Viral coefficient
Aussi : K-factor, K, viral factor, virality coefficient, coefficient viral
The average number of new users each existing user generates through referrals. Above 1.0, growth compounds on itself and becomes exponential.
What It Is
The viral coefficient (often written as K or K-factor) measures how many new users each existing user brings in through invitations, referrals, or shared content. It quantifies word-of-mouth growth as a single number.
The standard formula is:
K = i x c
- i = the average number of invitations sent per existing user
- c = the conversion rate of those invitations (fraction that become new users)
For example, if each user sends 10 invitations and 5% convert, then K = 10 x 0.05 = 0.5.
Why it matters
The critical threshold is 1.0:
- K < 1.0: each cohort of users generates fewer replacements than itself. Referral growth decays and eventually stops. You still need paid acquisition to grow.
- K = 1.0: growth is self-sustaining but linear.
- K > 1.0: each user generates more than one new user, producing exponential (viral) growth until saturation.
Most real products live below 1.0. That is not failure: even K = 0.5 can cut customer acquisition cost dramatically by amplifying paid and organic channels. Sustained K > 1.0 is rare and usually temporary.
How it is used in practice
- Modeling growth: combined with the viral cycle time (how long one referral loop takes), K predicts how fast a user base compounds. Shorter cycle times matter as much as a high K.
- Diagnosing loops: because K = i x c, teams optimize either the number of invites (product prompts, sharing hooks) or their conversion (landing page, incentive design).
- Blended growth: real-world growth mixes viral, paid, and organic. Analysts often report an effective K that includes all referral-driven signups.
Worked Example
A B2B tool starts with 1,000 users. Each user invites 4 colleagues, and 20% accept.
- K = 4 x 0.20 = 0.8
- Cohort 1: 1,000 users invite and produce 800 new users.
- Cohort 2: those 800 produce 640.
- Then 512, 410, and so on.
Total referred users converge to roughly 1,000 x (0.8 / (1, 0.8)) = 4,000, then stop. If instead c rose to 30% (K = 1.2), each cohort would be larger than the last and growth would compound without a natural ceiling until the market saturates.
Common Pitfalls
- Measuring K over too short a window (referral loops take time).
- Ignoring churn, which offsets viral gains.
- Confusing a one-time spike with sustainable K > 1.0.