18 Referral Program Global

What the source does

• FilesystemNone observedNo reads or writes in the source
• NetworkNone observedNo outbound calls in the source
• ShellNone observedNo commands run in the source
• DependenciesPinnedNothing floats
• SecretsNone foundNo keys in source

Five audit layers

Gates applied: no behavioural pass.

Audit 0.4.1 · source sha d435c7de76d1 · 15 Oct 2026Report an issue or request a re-scan

• Publisherminhnv0807 GitHub userOwner of the repository this skill is read from
• Source folderminhnv0807/ai-business-skills/skills/en/18-referral-program-globalThe exact folder we audited
• LicenceMITAs the repository states it
• Commit pinned for this verdictd435c7de76d1A new audit runs when the source changes
• Repository statusNot archivedLast push 32 days ago
• Sandboxed runNot runSkills are read, not executed; layer L4 is skipped

Its description: Use when customers should bring other customers — one-way versus two-way incentives, reward structure and economics, trigger moments, referral messaging, tracking and attribution, fraud controls, and anti-spam…

---
name: 18-referral-program-global
description: "Use when customers should bring other customers — one-way versus two-way incentives, reward structure and economics, trigger moments, referral messaging, tracking and attribution, fraud controls, and anti-spam compliance by region: TCPA for US, GDPR for EU, PDPA for SEA, LGPD for LATAM. Trigger on 'referral program', 'refer a friend', 'word of mouth', 'viral loop', 'how do I get customers to bring friends', 'affiliate rewards'. Also use when the user has happy customers and no system to use them. Not for — running your own community space, see `28-community-building-global`; posting inside third-party communities, see `38-community-seeding-global`; email flow mechanics, see `14-email-marketing-global`."
metadata:
  version: 1.0.1
  category: operations
license: MIT
triggers:
  - "referral program"
  - "refer a friend"
  - "word of mouth"
  - "viral loop"
  - "referral marketing"
related:
  - product-marketing-context-global
  - 14-email-marketing-global
  - 27-personal-brand-monetize-global
  - references/global-legal-compliance
---

# Referral Program (Global)

> Word of mouth is the highest-LTV acquisition channel in every region. But the LEGAL framework around how you contact referred prospects differs HUGELY: TCPA (US, SMS), GDPR (EU, all channels), PDPA (SEA), LGPD (LATAM). Pick the right region variant or get fined.

---

## For newbies

### Who is this skill for?

| Audience | Concrete example |
|----------|------------------|
| DTC brand wanting cheaper acquisition | Already at USD 30 CAC; want USD 10 CAC via referral |
| SaaS adding viral loop | Existing PMF; want negative CAC growth |
| Service business (coaching, agency) | High-LTV; want client referrals |
| Subscription brand | High retention; turn customers into ambassadors |
| E-commerce wanting AOV growth | Refer a friend = both get discount |

### Who is this NOT for?

- **Vietnam-only referral** -> Use `18-referral-program` (VN skill) — Zalo / Messenger optimized
- **B2B enterprise sales** -> ABM / partnership programs are different motion (not covered here)
- **Brand ambassador / affiliate** -> Use `27-personal-brand-monetize-global` for influencer-affiliate (when available)

### 30-second pre-read

This skill produces ONE referral program design with 6 components: model selection (1-way / 2-way / multi-tier), incentive math (% of LTV), tracking infrastructure, anti-fraud measures, launch sequence, and KPIs (K-factor, viral coefficient). Pick 1 of 4 region variants — the variant tunes the LEGAL rules for contacting referred prospects (especially via SMS/email).

### 3 common errors

1. **SMS-based referral in US without TCPA consent** -> Up to USD 1,500 per text fines + class actions
2. **Email-blast referred contacts in EU** -> GDPR violation; referral programs touching EU prospects need explicit consent from the prospect, NOT just the referrer
3. **Cash incentives that violate FTC endorsement rules** -> "Refer a friend, get USD 100" requires disclosed material connection if referrer posts publicly

---

## Why do you need this skill?

Without proper referral design:
- US: Risk TCPA class action (USD 500-1,500 per message)
- EU: GDPR violation if you store referred-prospect data without their consent
- SEA: PDPA Singapore strict — most referral programs need both-side consent
- LATAM: Brazil LGPD treats referred contacts as data subjects requiring consent
- Universal: incentive math wrong -> losing money instead of growing
- Universal: no anti-fraud -> 30-50% of "referrals" are self-referrals or bots

Plan the legal foundation correctly, get the incentive math right, ship a working viral loop.

---

## Workflow

```
Step 0: Check global context file
    |-- exists -> read product / customer / region
    |-- missing -> suggest user run product-marketing-context-global first
Step 1: Pick region variant (US / EU / SEA / LATAM)
Step 2: Confirm prerequisites (NPS, AOV, LTV, customer base)
Step 3: Choose model (1-way / 2-way / multi-tier affiliate)
Step 4: Calculate incentive (15-25% of LTV)
Step 5: Set up tracking + anti-fraud
Step 6: Design referral flow (7 steps)
Step 7: Launch sequence (30-day plan)
Step 8: Measure K-factor / viral coefficient
```

---

## Step 0: Check global context

Check `.agents/product-marketing-context-global.md`:
- **Yes** -> Read product, customer, region. Do NOT re-ask.
- **No** -> Suggest running `product-marketing-context-global` first.

---

## Step 1: Pick region variant

Ask: **"Which is your PRIMARY region: US, EU, SEA, or LATAM?"**

```
Where do most of your customers (and their referrals) live?
    |-- US / Canada       --> 01-us.md    (TCPA SMS rules; CAN-SPAM email; CCPA data)
    |-- EU / EEA / UK     --> 02-eu.md    (GDPR consent for ALL channels)
    |-- Southeast Asia    --> 03-sea.md   (PDPA per country; mostly opt-in)
    |-- Latin America     --> 04-latam.md (LGPD Brazil; LFPDPPP Mexico)
    |-- Vietnam only      --> Use `18-referral-program` (VN skill)
```

---

## Step 2: Prerequisites — does referral make sense?

### When referral works

- NPS >= 40 (customers actively like you)
- Customer has natural reason to share (visible result, social currency, peer-relevant)
- AOV high enough to fund meaningful incentive (USD 50+ ideal)
- LTV high enough to justify CAC investment
- Existing base of 100+ happy customers to seed

### When referral does NOT work (skip this skill)

- NPS < 20 (customers don't like you yet — fix retention first)
- Sensitive product category (financial advice, intimate health) — referrals feel weird
- Very low AOV (< USD 10) — incentive economics don't work
- Pre-launch or no customer base — no one to refer

### Ask the user

1. Product type? (DTC / SaaS / Service / Subscription)
2. Average AOV and LTV?
3. Existing happy customer count?
4. Goal: more new customers, lower CAC, or higher engagement?

---

## Step 3: Referral models

### Model 1: One-way (referrer gets reward, referee gets nothing)

**When:** 

Excerpt, first 6,000 characters of SKILL.md, verbatim from the audited commit. It is text from someone else’s repository, quoted so you can judge it before your agent reads it.

git clone https://github.com/minhnv0807/ai-business-skills
git -C ai-business-skills checkout d435c7de76d1

Works in

1. 2026-10-15
   ✓SAFE B 89 / 100

   first audit · source d435c7de76d1

Questions about 18 Referral Program Global

• github.com/minhnv0807/ai-business-skills/tree/master/skills/en/1 — the audited source at commit d435c7de76d1 and every finding aboveObserved · 15 Oct 2026

Spotted something wrong? Report a correction · Methodology