tl;dr: K-factor, or viral coefficient, shows how many new users one current user brings in on average during one viral cycle. The formula is K = I × Conv%, where I is the number of invitations per user and Conv% is the invitation-to-new-user conversion rate.
Updated August 10, 2026: I cleaned up the wording, the model limitations, and the example.
Successful founders understand perfectly well how important it is for a startup to build a customer base—users, subscribers, members, and so on—with as little time and money as possible. Viral growth helps with that by bringing in far more customers without increasing ad budgets. Getting real viral growth is difficult, though, and very few companies actually achieve it.
Online marketing has two key viral-growth metrics: K-factor, also called the viral coefficient, and viral cycle time. Together they show how the customer base changes over time.
What Is the Viral Coefficient (K-Factor), and How Do You Calculate It?
Imagine you are launching a new company and your growth plan relies on bringing in customers through viral growth.
It starts with a few friends. They become your first customers and invite their friends, who then do the same thing. You get a chain reaction.
The model might start with these parameters:
- Cust(0) = 10 — the initial number of customers, step 0
- I = 10 — the number of invitations sent to potential customers
- Conv% = 20% — the response or conversion rate
The first thing we need to calculate is how many new customers each existing customer can bring in. This is the extremely important variable known as the viral coefficient, or K-factor. Its formula is simply the number of invitations multiplied by the conversion rate:
K = I × Conv%
Now let us look at customer growth during one viral cycle. Our first 10 customers each sent 10 invitations and got a response from 20% of the invitees, or two new customers per person. The first cycle therefore brings Cust(1) = 10 × 2 = 20 new customers, while the total customer count becomes Cust(0) + Cust(1) = 10 + 20 = 30.
In later cycles, the inflow of new customers grows like this:
Cust(i) = Cust(i−1) × K, where i is the cycle number and Cust(i−1) is the number of new customers from the previous cycle.
Adding all customers at each step gives the total user count. If customers stay loyal and K = 2, the sequence is 10, 30 (10 + 20), 70 (30 + 40), 150 (70 + 80), 310 (150 + 160), 630, 1,270, and so on. You can put these formulas into any spreadsheet and run your own numbers.
To understand the model fully, it helps to look at later growth cycles. Notice that only customers from the previous cycle, Cust(i−1), send invitations in the formula. That is because every customer is unlikely to keep sending invitations in every cycle.
How K-Factor Affects the Viral Growth Cycle
According to this formula, viral growth becomes more dynamic as K-factor increases:
- if K-factor is above 1, the growth curve is exponential;
- at K = 1, you can expect only linear growth;
- if K is below 1, viral growth fades and the total customer count eventually levels off. External factors such as advertising can still produce growth, but it is no longer viral.
Fine, Give Me a K-Factor Example
Suppose we buy 10,000 activated installs at $2 each. An average user sends 0.4 invitations, and 25% of them convert into installs. We get K = 0.4 × 0.25 = 0.1: the first wave brings 1,000 installs, then 100, then 10, and that is basically it. This is not always a sad ending, by the way. For media buys with strong early ROAS, it is simply extra money: even low virality brings another 1,111 installs and lowers eCPI to $1.80. Everything counts.
Viral Cycle Time
A viral cycle can be broken down into these stages:
- A customer discovers your app.
- They decide to use it.
- They like it enough to share it.
- They send an invitation to their friends.
- The people who accept the invitation become interested in the app, and the cycle starts again.
The total time needed to complete these stages is the viral cycle time. Viral growth is limited by the speed of that cycle, which means you should optimize it wherever you can.
You can see how strongly cycle time affects viral growth in this formula:
Cust(t) = Cust(0) × (K ^ (t/ct + 1) − 1) / (K − 1),
where Cust(0) is the initial number of customers, t is the time frame in days, K is the viral coefficient, and ct is the viral cycle time.
If you make the cycle very fast, the result can be dramatic. With K = 2 and 10 starting users, after 20 days we get:
- about 20,000 users when ct = 2;
- more than 20 million when ct = 1.
This effect helps explain why cycle speed matters so much for products such as YouTube and Facebook, but it obviously does not explain their success on its own.
K-Factor Limitations
K-factor tells you nothing about user quality. You can get a beautiful viral-growth curve and then discover terrible retention and enormous churn. That happens all the time. Growth also slows on its own as the audience becomes saturated. Compare K-factor only across cohorts from the same source and period; otherwise a change in traffic quality or the measurement window can easily masquerade as virality.
How to Improve K-Factor
The formula is simple: number of invitations times their conversion rate. Thank you, Captain Obvious. In practice, that means working on either the number of invitations or their conversion. To improve I, make the invitation flow simple and tempting: users should want to show off or pull other people into the product. Conversion is the usual story—a clear offer and value, a simple landing page and onboarding, and the shortest possible time to activation.
Viral Properties of a Product
It is important to understand that marketing tactics or strategies cannot magically give any product a viral character. Those properties need to be considered while the startup, app, or other product is being created. Virality has to be designed carefully and implemented well because the product’s success can depend on it. Put differently: a product needs viral potential. Without that, virality is hard or impossible.
The most viral products are often those that need existing social groups to spread successfully. In other words, the product becomes most valuable when users can invite their friends. Social networks such as Facebook, Instagram, and Pinterest, as well as messengers such as Skype and Viber, are examples.
K-Factor Takeaways
K above 1 is a sign of self-sustaining viral growth within this simplified model. If K is below 1, each new wave fades without an external inflow.
All else being equal, shortening the viral cycle can accelerate growth more than a small increase in K-factor. Compare their effects over the same time horizon and across equivalent cohorts.
Viral marketing works best for a product whose viral properties were designed from the start. If those properties are too weak or have run their course, it is time to switch to other marketing methods such as SEO or SMO. You can find the other terms in the CPA glossary.
Viral marketing has enormous potential, but that potential is still finite.