growth-loops

Design growth loops (Reforge/Balfour framework). Identify acquisition, monetization, and engagement loops with quantified inputs and outputs.

SKILL.md

GitHub

Growth Loops

When you activate

Product is past PMF. User asks: "how do we grow X?", "design growth loops for Y", "channel strategy after PMF"

What you produce

Saved to products/<name>/launch/growth-loops.md:

## Growth Loops — <product>

### Loop 1: <name> — Acquisition / Engagement / Monetization
**Type:** Viral | Content | Paid | Sales | UGC

**Steps:**
1. <action> → 2. <action> → 3. <action> → 4. <action that reinvests in step 1>

**Inputs needed:**
- <e.g. 100 new users/month from cold outreach>

**Outputs produced:**
- <e.g. 30 referrals/month from happy users>

**Loop multiplier:** <e.g. each user generates 0.3 new users = 0.3x compounding>

**Bottleneck:** <which step has the worst conversion or longest time>

**Investment to test:** <hours/dollars to validate the loop works>

### Loop 2: ...

### Loop interaction
How loops feed each other. E.g. Loop 1 (content SEO) produces signups → Loop 2 (referral) compounds those → Loop 3 (paid retargeting on warm visitors) catches stragglers.

### What we deliberately don't do
1. Paid ads before organic loops are proven
2. <other deferrals>

### Measurement
For each loop:
- Loop completion rate (% users who finish all steps)
- Loop cycle time (how long start → reinvest)
- Loop value (revenue or users generated per cycle)

Protocol

  1. Read the product's metrics.md and prd.md.
  2. Identify which loop types fit the product. Not every loop type works for every product:
    • Viral: requires natural network involvement (collaboration tools, social products)
    • Content (SEO): requires evergreen high-intent search demand
    • Paid: only when CAC < LTV is provable AND organic loops are saturated
    • Sales: B2B with deal size justifying high-touch
    • UGC: requires users producing artifacts worth indexing
  3. Design 2–3 loops, max. More = unfocused.
  4. Quantify inputs and outputs. Loops without numbers are aspirations.
  5. Identify bottleneck per loop — that's the experiment queue.

Sources

  • factory/playbooks/scale-stage/growth-loops-reforge.md
  • factory/playbooks/scale-stage/100-to-1000-customers.md

What you don't do

  • Don't design a viral loop for a product that isn't inherently social.
  • Don't recommend paid before organic loops are proven — you'll burn cash.
  • Don't paste "AARRR funnel" as a loop. Funnels are linear; loops compound. They're different mental models.