How k-factor is calculated
K-factor multiplies how many invites users send by how often those invites work:
k = invites per user x invite conversion rate
A worked example: in October, 20,000 new users join. They send 50,000 invites in total, an average of 2.5 each. 6,000 of those invites become new installs, a 12% conversion rate. K-factor is 2.5 x 0.12 = 0.3.
Measure it per cohort and over a fixed window, such as 30 days after the inviter joined, so the number does not keep changing as late invites trickle in.
What k-factor does to growth
When k is below 1, each wave of users brings a smaller wave behind it. The total reach of one starting user is:
total users per seed user = 1 / (1 - k)
With k = 0.3, every paid or organic user leads to about 1.43 users in total. That turns directly into cost: a $5 CPI becomes an effective $3.50 per user. When k is 1 or higher, the waves stop shrinking and growth compounds, which is why a k above 1 gets so much attention, and why it almost never lasts as a market saturates.
Improving k-factor
- Make inviting part of the product. Splitting a bill, sharing a playlist or inviting a teammate builds invites into real use.
- Reward both sides. Credits for the inviter and a welcome offer for the invitee raise both invite rate and conversion.
- Remove friction after the click. Deferred deep links take the invitee straight to the right screen with the referral code applied, even if they had to install first.
- Shorten cycle time. Prompt invites at moments of value, not on day one before the user has seen any.
Measuring referral growth in Linkrunner
Accurate k-factor needs each referred install tied to its inviter. Linkrunner links carry a referrer parameter, and referred installs appear in a dedicated Referrals channel, separate from paid and organic. Deferred deep links carry the referral context through the install so the invitee lands on the right screen. Divide referred installs by the inviting cohort's invites and you have your k-factor per cohort.