What Is Viral Coefficient in Short-Form Video?

The viral coefficient is a product-growth metric that got borrowed by social media, and the cleanest definition of it belongs to Andrew Chen, now a general partner at Andreessen Horowitz. In a widely cited essay on the metric (andrewchen.com), Chen reduces the whole idea to one question, verbatim, "For every user coming into your site, how many friends do they bring?" per Chen. The formula behind that question is standardized: First Round Review states it as "K = (average number of invites sent by each user) × (conversion rate of invitees)" (review.firstround.com). On a video feed there is no invite link, so the translation is direct, the share is the invite and the new viewer who watches because of it is the conversion. The reason the metric matters is that it turns sharing from a number on a dashboard into a growth-rate input: a post is spreading on its own only when each viewer's shares bring in more than one new viewer.

Definition

Viral coefficient, also called the K-factor, measures how many new viewers each existing viewer generates through sharing. It is the product of two numbers: the average number of invites or shares each person sends, and the rate at which those invites convert into new viewers. A coefficient above 1 means each viewer brings in more than one additional viewer, which compounds.

What It Means

The viral coefficient came from product growth, not social media, and the cleanest framing of what it measures is Andrew Chen's. Chen, now a general partner at Andreessen Horowitz, wrote a widely cited essay on the metric (https://andrewchen.com/viral-coefficient-what-it-does-and-does-not-measure/) reducing it to one question, verbatim, "For every user coming into your site, how many friends do they bring?" per Chen. The formula behind that question is standardized: K equals the average number of invites sent per user multiplied by the conversion rate of those invites, expressed by First Round Review as "K = (average number of invites sent by each user) × (conversion rate of invitees)" (https://review.firstround.com/glossary/k-factor-virality/). On short-form video there is no invite link, so the share is the invite and the new viewer who watches because of that share is the conversion. The operational point is that the K-factor reframes sharing from a vanity metric into a growth-rate input: a post is spreading on its own only when each viewer's shares pull in more than one new viewer.

Where It Shows Up in Content Work

For social media managers, the value of the K-factor is not the exact number, which short-form analytics rarely let you compute cleanly, but the question it forces before production. Would a viewer send this to a specific person, and why. Content that earns shares usually gives the sender something: social currency, practical value to pass on, emotional resonance, or a way to express identity. Treating shareability as a creative input at the scripting stage, rather than a reporting metric after the fact, is what the coefficient is useful for.

What the viral coefficient actually measures

The viral coefficient measures pass-along: how many new viewers each existing viewer generates through sharing. Andrew Chen frames it as the single question of how many friends each user brings (andrewchen.com), and the K-factor formula gives that question a number, the average shares per viewer multiplied by the rate those shares convert into new viewers, per First Round Review's expression of it (review.firstround.com).

The threshold that matters is 1. A coefficient above 1 means each viewer brings in more than one additional viewer, so the audience compounds without paid acquisition. Below 1, every wave of viewers is smaller than the last, and growth depends on a continuing outside push. The number itself is less useful on social video than the framing, because it forces the right pre-production question: does this post give the viewer a reason to recruit the next viewer.

How to calculate it, and the worked example

The formula is K = i × c, where i is the average number of invites or shares each viewer sends and c is the conversion rate of those invites. First Round Review writes it as "K = (average number of invites sent by each user) × (conversion rate of invitees)" (review.firstround.com). On short-form video, i is how many people each viewer shares the clip with and c is the share of those recipients who actually watch.

A worked example, labeled illustrative and not from any specific post: suppose 1,000 viewers each share a clip with an average of 0.4 people (most share with no one, a few share with several), and 50 percent of those recipients watch. Then i = 0.4, c = 0.5, and K = 0.4 × 0.5 = 0.2. That is well below 1, which is the normal case: most content does not spread on sharing alone. To cross 1 on those same numbers, either the share rate or the conversion rate would have to more than double, which is rare and is what a genuinely viral clip achieves.

Because short-form analytics do not expose which new viewers arrived from a specific share, the honest field proxy is share rate, total shares divided by total views, compared against your own account baseline. Share rate is not the K-factor, but a clip whose share rate is well above your baseline is the clip where the pass-along behavior the coefficient describes is actually happening. Track it monthly by format and topic rather than chasing a single absolute number.

The deeper reason not to over-trust the coefficient is that it ignores retention. Andrew Chen's essay is explicit that the metric does not tell you whether "your product is sticky" or whether new users stay (andrewchen.com). A clip can have a high share rate and still grow nothing durable if the new viewers it pulls in never follow or watch a second video. Sharing gets them in the door; retention is a separate question the coefficient cannot answer.

How to use the K-factor as a creative filter

Run the question before production, not after. For each planned post, ask Chen's question in concrete form (andrewchen.com): would a viewer send this to a specific named person, and what does sending it do for the sender. If the honest answer is that no one would forward it, the post may still be worth making for reach or saves, but it will not spread on sharing, and you should not expect it to.

Separate the share job from the save job. A post engineered for shares needs social currency or identity payoff, a reason the sender looks good or feels understood by passing it on. A post engineered for saves needs reference value to the viewer. Designing for both at once usually produces a clip that does neither well, so pick the job per post.

Read share rate against your own baseline and against retention together. A high share rate with weak follow-through means the clip spread but recruited nothing durable, which is exactly the gap Chen warns the coefficient hides (andrewchen.com). The formats worth repeating are the ones that score on both: people share them, and the new viewers they bring in stay.

Common mistakes

The first mistake is treating a viral coefficient or a high share rate as proof of durable growth. Andrew Chen's own framing of what the metric does not measure (andrewchen.com) includes whether new users stay, so a spreading clip can still grow nothing lasting if retention is weak.

The second mistake is chasing a universal K-factor target. There is no industry-wide right number on social video, and most content sits well below 1, as the worked example shows. The useful comparison is against your own baseline by format, not against a benchmark you read somewhere.

The third mistake is reporting shareability after the fact instead of designing for it. The coefficient is most valuable as a pre-production filter, the question of whether the post gives the viewer a reason to recruit another viewer, which is the input side of Chen's how-many-friends question (andrewchen.com).

Where a planning-first tool fits

Superdirector analyzes reference content in a niche to identify the emotional patterns, format structures, and topic angles associated with strong share behavior, which is useful at the scripting stage where shareability is actually decided rather than at the reporting stage where it is only counted. Whether a given idea earns the share, and the execution that makes a viewer want to forward it, is the operator's creative call.

Disclosure by Bell Chen, founder of Superdirector: the reference-analysis features mentioned here are part of the product I build. The viral-coefficient definition and formula in this piece are sourced from the linked Andrew Chen essay and First Round Review glossary; the worked example is illustrative and not drawn from any specific post. Treat the tooling note as one input among several.

Related Terms

Frequently asked questions