The viral coefficient (K-factor) equals the number of invites each user sends multiplied by the percentage of those invites that convert into new active users. When K is greater than 1, every existing user replaces themselves with more than one new user, and growth compounds without paid spend. In practice, K sits below 1 for almost every product — the real engine is amplification, not self-sustaining virality.
Quick Answer:
K = invites sent per user × conversion rate of those invites. K > 1 means a self-sustaining viral loop; K just under 1 means each user still adds meaningful "free" growth on top of paid or organic acquisition, but the loop dies out on its own without new seed users.
What Is the Viral Coefficient and Why Does It Matter
The viral coefficient (K-factor) measures how many new users each existing user generates through invites, referrals, or shares, before any paid acquisition is added. It matters because it tells a founder or growth PM whether the product genuinely grows itself, or whether "word of mouth" is actually just marketing lift dressed up in growth-team language.
Board decks love the word "viral." It implies growth that costs nothing and compounds forever. But K-factor is a precise, falsifiable number, and most teams that say "we're going viral" have never actually calculated it. That gap between vibe and math is exactly where growth PMs earn their keep.
Three things K-factor tells you that vanity metrics don't:
- Whether user growth is self-funding (K > 1) or merely self-assisting (K < 1 but meaningfully positive).
- Where the loop is weakest — the invite rate, or the conversion rate — so you know what to fix.
- Whether a spike in signups is a real structural shift or a temporary seeding event (a launch, a press hit) that will decay once the seed cohort stops inviting.
Ignoring K-factor doesn't just cost you a vanity metric — it corrupts the growth model underneath your activation metric and your retention forecasts, because both assume you know where new users are actually coming from.
How to Calculate K-Factor: The Formula and a Worked Example
K-factor equals invites sent per existing user multiplied by the conversion rate of those invites into active new users: K = i × c. Run this per cohort, per time window (typically 30 or 90 days), never as a lifetime average — averages hide whether the loop is accelerating or decaying.
The Formula, Broken Down
i(invites per user): total invites sent by a cohort, divided by the number of users in that cohort. Not "invites available" — invites actually sent.c(conversion rate): of those invites, the percentage that result in a new user who completes signup and reaches an active state (not just clicks a link).K = i × c: the average number of new active users each existing user directly produces.
Worked Calculation
Take a cohort of 1,000 users on a collaborative tool. Over 30 days, they send 2,500 invites combined — that's i = 2.5 invites per user. Of those 2,500 invites, 180 recipients sign up and become active — a c = 0.072 (7.2%) conversion rate.
K = 2.5 × 0.072 = 0.18
| Variable | Value | What it represents |
|---|---|---|
| Cohort size | 1,000 users | Starting population |
| Invites sent | 2,500 | Total outbound invites in 30 days |
i (invites/user) | 2.5 | Invites ÷ cohort size |
| New activations from invites | 180 | Invited users who reached active status |
c (conversion rate) | 7.2% | Activations ÷ invites sent |
| K-factor | 0.18 | i × c |
At K = 0.18, each original user generates roughly 0.18 additional active users through invites alone. That's a real, useful growth contribution — but nowhere near self-sustaining. A K of 0.18 needs continuous injection of new seed users (paid, organic search, content) to keep the total growing; left alone, the invite loop shrinks with each generation, since 1,000 × 0.18 = 180, then 180 × 0.18 = 32, and so on toward zero.
What K Greater Than 1 Actually Requires
For K to clear 1.0, you need either a much higher invite rate, a much higher conversion rate, or both. Holding conversion at 7.2%, you'd need i ≈ 14 invites per user — a number almost no product sustains without spam-like behavior or a use case (like a payment app or a document you must share to collaborate) where sending is functionally required, not optional.
This is why K > 1 is genuinely rare outside a narrow set of categories: communication tools (each message requires a recipient), payment apps (you must invite someone to pay them), and early-stage social networks in a still-uncrowded niche. Andrew Chen, who studied this extensively while at a16z and previously at Uber, has written that true K > 1 virality is closer to a structural property of the product category than a marketing achievement you can bolt on.
K Greater Than 1 vs. K Just Under 1: Two Very Different Growth Stories
A K above 1 means the product grows on its own, compounding without new paid spend, while a K below 1 (but meaningfully positive) means invites amplify — but don't replace — your paid and organic acquisition. Confusing the two leads teams to defund paid acquisition on the mistaken belief that "virality" will cover the gap, which it structurally cannot at K < 1.
The Compounding Math
| K-factor | Growth behavior | 5-generation outcome from 1,000 seed users |
|---|---|---|
| 1.5 | Exponential, self-sustaining | ~7,594 new users from invites alone |
| 1.0 | Flat replacement, borderline | 1,000 new users per generation, indefinitely |
| 0.5 | Decaying, needs reseeding | 1,000 → 500 → 250 → 125 → 63 (converges near 2,000 total) |
| 0.18 | Amplification only | 1,000 → 180 → 32 → 6 → 1 (converges near 1,220 total) |
The takeaway: below K = 1, the invite chain is a geometric series that converges to a finite total, not a runaway loop. You can compute that total directly — it's seed users ÷ (1 − K). At K = 0.18, 1,000 seed users eventually produce about 1,220 total through invites (the extra ~220), then the loop is spent. That's a real, worthwhile amplifier on top of paid acquisition. It is not, on its own, a growth strategy.
