Referral Mechanics — The 13 Metrics That Tune the Loop
Every referral program is the same loop with different levers. The 13 metrics that tune it — and the step-by-step mechanics of a referral flow that compounds.
DRAFT intro — AJ to finalize voice.
Most referral programs fail for the same reason: the team ships an incentive and ignores the loop. But a referral program is a system — a loop where every step has a metric and a lever. Get the loop right and growth compounds for free. Here are the 13 metrics worth knowing cold, and the mechanics of a flow that actually loops.
The 13 metrics
| Metric | Definition | The lever |
|---|---|---|
| K-factor (viral coefficient) | invites per user × invite→signup rate | K > 1 = self-sustaining exponential growth. The north star. |
| 30-day K-factor | same, counted within 30 days of joining | The operating target — move invite rate, invite→signup, or cycle time. |
| Viral cycle time | avg time from a user joining to their invitee joining | Shorter compounds faster; often higher leverage than raw K. |
| Participation rate | % of eligible users who send ≥1 invite | Top of the referral funnel — trigger timing + prompt placement. |
| Invites per sharer | avg invites among users who share at all | Channel choice + friction (prefill, SMS vs link vs social). |
| Invite CTR | % of invites that get clicked | Sender personalization, social proof, messaging. |
| Visit→signup | % of referred visitors who create an account | Dedicated referred landing page, incentive clarity. |
| Signup→activation | % of referred signups who reach first value | Onboarding — tie the reward unlock to activation, not signup. |
| Referral rate | % of new customers attributable to referrals | The headline "how much growth is organic." |
| Referral CAC | (incentives + program cost) ÷ referred customers | Compare to paid CAC; usually the cheap channel. |
| LTV: referred vs non-referred | retention/value of the referred cohort vs the rest | Referred users usually retain better — the quiet superpower. |
| Reward redemption rate | % of earned rewards actually claimed | Incentive design; unredeemed = friction or weak incentive. |
| Fraud / gaming rate | % of referrals flagged as abuse | Critical with cash incentives (and compliance). |
The loop, step by step (and the lever at each)
- Trigger / eligibility — prompt at the "aha" moment (post-activation, first result). Lever: timing.
- Prompt / exposure — the CTA surface, copy, frequency. Lever: placement.
- Share / invite sent — channel + friction. Metrics: participation rate, invites per sharer. Lever: prefill, fewer steps.
- Delivery / landing — the invite arrives and is seen. Lever: channel deliverability.
- Click / visit — Metric: invite CTR. Lever: social proof, sender name.
- Signup — Metric: visit→signup. Lever: dedicated referred landing page, clear double-sided incentive.
- Activation — the referred user reaches value. Metric: signup→activation. Lever: onboarding + reward unlock tied to activation.
- Reward — both sides paid. Metric: redemption. Lever: incentive design + fraud control.
- Re-loop — the new user becomes a referrer. Metric: time-to-first-invite. Lever: prompt new users early — this is what drives cycle time.
The number that ties it together is the K-factor: invites per user × invite→signup rate. Above 1, every cohort more than replaces itself and growth sustains itself.
See it in the wild: Dropbox's referral loop took it from 100K to 4M users in 15 months.
Related
Filed under Referral, Activation, Retention. See also Apple Just Proved the Lock-In Thesis, The 7 PLG Metrics That Actually Matter, The Complete Guide to Activation Rate Optimization.