT-0
All-time Installs 3
Active Installs 3
Total Versions 1
Frequently Asked Questions

What is Growth Hacker (Early Stage)?

Rapid user acquisition, viral loops, conversion optimization, and growth experiments. Use when working on: getting first users, improving signup/activation r... It is an AI Agent Skill for Claude Code / OpenClaw, with 452 downloads so far.

How do I install Growth Hacker (Early Stage)?

Run "/install growth-hacker-early-stage" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is Growth Hacker (Early Stage) free?

Yes, Growth Hacker (Early Stage) is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does Growth Hacker (Early Stage) support?

Growth Hacker (Early Stage) is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created Growth Hacker (Early Stage)?

It is built and maintained by Bullkis1 (@bullkis1); the current version is v1.0.0.

💬 Comments