Sample frameDay 270·Growth · Acquire

Some products carry their own distribution. One number tells you whether yours is one of them.

Growth.

The Viral Coefficient

5 min read·Apply by lunch

The question

Of everyone who arrived last month, how many came because an existing customer put you in front of them rather than because you paid, posted or pitched — and how many days sat between that customer joining and the person they brought? Most founders can estimate the first number. Almost nobody has measured the second.

The idea

Day 17 taught you to build a compound loop. This is the instrument that tells you whether yours actually runs. David Skok founded four software companies, three of which went public, before joining Matrix Partners as a general partner in 2001, and on his blog, For Entrepreneurs, he wrote the piece that finally put the arithmetic down: Lessons Learned – Viral Marketing. A viral loop has two inputs. Each user sends some number of invitations, call it i, and some fraction of those invitations convert, call it c. Multiply them and you have K, the viral coefficient. Above one, each cohort produces a larger cohort and the product grows without you touching it. Below one the loop still amplifies everything else you do — at a K of 0.5, every hundred customers you buy become two hundred. Skok's real contribution was the second variable. Growth is not K on its own; it is K compounding once per viral cycle time, the gap between somebody joining and the people they bring joining. Cycle time sits in the exponent while K sits in the base. The trap is reading the coefficient without the clock: a K of 1.1 on a daily cycle turns a hundred customers into roughly eighteen thousand inside a month, while the same 1.1 on a monthly cycle turns a hundred into two hundred and ten.

The framework

01
K has two halves, and usually only one of them is broken.
Measure i and c separately or you'll spend a quarter fixing the wrong one. Invitations sent per user is a product problem — how visible the share action is, when in the experience it's offered, how many people a user can plausibly name. Acceptance rate is a message problem — who the invitation appears to come from, what it promises, and how much work the recipient does before seeing anything of value. High i with dismal c means you're spamming people on their behalf. High c with dismal i means you built something people would happily recommend and never got round to asking them.
02
How long is your cycle, measured end to end?
Cycle time runs from a user joining to their invitees joining, and it's assembled from ordinary delays: the days before you ask, the friction in the asking, the invitation sitting unopened, the sign-up they postpone until Sunday. Instrument it once and the work becomes obvious — move the invitation earlier in the first session, let a recipient see something worth having before they register, delete a confirmation step. None of that changes K at all, and all of it changes the growth rate, because you're compounding more often.
03
The saturation ceiling.
K is not a constant, and treating last quarter's figure as a permanent property of the product is how forecasts go badly wrong. Your first users invite the people most likely to want the thing — the colleague who already complained about the problem, the friend who asked what you were building — so the earliest cohorts flatter you. As the reachable network fills up, both halves fall: there are fewer obvious people left to name, and the ones who remain have already watched the invitation land in somebody else's feed. A loop that ran above one can drift below it without a line of code changing.
04
Measure K by cohort, or you will never see it fall.
An all-time K averages a rising curve and a falling one, and hides both. Compute it per cohort — this month's joiners, the invitations they sent, the ones that converted — and set the last six months in a row so the direction shows. What you're reading for is a declining K on rising volume: more customers arriving, each bringing fewer. That combination flatters a top-line chart and is the most dangerous input a forecast can take, because it assumes today's ratio survives at ten times the size. What it actually describes is a business reverting, month by month, to whatever it pays for.

Members only · 4 principles + template + AI mentor

364 more frameworks are waiting.

The Viral Coefficient is Day 1 of 365. One framework every morning for a year — across strategy, sales, negotiation, leadership, and more.

$1 /day

Billed $365/yr · cancel any time

365 frameworks

12 topics · templates · AI mentor

Reserve your place — $365/yr →

One payment a year · the whole library · no auto-renewal.

More Growth frameworks

All 24 growth frameworks →