newsletter_referral_programs v1.0.0

name: "newsletter_referral_programs" description: "Design and run a newsletter referral program from scratch: tier/reward design with real cost math, fraud control, leaderboard mechanics, launch sequencing, and tracking. Triggers on 'newsletter referral program', 'referral tiers for my newsletter', 'stop referral fraud', 'referral leaderboard', 'launch a newsletter referral program', or 'subscriber referral rewards'."

Newsletter Referral Programs

Purpose

Build the referral engine that grows a newsletter subscriber base: a tiered reward ladder priced so each referred subscriber costs less than paid acquisition, a fraud screen that keeps the list clean, a leaderboard that turns top referrers into promoters, and a launch sequence that gets the flywheel turning. This skill covers the referral program itself in depth. It does not cover lead magnets, cross-promos, or welcome sequences (see newsletter_growth_loops for those).

Workflow

1. Size the economics first

Compute what each referred subscriber can cost and still beat paid channels.

Max reward spend = (Subscriber LTV x Gross Margin) - Target CPA
Per-tier CPA    = (Reward cost + fulfillment cost) / Referrals needed

Rules of thumb, verified against Morning Brew's published program (Tyler Denk):

• Tier 1 should unlock at 1-3 referrals and cost you almost nothing: bonus content, a community invite, early access. Morning Brew's first reward (Sunday edition Light Roast) cost $0 to produce.
• Low-cost physical rewards sit in the mid tiers. Morning Brew's stickers cost ~$1.25 mailed for 5 referrals, a $0.25 CPA versus $3-5 paid CPA. At 15 referrals their total reward cost ($3.40 mailed) still landed at a $0.23 CPA, cheaper per-subscriber than the tier below.
• Keep tier CPA flat or falling as tiers rise. If a higher tier raises CPA, reprice the reward, not the tier.
• Never let total program CPA exceed your paid-channel CPA. That is the ceiling for the whole design.

Run bin/tier_calculator.py with your numbers before finalizing any ladder. See references/tier-design.md for the worked Morning Brew example.

2. Design the tier ladder

• 3-5 tiers is the sweet spot. More tiers than that and nobody can hold the ladder in their head.
• Put the first reward close (1-3 referrals). The first win teaches people the game is winnable.
• Space mid tiers 2-3x apart (3, 10, 25 is the common newsletter default, per beehiiv's own launch checklist).
• Make the top tier genuinely aspirational: status, access, a call, dinner. Things money can't buy beat things money can buy at the top.
• Show progress in every issue: a dynamic meter in the footer that says "you're 2 referrals from your next reward." Show total referrals and referrals-to-next-tier on the referral page.
• Tease the next tier inside every reward-unlock email. The redemption moment is the highest-motivation moment you'll ever get.

See references/tier-design.md for ladder templates and references/leaderboard-design.md for the leaderboard layer.

3. Lock down fraud before launch

A referral program with no abuse controls will fill your list with fake subscribers who trash your deliverability. Require all of these from day one:

1. Double opt-in for referred subscribers. Unconfirmed referrals stay in a pending state and never count toward rewards (this is how beehiiv's referral program works by default).
2. Disposable-email blocklist on the signup form. Block the known temp-mail domains, keep real free providers (Gmail, Yahoo, Outlook).
3. Rate limiting: cap signups per IP per hour and flag bursts. Track signups per referrer and investigate spikes.
4. Honeypot form field: hidden field bots fill, humans don't. Zero friction, real signal.
5. Self-referral rules: the same email can't refer itself; watch for referrer/referred email similarity and shared IPs within a referral chain.
6. Delay fulfillment: ship rewards 7-14 days after the tier is hit so fake referrals that bounce or get filtered don't trigger payouts.

See references/fraud-controls.md for the full pattern list (fake-signup farms, email-forwarding abuse, incentivized click rings) and how to catch each one.

4. Run the leaderboard right

• Leaderboards motivate the top 1% and can demotivate everyone else. Fix this by ranking in cohorts (this week, this month, all time) so new referrers see a board they can still climb.
• Update in near real time. A weekly board that lags feels dead.
• Make ranking positions shareable: "I'm #3 this month, help me hit #1" posts are free promotion. Public share links per position.
• Reward status, not just stuff: top-10 spots get a name in the newsletter, a community role, a call with the author. Status costs nothing and scales forever.
• Run 30-day sprints: temporary leaderboards with one juicy prize, reset monthly. Sprints re-engage dormant referrers.

See references/leaderboard-design.md.

5. Launch in sequence

1. Seed (week -2): turn on tracking and referral links for existing subscribers before announcing. You want attribution working before the first share.
2. Soft launch (week -1): announce to your most engaged segment (top 20% openers). Let them find the bugs.
3. Public launch (week 0): dedicated issue + footer CTA in every issue. Every subscriber should know the program exists within two sends.
4. Watch and tune (weeks 1-4): check the fraud dashboard daily, confirm tier CPAs against projections, adjust tier spacing if a tier is empty or a reward is over-subscribed.

See references/launch-sequence.md for the full checklist.

6. Track what matters

Per referrer: referrals sent, confirmed, pending, reward tiers hit, fraud flags. Per program: referral share of total growth, blended referral CPA, referred-subscriber open rate vs organic, fraud rate (rejected / total).

Alert thresholds: fraud rate over 5% means your controls are leaking; referred open rate below organic by 10+ points means the program is pulling in the wrong audience.

Output Contract

A finished referral-program design ships with:

• The tier ladder: referrals required, reward, and per-tier CPA at each step (from the calculator).
• The fraud control list with exact rules and thresholds.
• The launch sequence with dates and owner for each step.
• The tracking spec: events to log, dashboard fields, alert thresholds.

Operating Rules

1. Every factual claim about a case study, number, or platform behavior must come from a verified source. Unverifiable numbers get cut, not published.
2. Never promise a CPA without running the calculator against the newsletter's real LTV and fulfillment costs.
3. Fraud controls are not optional and ship with the launch, not after the first abuse wave.
4. Style: plain talk, no em dashes, no filler. If a sentence sounds like a press release, rewrite it.