Why Most "Viral" Products Are Actually Amplification Loops
The sober reality: most products that get called viral internally are running K somewhere between 0.15 and 0.4, layered on top of paid acquisition, content/SEO, or sales — not a standalone K > 1 engine. Sean Ellis and Morgan Brown's Hacking Growth explicitly frames this as amplification, distinct from true virality, and argues most "growth loops" praised in case studies are amplification loops with good instrumentation, not organic K > 1 miracles.
This distinction changes what you tell the board:
- True K > 1: "Growth is structurally self-funding; paid acquisition is optional, used only to accelerate the curve."
- Amplification (K < 1): "Every paid or organic user we acquire brings roughly
K ÷ (1 − K)bonus users for free — that's a real CAC discount, not a replacement for CAC spend."
Saying the second thing honestly, with the actual multiplier attached, is a far stronger board narrative than an overclaimed "we're viral" that gets falsified the first quarter paid spend pauses and growth stalls.
Where K-Factor Fits in Your Broader Growth Model
K-factor is one input into a growth model, not a replacement for one — it describes only the invite-and-conversion loop, and says nothing about whether invited users actually retain. A high K-factor feeding low-retention signups is a leaky bucket with a fast tap, not a healthy engine.
Three connections growth PMs consistently underweight:
- K-factor without an aha moment is hollow. If invited users convert to signup but never reach the product's core value moment, your
c(conversion rate) is measuring account creation, not real activation — inflating K on paper while retention quietly rots. - Time to value shapes conversion rate directly. An invite that lands a recipient in a confusing, slow-to-value product converts worse than one that lands them in an obvious first win — so shrinking time-to-value is itself a K-factor lever, not a separate initiative.
- The invite trigger point is a customer journey design decision. When and how you prompt an invite (immediately after signup vs. after a first success) materially changes
i, and mapping that moment against the user's emotional state is exactly the kind of analysis a journey map is built for.
For the fuller picture of how K-factor sits alongside activation, retention curves, and cohort-based growth accounting, the growth and retention field guide walks through how these metrics compose into one model instead of competing dashboards.
Making the K-Factor Loop Visible Instead of Abstract
The hardest part of K-factor in practice isn't the formula — it's that the loop is a chain of dependent steps (user acts, invite sends, recipient sees, recipient converts, new user acts again) that's easy to describe in a spreadsheet and hard to actually see as a system with feedback and decay.
This is the kind of structure Prodinja's Systems Engineering tool is built for: you describe the invite-and-conversion chain as a causal loop, and the tool renders it as a feedback diagram, making each multiplier — the invite rate, the conversion rate, the decay across generations — a visible node instead of a buried spreadsheet formula. Seeing the loop drawn out makes it far more obvious, at a glance, whether you're looking at a self-sustaining loop or a well-instrumented amplification effect, before you build a board slide around the wrong one.
That framing matters because K-factor is fundamentally a systems problem disguised as a single metric — a small change in conversion rate at the top of the chain compounds (or decays) through every subsequent generation, exactly the kind of dynamic that's easy to miscalculate in a static spreadsheet and easier to reason about once it's drawn as a loop.
Key Takeaways
- K-factor equals invites per user multiplied by conversion rate (
K = i × c) — calculate it per cohort and time window, never as a lifetime average. - K greater than 1 means self-sustaining growth: each user replaces themselves with more than one new user, compounding without added paid spend.
- K greater than 1 is rare outside communication, payment, and early social-network categories, where sending an invite is functionally required by the product's core use, not optional.
- Below K = 1, invite chains converge to a finite total (
seed users ÷ (1 − K)), not a runaway loop — useful amplification, not replacement growth. - Most "viral" products run K between roughly 0.15 and 0.4, layered on top of paid or organic acquisition — an amplification loop, not true virality.
- K-factor is hollow without retention context — pair it with your activation metric and time-to-value data before presenting it to a board.
Frequently Asked Questions
How do you calculate K-factor for a mobile app?
Calculate K the same way as any product: count invites sent per user over a fixed window (30 or 90 days), divide the resulting new activations by invites sent to get conversion rate, then multiply the two. Mobile apps often add a second layer — organic app-store discovery driven by installs — which is a separate metric (sometimes called viral lift), not part of K itself.
What is a "good" viral coefficient?
There's no universal good K-factor, because the right benchmark depends entirely on category and cost structure. A K of 0.15–0.3 is a strong amplification result for most B2B SaaS products; a K near or above 1 is exceptional and typically limited to communication or payment-style products where invites are structurally required to use the product at all.
Is K-factor the same as virality or a viral loop?
K-factor is the number that measures a viral loop's strength; the loop itself is the mechanism (invite trigger, message, landing experience) that produces invites and conversions. You can improve K by redesigning the loop — where and how an invite is prompted — without touching the underlying formula.
Why does my K-factor look high but growth still isn't compounding?
A K-factor that looks high but doesn't compound usually means the conversion rate is counting signups, not real activated users, or that invites are concentrated in a small power-user segment that isn't representative. Recheck your conversion definition against your actual activation metric before trusting the number.
Can you increase K-factor without spamming users?
Yes — the more durable lever is usually redesigning the invite moment around genuine utility, such as inviting a collaborator to a document you must share to get value, rather than increasing invite volume or incentive size. Products in the Jobs-to-be-Done tradition tend to find these natural invite points by mapping the underlying job a user is hiring the product to do, covered in the Jobs-to-be-Done field guide.